Timestamps in the description as well. The Patrons Ask GN episode is now live: www.patreon.com/posts/35658549 -- thanks to all our supporters for the help, especially during this time. You can submit questions for the next episode via the YT comments or via Discord (for Patreon backers). Watch TiN explain to us his e-power video cards: th-cam.com/video/ewkVNoMbcCc/w-d-xo.html Find Joe here: th-cam.com/channels/HcGSnu7Izq231nAS5zFmsw.html
To transplant a GPU, you don't need to do the procedure with tweezers and heat gun to restore the solder balls. What you need is: - laser cut stainless steel BGA template - equally sized solder balls (you can buy them in different sizes) - solder wick to remove old solder - heat gun - IR heated plate -- this is so the board won't deform The process is called BGA reballing. This will have much higher chance of success than doing it manually. Also it doesn't take that long.
I feel like part of something special here at GN thanks Steve and the rest of the Gamers Nexus team and all the community as well. There's always a friendly bunch of people here in the comments and it's nice. Thanks again 👍
14:53 That whole question should be answered by Louis Rossman, he has the tools to do CPU/GPU swaps and does the work often enough fixing Apple's broken crap. Theres a possibility that he would do it too if asked for the right price... anyone willing to do it should be able to afford it considering how much they paid for in video cards just to have the GPU swapped to begin with.
I have watched quite a few Rossman videos and in one of them he mentioned that he would never take a job that involved desoldering a GPU on a Macbook because of how much of a hassle it is. I doubt that you would be able to convince him to do one on a graphics card without forking over more cash than what it would take to just buy a new graphics card.
@@emu071981 Probably correct but the problem is that what these people want is not something you can buy. They are trying to compete with people like Kingpin that can go through dozens or hundreds of cards, find the ones with the best ram and the best dies, swap around to combine the two best into one board. You can't just buy that unless you get extremely lucky. The only way to compete is to do the same. Buy lots of cards, find out which has the best ram and which has the best gpu and then swap out parts to combine.
Hey Steve, in case you want some more info on the Indium sheets, it's nothing specifically made for CPU's but usually used to cool down lasers. It's used as a TIM and it's bassically just a very thing sheet of Indium that gets crushed in between. Because indium has thermal conductivity of 86W/mk and is very soft, with enough preassure it will plastically deform and fill those microscopic gaps. The perforated sheets exist because since it's 2/3 or 1/2 of the material without the perforations, it takes that much less force to deform them so it might be more doable under a CPU block. You can get them on ebay and aliexpress, and there are some specialized companies making them.
In my experience CPU and GPU mounting pressures were far too low to get indium foil or even micro-spring/micro-lattice sheets to work adequately sadly. There might be a thicker and more easily deformable product I didn't encounter, as my experience with Indium was somewhat limited and "laser"-focused (ha) on FPGA immersion cooling. A thicker and softer version might better suit the low mount pressures found in consumer hardware better. Seems like Indium-based sheets require pressures of 60PSI and beyond, while consumer hardware only usually mounts around 30-40PSI. The exponential curve in thermal resistance vs pressure for indium sheets results in intolerably weak performance at that mark.
@@soulofjacobeh On the GPU, since it's the bare naked chips you can't really put more preassure, but on the CPU side on the other hand, with the IHS I think you can get enough preassure (maybe with a plastic washer) on the perforated sheets. I haven't been able to test it yet though.
@@pagaiba3 Yep you're right about the low pressure situation with GPUs. I've been trying to track down some manufacturer specs for AM4 and Intel platforms to see if there's a safe way to get it high enough. For the CPU I had to test with in the office at the time, I did crank the pressure up, but it was such a low power CPU that I really couldn't use any of the data. I also couldn't measure the pressure I was applying to see how high I got it. Edit: As an aside, bringing pressure too high on CPUs usually slightly flexes the motherboard. If it doesn't immediately cause issues, it certainly does long term in most cases. Hence why I didn't yolo-test it much. I notice that on AM4 the mounting pressure is low enough that I don't get superior thermals (vs thermal grease) with even consumer-targeted graphite pads (e.g. IC graphite and Thermal Grizzly's fabric-like one), whereas on Intel Xeon 2000/3000 series the pressure was so high that IC graphite was better than every consumer paste I could test. I need to try the Indium sheets on that Xeon platform again. Stock mounting pressure on that was ridiculous. Based on its behavior with the graphite pads, I might be able to substitute the somewhat compressible IC graphite pads in for Indium for short-term testing platforms.
Following the side intake fan question, how about GPU exhaust fan? I have an 80mm fan strapped to PCI cutouts under the GPU and it improved the thermals noticeably, in theory because it sucks air under the GPU. But I have an old crappy case and I would like to see how it performs on broader variety of cases. And BTW it's way easier not to do holes in the glass, but instead make a notch for a small metal mesh piece on the side of the glass panel.
I've heard the first Ryzen release didn't have great memory compatibly, but that this has greatly been improved for the newer platforms. What components or firmware are responsible for this change? I'm curious because of the broad backwards compatibility between old and new motherboards and CPUs (eg using a new gen cpu on an old mobo platform)
As for the backwards compatibility, think there's a table around for the amd am4 chipsets and the processors themselves. Basically, the 3xx has, at most, beta support for zen2, while x570 is not compatible with zen1. That means that full compatibility, at the moment, is zen+ and the 4xx chipsets. They all run, and can be run by, any am4 motherboard.
Side intakes helped a lot with my GPU and NVME SSD...then again, the case also uses a 200mm side fan (and could possibly mount a 230mm looking at the mounting holes)
@@DaleEarnhardtsSeatbelt My case realistically can't even be fully utilised by any modern hardware, so unless I wanted something small, there is no upgrade path period. I can't even get motherboard that'd fill it. It can fit motherboards with 10x PCIe slots, and as far as I'm aware they don't exist outside of prototypes.
27:37 I still use an old Cooler Master HAF 922 with a 200mm fan in the side door. I simply ran an extension power cable to it. All of the 200mm fans are the original ones from my i5 3570K build in 2012.
In regards to the fan connector for a side mounted fan and how they can get ripped off easily when taking the side of the case off and forgetting to disconnect you could just as well make a magnetic Breakaway connection for the fan power connection
@@huri9946 TecLab I think? Might be them, they work in the area at least and I see them talking about power supplies and doing these zombiefications all the time. They're great. If not them, try looking up RTX 2080ti Super and see what comes up
Wish I knew how to do that a few years back when I bought a 9800 GX2 and killed the memry on it by trying to get more go out of it. Stupid really, at the time it was a complete monster :~)
People think that Joe only knows one word, but what they don't know is that his intonation changes for that one word, and each intonation has a very different meaning.
27:37 cube like or deeper cases with the psu in the back panel instead of the bottom shelf, you can have fans on the bottom, Corsair has an itx case like this right now
I really appreciate your honesty and knowledge. It’s clear you care and take time with each response and explanation. The pc community owes you for sharing your passion with us
Hi Steve. Would you ever consider building a case yourself? You're so good with spotting flaws immediately - I suspect with a day with a metalworker you could make something rivalling the HAF-X in performance
Hi Steve@GN, the gauge and length of the wire causes the resistance. A small gauge wire say 24 and 12" long will have way more resistance than a larger 6 gauge wire 2" long. Also m (milli) ohms is .001 ohms and M (Mega) ohms is 1,000,000 ohms, that is a huge difference. Just in case anyone wants to know. Very informative video, thanks...
On the topic of side intake fans I think a little better idea for case manufacturers would be a rear bracket under for the GPU that would change out the un-used expansion slots and replace them with fan mounting holes. That would allow for an intake/exhaust fan for the GPU.
RE the wires and breadboard question: you could compensate for the resistance of the wires by just using a slightly smaller shunt resistor on the breadboard. the real problem is the inductive loop you'd create with the wires going off to the breadboard, which would create all sorts of weird frequency-dependent effects
Thanks Steve! I'll check these places when I visit Taiwan. As you can imagine, my trip is postponed. Right now, you would need to provide documents of granted entry + stay in quarantine for 14 days. Nevermind the fact that most flights are cancelled.
For Ask GN: Apart from the motor or the bearing, what improves case fan performance more? Is it the number of blades, the angle of the blades, the design of the blades, etc.? Is static pressure vs airflow still a thing, or are there fans that have a good balance between the two? Can buyers trust the airflow, static pressure, and noise numbers listed by fan makers and vendors? Thanks GN!
I have a question for you Tech Jesus: How does one learn BIOS F/W editing and creating F/W for Laptops and desktops? Didn't really find too many resources that give a fundamental based idea on how one can go about it. This has to do with making OEM locked hardware work anywhere (MXM gpus for example) and bypassing voltage and power limits on laptops for example.
Would be cool to know, mostly it's a matter of buying a popular laptop, I've the Op 17+ and was able to get an unlocked Bios to mod it like a desktop, was able to fine tune the voltage, clock, fan speed, to dramatically lower the temps
I suspect you might be able to get help with swapping GPU packages from Louis Rossmann because he has a BGA machine, but I have no idea how he would respond on the personal side.
Ask GN: Aside from isolating from the internet and dealing with update, do you guys do anything else to try and limit background processes on Windows before testing? Do you run scripts or de-bloat / lock down tools? Or do you go in and disable certain things manually? Also, do you guys try and clean up Windows on you’re personal rigs or try to limit telemetry etc? Thanks, keep up the awesome work.
29:48 Thermaltake did that with their Spedo case. It had pad contacts on the bottom inside facing of the side panel and another contact pad on the bottom of the case where the side panel touched and it would power the 200mm fan that was on the side panel. No Wire from the board to the fan. That's one part of the design I wish Thermaltake kept for their other cases. But Thermaltake these days chases trends.
33:27 I love how, in the middle of all of these signs in Chinese and Taiwanese, there's a random "OPEN" sign XD Aside from the Intel logos in the top right and the "FULL HD" above one of the entryways, it's the only English thing in the entire frame.
Why doesn't AiO coolers just use standard car coolant? I've used some extremely old new stuff (from Woolco to help date it) to restore a cooler that was completely gunked up. I've left that sitting for around 2 years because I didn't have a use for it/didn't have the proper mounting hardware (only had the LGA 2011 screw plate) but I recently took it and ziptied it to a 3770 and it worked perfectly. Absolutely 0 issues except maybe I underfilled it as it was noisy in some orientations, but that didn't affect performance. Also tried some experiments by cooling the radiator in a bucket of dry iced coolant, which also worked fine.
i was wonderingm first some context. i am a over the road truck driver. i like to video edit, photo edit, and video game. even without doing my hobbys i require a working computer at all times. in the past ive had to buy new laptops when my main laptop breaks. which is a problem as i have to reserve a ssd slot. i also carry multiple external harddrives up to 10 and may need to have multiple plugged in at once. so my two questions. Is there a portable desktop solution that allows me to reach 64 gigs of ram and still be portable enough in all weather conditions whether it is covered or not? and since i can experience a couple hundred thousand miles a year with all the vibrations, are expensive ssds my only options?
I remember there is something like this: www.silverstonetek.com/product.php?pid=644&area=en I first learn about it from LTT, when Linus was having a blast taking the piss out of Silverstone's exhibits (just jokingly): th-cam.com/video/Bvl5-t9gvR0/w-d-xo.html
Working in storage for a couple of years. Enterprise grade HDDs can take a bit more vibrations than desktop HDDs, as they are designed to run in chassis with 60-100+ drives at once. However, the shocks on the road are something completely different. There are special server cases available for extended mobile use even in rough terrain, mainly used by the military and expeditions. That will get more expensive than a bunch of enterprise grade SSDs, though. Regarding weather conditions, there are components and complete portable systems available that support extended environments from -20°C to +70°C, high humidity and dust. Thats mostly industrial stuff, passively cooled and/or with extensive filtering which usually ends up with lower performance and regular maintenance intervals. Something in between this and a regular consumer notebook would be a Panasonic Toughbook or something similar. Those usually last for years even under the hardest conditions, but should be replaced after the warranty ends to avoid any trouble.
@@cb139 The problem with laptops is that it is difficult to self service. If a component in a desktop breaks i can order it at a city ahead of me, go and pick up the product at the store. if a laptop does that I have to go and buy another laptop while sending the first one in for RMA. That could take months to send it and then receive it. And the cheap tempory laptop i bought would break pretty quickly. Like i just recently got rid of 25 cheap laptops. Ideally i dont want to buy them, but i cant work without one. and buying two laptops does not serve me well since i would still be have a issue with performance. while i am waiting to see how the amd lineup holds for laptops coming up, in the past laptops took a really long time to put a video together. While a amd desktop could take way less of a time if i got one right now. And on storage the temperature range be whats inside the semi truck. Most of the time it should not get too hot or too cold as the humans inside it would do what it can to adjust. The range it usually gets would be -10 degrees f to 160 f degrees if no one is inside the vehicle. that wont be operating temperatures and temperature under the bed would be more 85 degrees f. In fact usually keep room temperature items under the bed during the summer. although during the winter or in cold times you want to keep room temperature items ideally on the top bunk as that usually is the hottest part of the truck. it varies.
Regarding side intake fans, I wouldn’t have an air cooled card without a fan there to get fresh air to it. I’ve modded cases to put one on a case that didn’t come with one. Worst case scenario, take the side cover completely off and cool the graphics card with an external fan lol I did that for years before I got into liquid cooling. I had multiple multi GPU setups of 2, 3, and 4 GPUs. I had three side intake, top blowing, pin wheel style air filters that moved a lot of air on high or moved moderate air on medium while remaining silent. My systems never got dusty lol and even with 3 graphics cards sandwiched together, temps remained decent. If you know anything about blower cards when you sandwich 3 together, the middle one ordinarily gets very hot compared to the other two. Although, to be fair, I did have two GTX 280s die on me once out of a 3 way SLI setup. But XFX replaced them for free, so who knows what happened there?
Yup i have a Fractal Design Define R5 with a side intake fan!! Easily lowers my GPU temps about 6-8C compared with no side intake. I also recently bought a thermal probe sensor which is supported on my Asus Maximus X Hero motherboard and stuck it to the back of my GPU. I use the thermal probe temp to control the side fan using the Fan control in the BIOS. Works amazing the side intake fan hits about 50% fan speed at 60C and will only hit 100% fan speed at 75C which pretty much never happens anyway.
did you also have the XFX 790 i mobo...I had that setup ...3 GTX 280's sucked a lot of juice....I took the grill plate off the backs of them, the bit that screws the cards to the chassis, those things were quite restrictive..certainly helped...
psycronizer actually I had an ASUS Striker II Extreme 790i, that was like the God of mother boards back then. Back when $300 was considered crazy for a mother board. It was such a great board. I had an EVGA 780i before that and the chipset used to get really hot, around 70c with 3 way SLI. That’s when I started liquid cooling, hence the Striker II, it had a water block to cool the chipset. Although, after I upgraded, I removed my cooling system and added this funny little fan I made. It was a tiny 40mm fan blowing through a little funnel and a tube so I could get air through the water block lol, surprisingly it worked well enough to keep temps in check, but I wasn’t running 3 way SLI anymore at that point, I was running 4 way Crossfire with HD 5970s on my next system. Actually what’s great about that system? I sold that ASUS board years later for $350 and the Q9650 that was good for 4.05 GHz at 1.325v on air for $300.
I have a inwin 101c with air cooling intake at back, 2 intakes at bottom yeeting air at the gpu and 2 side exhausts near the front. Gpu and CPU temps are good because they both get fresh air
@@Shadowdane I have the now retired Fractal Arc Midi. I have a slim 180mm Silverstone fan mounted on the side as an exhaust which seems unorthodox but I have crossfired GPU's so helps my CPU. I see the R5 is still listed on Fractals site so still an option.
Yep. It's a part of the reason why I have refused to ditch my Cooler Master HAF X. The airflow in it is amazing and glass can go get bent if I can have good temperatures instead.
Nvidia bins them before sending them to any of those people, and keep the best for themselves. So those FE GPU cores are almost always going to be the best.
A few questions on your schedule in one, I hope you can answer at least a few of them. Why did you stop uploading frequent Ask GN episodes? I think in the last two years there were only 16 or so, while we got almost weekly episodes before that. When is the popular fans roundup coming that was promised around 2 years ago? With all of the expensive PSU testing gear you aquired, are we going to get more PSU reviews?
The GPU can be done, I am a Rework engineer who worked to install and do research on the first BGA machine. It is NOT for the average hobbyist. It takes a stereo microscope a set of hot micro tweezers a ton of patience and a bunch of time. In the defense and aerospace industry I have done them to increase memory capacity of obsolete systems on the B-52 avionics, I did it in the early 2000s to update the encoder boards in the Hubble telescope (It was cool because I built the originals at Sperry univac when I first started out) I also did them to update the U2 for NASA as part of a avionics update to allow it to operate on civilian networks. BGA processors are not easy but it can be done as a once in a lifetime experience. Btw, if you overheat the board, its toast forever.
Indium Corp sells indium alloy sheets. The sheets need a certain amount of pressure to "liquify" them. I have been considering getting some but it is a little bit expensive (they only sell them in lots).
I used to own an Aerocool VX-E pro as my first case. AMD Phenom II X6-1090T BE, HD 6870, 4GB ram. Man, that thing had one hell of a side intake. One hell of a front one too!
I was wondering about the side intake fans as well. My first PC build ever (back in like 2004) had a side intake fan that was pretty much essential for keeping my GPU stable. I should have known that form over function was the culprit. Although the glass issue is probably the bigger culprit, because I'm sure someone could figure out how to do it in an aesthetically pleasing way.
Somebody ain't sleepin. Question is - let's see liquid helium overclocking at GN livesteam :P So you can answer once and for all, those question that popup every OC livestream "pff, LN2 OC, why not LHe instead?!". There ;)
@@NovemberOrWhatever Some CPUs can still work at temperature below -230...-240 °C. But yes, using helium to do overclocking is a big waste of non-recoverable resource paramount for scientific and technological advance of humankind.
Thank you for taking my question! The issue you had with the "liquid metal pads" might be addressed by the micro-spring indium TIMs, as they are designed to adapt (somewhat) to the surfaces that they are adapting. The ones I was looking at were from Indium Corporation here www.indium.com/products/thermal-interface-materials/#heat-spring.
Ah, I actually address this specific product in a long-winded answer in a main comment somewhere, but the tl;dr here is that the HeatSpring (and non-spring indium foil) require such extreme mounting pressure to achieve their ridiculous performance potential that it isn't practical (and usually impossible) on consumer hardware. >Also god bless those things are expensive vs. liquid metal.
Oh cool, you have a Blender guy on staff! I guess that explains what your logo is likely made in. 👍 I have just started watching some of your videos since they show up in my Google News feed. Neat stuff, although I'm not so much into the hardware as you, it's good info! I actually deal with some of the Blender administration for the data center and studio, so always nice to hear it used! Anyway, cat wants in, cheers! 🍻
I remember owning a side intake fan case cut into acrylic back in the day. The fan sat right over my AMD Athlon 2800+ Barton. It was so ugly and made cable management a complete pain; though with P-ATA/IDE cables it was damn near impossible like in today's rigs. Kids like Steve don't know what it is like to have had to walk to the computer store uphill both ways in the snow while barefoot just to get the parts to build your computer. Oh, how I miss the days of using a Dremel to cut a hole in my case for a window.
My Corsair Carbide Series 400R case from 2015 has side panel air vents instead of glass. I mounted 2 Noctua 120mm static pressure fans and a filter that blows directly on my RTX 2080 Super. Between that, my top mounted 2 120mm fan Corsair Hydro Series H100i radiator and the 2 120mm front intake fans I can run most games at 1440p 60 ~140 fps with epic settings and stay in the 60 ~ 70c range.
I had an Nvidia 9800 GTX a long time ago. Then I modded my case by cutting a hole to the bottom part of the side panel and installed an intake fan there, blowing air at the GPU. My GPU thermals went up. So it doesn't always help. The 9800 GTX was a blower design, though, and I think the extra intake fan somehow messed up the intake airflow of the radial blower fan.
On the indium TIM, I wonder how the mounting pressure comes into play? Computer hardware is probably less tolerant of higher mounting pressure than other (Industrial?) applications. I suspect that a sheet of indium would require higher mounting pressure than more conventional TIM.
@@GamersNexus Indium would likely work best if you can get it to melt slightly. A tough task for sure. The other option would be to soften it a bit with a different eutectic mixture where it flows at a more reasonable temperature.
@@MrMackievelli You could probably use ultrasonic welding techniques to input a relatively low amount of heat and still flow the indium. There are ultrasonic soldering irons that can make indium solder flow - a modification to the head should be able to get it to flow and spread across a wide area. www.cheersonic.com/portfolio-items/industry-ultrasonic-soldering-iron-for-indium/
when i worked for an electronics manufacturer that did their own refurbs and repairs in house, we used pads that required "excessive heat" to flow, but once initial flow happens, you dont have to wait for temps to hit "holly hell that coolers boiling" for it to start working properly... it was mostly used in situations where extreme heat and cycling was expected, where reflow was also expected, often the units fans would only kick on after reflow temps were hit, then the fans would kick in and cool the whole unit down drastically... not a normal common use situation, less so now then back then even since electronics are more efficient and the units we built where some sort of industrial inverter with both stepup and step down, only thing i ever saw them used for was when tested off a huge bank of batteries and running various test equipment ac and dc, or the same units after repair used to re-charge batteries after testing, funny part was, we did all the tech work, then, the units got shipped someplace to be put in a new shell... despite us having those shells on the other side of the complex... very strange way they handled that...but... it was a job.. and watching my boss sit on a live, and amazingly hot power supply board, that was on test stool...that he was always telling people not to sit on or sat things not being worked on/tested on them... burned his ass threw his levis... so bad he ended up staying home for the rest of the week... LOL.. he was such a snotty shithead....
This is what the perforations or micro-springs are for, concentrating the clamping force down onto a smaller amount of indium or using spring forces from the indium itself. These products are actually marketed specifically for computer hardware TIM applications, which was why I brought them up.
14:53 Maybe two years ago I found guy on some laptop upgrade/mod forum that transplanted and showed the whole process of putting desktop GTX980 chip into GTX980m MXM 3.0b board and it worked, so I think It would be worth trying transplanting for example Titan RTX chip into Kingpin's 2080Ti PCB and others...
I stick to my Define R5 just because it's got a side panel where i can mount a fan. It's a good case very quiet and good temps if you configure the fans in a good way, I hate tempered glass and if i wanted a glass case I'd buy a fish Tank or a see through plastic bucket. as for RGB I couldn't give a toss about it. I only have a RGB Keyboard and Mouse just so I can see the keys etc in the dark.
One thing I want to add to the question why cases don´t have side intakes anymore: I saw one case, the Geeekcase A50, it´s a 2 chamber ITX case (similar to stuff like the DAN case) but with windows on both sides. I´m pretty sure it´s not tempered glass, but it still looks pretty good. It actually has holes in the windowed side panels! Technically it doesn´t have side intakes but the GPU fans and the CPU fan are close enough to the holes I think. I would be really interested in the performance of that case.
Hello Guys hello, my friend bought one, it’s not that bad of a card, heating is an issue though. We got the NZXT G12 bracket and a NZXT AIO (240mm liquid cooler) and it’s great. Look up the NZXT G12 bracket for a 5700xt. Best after market solution. The bracket is like $20, and the AIO, depending on which one you get, make sure it’s compatible, can be around $130. Super good results. Very quiet too.
You could also opt for an aftermarket air cooler, the arctic accelero appears to be a good option but I have little experience with aftermarket GPU solutions.
@DANK sure, you can buy the whole kit, but for someone who can't fit/doesn't want to do a full custom pump it's a good idea. Full custom is better tho.
I am currently using this exact "LIQUID MetalPad" Which I am pretty sure is the product you were talking about. and I can agree, these things are so incredibly thin (feels as thin as gold foil) that there's a good chance that, depending on how you tightened the screws on your cooler, there's noticeable imperfections. You're supposed to switch off your cooling and let the CPU "burn in" for 30+Minutes at at least 70°c for the metal to liquefy and spread out properly. This does not perform any better than standard MX-4 thermal paste. In fact I might have trouble with hot spots on my Ryzen 3900x IHS so this stuff is definitely coming off the next time I change something on my build.
I wonder, with the glass sides if they could lay some transparent tracks on the glass like in an lcd panel to get power to a fan, or if there would not be enough material to handle the current.
What are the criteria for reviewing stuff (e.g. AIO-Coolers)? I was kinda sad that the Arctic Freezer II was not review by you guys because it seemed pretty innovative (VRM-Fan [not technically new cause Cryorig], their own pump design, one header for all Motors). So that made me wonder what your process for selecting stuff for review is.
Yeah side intake fans do a great job in keeping GPU thermals in check. Have an old Antec three hundred that I bought new years ago. My first case with a side intake
Question: Have you ever consider making a page on your web with full charts with comparisons of pc cases CPU temperatures, GPU temperatures, dB levels, coolers and others? I will love to see it there to compere every case you tested in one chart.
Side fans are amazing, when i changed my system i went from my old case with side fan to a typical stupid glass panel. Since then i tested to run my system in my old case: ryzen 2700 at 4.15 ghz cooled with 240mm AIO, X470 Strix and Palit 2080. I dont have proper thermal gun, but AIDA showed that temps went from: 69 to 61 on CPU, 84-85 to 80 on GPU and most importantly from 87 to 69 on VRM. If my old case had dust filters and wasn't so out of date i would rather keept using it. So when i will be looking for a new case, a side fan will be a big factor for my purchase. Ps: i would also love if manufacturers once again started making custom VRM heatsinks, this would benefit many users with boards below 250$/€.
For the Indium foil and Indium HeatSpring (and similar metal "pads") vs. liquid metal question: most of the "liquid metal pads" require excessively high mounting pressure (such as found in FPGAs and other less-consumer-oriented devices) to achieve winning performance. Where a traditional consumer GPU might have mounting pressure specs around 35PSI, a well-made micro-spring metal sheet will only begin to approach low-end thermal grease - best case. Most publicly available metal pads only begin to really shine around 60-80PSI and beyond (source from one of the materials I researched: chart 3 on www.indium.com/products/thermal-interface-materials/ ). In my experiments, I too did the stacking that Steve mentioned. It was the only way on 80% of consumer platforms to achieve reasonable pressure, but this only brought it down to temps matching thermal paste. I even started with this approach with some of the lower mounting pressure FPGAs before opting to push limits on pressure specs instead. By recklessly ignoring pressure specs, I was able to reach some impressive performance. With lapped or otherwise better-mated IHSsink surfaces AND higher mounting pressure, the metal pads were exceptional and great for the application I was targeting (two-stage immersion cooling of FPGAs). Making an educated guess based on these findings, I believe liquid metal would've likely performed similarly (possibly better, as there would be less of a 'barrier' than with a non-displaceable pad), but it couldn't be used in our immersion baths (imagine liquid metal escaping and floating freely in another liquid surrounding your precious computer bits) so I didn't get around to testing it at the higher mount pressures :(. Also, the micro-spring or latticed metal pads are ungodly expensive in most cases. Non-laticed/non-spring indium foil on its own was barely worth looking at, as it carried about half the performance of the micro-spring adaptation, was painfully easy to permanently deform, and not particularly reusable as it would carry an imprint of the imperfections of the surfaces it was mating. The micro-spring was somewhat more reusable, but did also carry some deformations/imprints. --- Sorry for the text wall, but it's a neat question and subject. I rarely get to share :P Edit: If Steve sees this I'll be so happy ;_;
Pin-to-pad would be stupid-complex to engineer and implement. You're talking about an additional brace along one side with not just contacts for four-pin / three pin fans but also the RGB connectors, which then would make it a miniaturized 8-pin connector.The pads would have to be mounted to the side panel, and the pins would have to be mounted to an otherwise-useless support mechanism which would not permit easy access. There could be slots in the case body to permit drop-in installation and some tabs with indents to pop the rail in place, giving it _some_ semblance of attachment, _or_ an indented tab on front with a thumbscrew in back ala old-style cases which allowed only eight screws to hold in an optical device (and usually, it was _only_ two of the eight holes anyway) which would provide both a temporary and permanent attachment. But the means to fasten it still has to be engineered into a case. _That's not ever going to happen. What I would like to see manufacturers of cases do is to have a cutout for the GPU, entirely. We see this with sandwich-style boards where the GPU is side-mounted with riser, so why not purpose that with mesh on a steel panel to draw in cool air from the outside? The GPU heatsink willl still work as intended to spill its hot air out, but the case, allowing for air to come in from outside directly rather than a convoluted 90-degree rotation from top, back or side would be better than in the middle next to other components. It would make for a wider case, but that's what needs to be done to re-introduce side intake. Sapphire's Pulse RX AMD AIB cards has a very ingenious fan system where only one screw holds it on, with the fan motor assembly on a thin plastic frame. For people who would rather side _exhaust_ with their GPU, Sapphire's partners could easily produce a flipped version which spits the air out directly from the top and draw in from the sides, which would make the GPU function similarly to side-mounted exhaust for the entire system should a user prefer. More video card manufacturers making their fan assemblies easily removable from cooler housings would permit more of an interest in cases with side ventilation, provided people want this, and people _do_ want this, but they're a vocal minority the case manufacturers are not going to heed because the idea is too expansive to manufacturer, and many users... simply don't care about performance. We've seen that 2018, we've seen that 2019, we're seeing it today. Shame it came to that, but manufacturers need to satisfy the _majority_ of users, and avoid being too niche to remain profitable. You voting with your wallet is irrelevant, and the TH-cam tech scene is also partly to blame for this, showing off concepts which choke airflow back in 2018 and people just _eating that up. _Couldn't have enough!_ Also sadly, even though some people in the tech scene _eventually_ made their cries for airflow heard, and cases are being made with airflow in mind thanks to the efforts of every Steve in the scene (people who know will get the joke), the airflow trend is going to last for maybe a year or two before case manufactures cock it up again and block off _all_ of the holes for cost savings, passing it off as for looks. It's only a matter of time.
Follow up question : I know this will porbably dont work but will still ask anyway .What about a 1x2 Chip socket instead of a full breadboard , soldered paralel to the saunt resistor , then just attach legs from a jumper cable onto the saunt resistors you want to test out .
Bit of a followup to the side panel fans not looking good, where personally do you draw the line between form and function with a computer? Like how much in terms of looks or acoustics are you willing to sacrifice in the name of performance?
Tempered glass it not needed for the side panel. For example Fractal design should save up and purchase the Define R5 case and examine how they were able to add a side panel fan to their solid steel side panel, and use that info to design a side panel for their newer cases.
my 2010 Area 51 Alienware case that I still have and like to play around with it because it has an ASUS P6T V2 X58 Motherboard in it... This case has Pin to Pad fan power on both side panels. I've even thought about taking the hardware off of the case in order to use it on another case to do something similar.
Would love to see you guys use the EVC2 by Elmor to push the Voltage on a 2080 Ti or a Titan RTX. I would be interested to see Turing's scaling with more voltage, that 1.093 voltage limit is annoying on Pascal.
What about those fans you used to get that you could fit into your spare pci cutouts on the back of the case. I'd be interested to know if lining up one or more of those under the GPU helped thermals at all. I think they were exhaust when you could get em, but i bet you could flip the fan.
indium sheets can perform very good as tim, but you need A LOT of mounting pressure, wich is not reasonably achieveble, especially between the die and IHS. the reason you need that much mounting pressure is the same reason you need TIM in the first place, you want to fill in microscopic gaps. to achieve that, the mouning pressure has to overcome the point where the TIM starts to expierience plastic deformation. I'm not sure if you need to maintain that level of mounting pressure or if a one time "pressing", maybe at high temperatures would be enough.
And what about Graphene thermal pads? I've seen some companies make them and prices do look similar to normal thermal pads. How is that performance-wise? On another topic - GPU transplants were common in laptop industry because there was a series of faulty products back around 2009. What you generally need is a sieve for specific chip to place BGA's in correct positions and hotair soldering station. Until now I didn't seen anyone setting BGA's by hand.
Yeah Im using IC Diamond Paste on my gaming build, seem to cool good, my 3900x never go over 70% under full load. However,, not the easiest paste to work with its very thick, so ideally the tube under the sun 10 min before applying would help it spread.
BTW, don't try to drill through tempered glass, because in the tempering process the glass gets a ton of internal stresses, and if you drill into it the forces will be released, and the glass explodes.
Dear Steve and GN. I was wondering if turning CPU to real time or high etc really does help gaming or multithreaded operations in any way whether it be fps, latency, etc. I sweat that 1% lows or avg fps seem to be higher when I go to real time
I still use my old antec case that I got when I built my first computer in 2012-2013 and it has a fan mount on the side panel. That thing is fucking amazing for my GPU temps and I never want a case without it. Sure you can make arguments that if I had a bigger case with a better design I could also improve my GPU temps, but that side mounted fan REALLY helps my GPU in this case. I hate glass-sides on computer cases, I just want a black box that's functional and that's what I've got.
Could we have some content covering silicon manufacturing technologies- we see clips of fabs, and mentions of different processes, but it'd be nice to understand what actually happens in a fab- some forms of lithography, apparently, but more detail(as far as processes being proprietary can take us) would be great?
The first point about contact being an issue, makes me think, would it be feasible for CPU manufacturers to make the die from a solid piece of metal that includes the heatsink? So like the die, IHS and heatsink are all 1 piece and the chip is built around that? considering the difference between some of intel's stock TIM and regular paste VS liquid metal between IHS and heatsing, you could probably get insane cooling on air alone with relatively small heatsinks. Maybe make instead of a heatsink make it 1 piece as a standard fitting waterblock that you can attach any pump/rad combo to. I know when I delidded and repasted my i7-7700k it dropped from like mid 90s to 65°C peaks from paste alone.
I had coolaboratory with H100i + i3570k. Two years. Then it let the exact same print on the block and CPU as the original Corsair paste. The OEM was on it for 5 years before. You have to polish both surfaces as the instruction says. So warranty void. It had 5c better delta. In 550D full with HDDs and air cooled XFX RX 580 GTR-S uperhot. Low RPM intake and exhaust.
The alternatives to liquid metal video would be extremely interesting. If possible could you test the graphene pad as alternative to thermal paste. that LTT did a video on?
@GN The Pins to Pads Idea for Side Mounted Fans had been done before, Thermaltake did it with their Level 10 GT Case, I still use my Level 10 GT Snow Edition Case and its Side Fan has Connectors at the Hinge. I don't know if any other manufacturers did or if Thermaltake had Patended the idea or not. I do know why they stopped doing side mounted fans is mostly due to the glass trend like you said however. I do think that they could look decent together with the right designs though, kind of like how the Level 10 looks except with a bigger glass panel possibly and maybe a smaller fan, but it definitely could work!
Indium solutions can be as good as 80 W/mK, but there are numerous issues with using them as thermal pads in DIY computer applications. You mentioned the issues with compression and flow at low pressure and temperature, but there are also issues with oxide formation, especially with high In content solutions. Mostly I've seen them in power semiconductor and photonic applications, as well as cryogenic applications. It might be interesting to use In for cryo cooling for extreme OCing, but applying it would be more like soldering the die to the pot than applying thermal grease. Another potential use case would be in high power modular systems, like those found in military and aerospace environments. Ultimately, as cool as the idea is I can't see it being practical outside of extreme applications.
I work in the architect/construction industry and deal a lot of with Revit workloads. Would you consider adding a Revit benchmark like RFO to your CPU testing. I think that that Revit is mostly single core frequency dependant workload, but I'd be interested to see where ryzen 3k stacks up.
so if wire has enough resistance couldn't you find the right gauge wire and just jump from one side of the shunt to the other? i mean a thick enough or long enough wire would cause resistance?
Thermaltake lvl 10 gt snow edition full tower has a large side intake fan with pin to pad power connection without wires from the motherboard to the fan itself.
Timestamps in the description as well. The Patrons Ask GN episode is now live: www.patreon.com/posts/35658549 -- thanks to all our supporters for the help, especially during this time. You can submit questions for the next episode via the YT comments or via Discord (for Patreon backers).
Watch TiN explain to us his e-power video cards: th-cam.com/video/ewkVNoMbcCc/w-d-xo.html
Find Joe here: th-cam.com/channels/HcGSnu7Izq231nAS5zFmsw.html
can you try pure potassium metal as a thermal paste? with the heatsink vacuum sealed on the cpu.
Yeah... one video solder GPU into water block...kakaka
What do you think is the best budget card, new or used? I would assume it's the RX 580 but I would love to hear what you think.
1st I got the shirt I ordered, took a while to get here to Australia, thx. Are you going to do a covid folding at home shirt?
You should find a dead Titan RTX, and transplant it onto a 2080 Ti board, add in the additional memory chip and see if you can get that working!
Good guy GN having descriptions with their timestamps. True heroes here, kids.
To transplant a GPU, you don't need to do the procedure with tweezers and heat gun to restore the solder balls. What you need is:
- laser cut stainless steel BGA template
- equally sized solder balls (you can buy them in different sizes)
- solder wick to remove old solder
- heat gun
- IR heated plate -- this is so the board won't deform
The process is called BGA reballing. This will have much higher chance of success than doing it manually. Also it doesn't take that long.
For the first time in a while, Steve doesn’t wear black
Its been the same shirt, all this time.
The shirt has simply shunt modded to appear blue.
It's a host. High fidelity, must say.
If you checkout some of his older videos he's wearing blue too.
Dude I knew something was off, thanks lol
I feel like part of something special here at GN thanks Steve and the rest of the Gamers Nexus team and all the community as well. There's always a friendly bunch of people here in the comments and it's nice.
Thanks again 👍
14:53 That whole question should be answered by Louis Rossman, he has the tools to do CPU/GPU swaps and does the work often enough fixing Apple's broken crap. Theres a possibility that he would do it too if asked for the right price... anyone willing to do it should be able to afford it considering how much they paid for in video cards just to have the GPU swapped to begin with.
Can be done relatively easy with the right equipment, can it be done reliably? No not really.
Use the zhuomao!
I have watched quite a few Rossman videos and in one of them he mentioned that he would never take a job that involved desoldering a GPU on a Macbook because of how much of a hassle it is. I doubt that you would be able to convince him to do one on a graphics card without forking over more cash than what it would take to just buy a new graphics card.
One russian channel on youtube put a gtx 1070 chip onto a 1060 board, also added memory chips so it's 8GB
@@emu071981 Probably correct but the problem is that what these people want is not something you can buy. They are trying to compete with people like Kingpin that can go through dozens or hundreds of cards, find the ones with the best ram and the best dies, swap around to combine the two best into one board. You can't just buy that unless you get extremely lucky. The only way to compete is to do the same. Buy lots of cards, find out which has the best ram and which has the best gpu and then swap out parts to combine.
Hey Steve, in case you want some more info on the Indium sheets, it's nothing specifically made for CPU's but usually used to cool down lasers. It's used as a TIM and it's bassically just a very thing sheet of Indium that gets crushed in between. Because indium has thermal conductivity of 86W/mk and is very soft, with enough preassure it will plastically deform and fill those microscopic gaps.
The perforated sheets exist because since it's 2/3 or 1/2 of the material without the perforations, it takes that much less force to deform them so it might be more doable under a CPU block.
You can get them on ebay and aliexpress, and there are some specialized companies making them.
In my experience CPU and GPU mounting pressures were far too low to get indium foil or even micro-spring/micro-lattice sheets to work adequately sadly. There might be a thicker and more easily deformable product I didn't encounter, as my experience with Indium was somewhat limited and "laser"-focused (ha) on FPGA immersion cooling. A thicker and softer version might better suit the low mount pressures found in consumer hardware better.
Seems like Indium-based sheets require pressures of 60PSI and beyond, while consumer hardware only usually mounts around 30-40PSI. The exponential curve in thermal resistance vs pressure for indium sheets results in intolerably weak performance at that mark.
@@soulofjacobeh On the GPU, since it's the bare naked chips you can't really put more preassure, but on the CPU side on the other hand, with the IHS I think you can get enough preassure (maybe with a plastic washer) on the perforated sheets. I haven't been able to test it yet though.
@@pagaiba3 Yep you're right about the low pressure situation with GPUs. I've been trying to track down some manufacturer specs for AM4 and Intel platforms to see if there's a safe way to get it high enough.
For the CPU I had to test with in the office at the time, I did crank the pressure up, but it was such a low power CPU that I really couldn't use any of the data. I also couldn't measure the pressure I was applying to see how high I got it.
Edit: As an aside, bringing pressure too high on CPUs usually slightly flexes the motherboard. If it doesn't immediately cause issues, it certainly does long term in most cases. Hence why I didn't yolo-test it much.
I notice that on AM4 the mounting pressure is low enough that I don't get superior thermals (vs thermal grease) with even consumer-targeted graphite pads (e.g. IC graphite and Thermal Grizzly's fabric-like one), whereas on Intel Xeon 2000/3000 series the pressure was so high that IC graphite was better than every consumer paste I could test.
I need to try the Indium sheets on that Xeon platform again. Stock mounting pressure on that was ridiculous.
Based on its behavior with the graphite pads, I might be able to substitute the somewhat compressible IC graphite pads in for Indium for short-term testing platforms.
Silicone Lottery - When you come out of the procedure looking natural.
This cracked me up 🤣
Following the side intake fan question, how about GPU exhaust fan? I have an 80mm fan strapped to PCI cutouts under the GPU and it improved the thermals noticeably, in theory because it sucks air under the GPU. But I have an old crappy case and I would like to see how it performs on broader variety of cases. And BTW it's way easier not to do holes in the glass, but instead make a notch for a small metal mesh piece on the side of the glass panel.
I've heard the first Ryzen release didn't have great memory compatibly, but that this has greatly been improved for the newer platforms. What components or firmware are responsible for this change? I'm curious because of the broad backwards compatibility between old and new motherboards and CPUs (eg using a new gen cpu on an old mobo platform)
Memory controller update, AMD mentioned it when zen+ was launched along with a major selling point for zen2
As for the backwards compatibility, think there's a table around for the amd am4 chipsets and the processors themselves. Basically, the 3xx has, at most, beta support for zen2, while x570 is not compatible with zen1.
That means that full compatibility, at the moment, is zen+ and the 4xx chipsets. They all run, and can be run by, any am4 motherboard.
@@GrimpakTheMook no, thats b350 with semi support for zen2, but x370 full support for zen2 all around
@@Tallnerdyguy then the data I saw was outdated (showed beta support only for those), thanks for the correction!
@@Tallnerdyguy Not quite full support, most 3xx and 4xx boards don't support pcie-e4 but that's not really a problem.
Side intakes helped a lot with my GPU and NVME SSD...then again, the case also uses a 200mm side fan (and could possibly mount a 230mm looking at the mounting holes)
Same here. I just left a comment about it. There are no case upgrades after 2010 unless you value form over function.
@@DaleEarnhardtsSeatbelt My case realistically can't even be fully utilised by any modern hardware, so unless I wanted something small, there is no upgrade path period.
I can't even get motherboard that'd fill it. It can fit motherboards with 10x PCIe slots, and as far as I'm aware they don't exist outside of prototypes.
27:37 I still use an old Cooler Master HAF 922 with a 200mm fan in the side door. I simply ran an extension power cable to it. All of the 200mm fans are the original ones from my i5 3570K build in 2012.
In regards to the fan connector for a side mounted fan and how they can get ripped off easily when taking the side of the case off and forgetting to disconnect you could just as well make a magnetic Breakaway connection for the fan power connection
Gamers Nexus is my favourite hardware channel up there with Paul's Hardware and Laurent's Choice. Keep up the great work!
There’s a Brazilian channel that did a 2080 Ti “super”. Transplanting memory from a 2080 super onto a 2080ti card.
what's the name of the channel?
@@huri9946 TecLab I think? Might be them, they work in the area at least and I see them talking about power supplies and doing these zombiefications all the time. They're great. If not them, try looking up RTX 2080ti Super and see what comes up
Wish I knew how to do that a few years back when I bought a 9800 GX2 and killed the memry on it by trying to get more go out of it. Stupid really, at the time it was a complete monster :~)
@@pegasusted2504
Ha ha, still got my 9800 GX2, keeping it as a relic of the past.
I was a pretty common procedure to do on the ps3
Slacker!?!?
People think that Joe only knows one word, but what they don't know is that his intonation changes for that one word, and each intonation has a very different meaning.
@@GamersNexus So he is a pokeman?
I’m still waiting for my LN2 Pots from Roman haha
Taiwan is my favorite place, it’s like florida with mountains ahhaha
Brad Haines florida man is real haha
27:37 cube like or deeper cases with the psu in the back panel instead of the bottom shelf, you can have fans on the bottom, Corsair has an itx case like this right now
I really appreciate your honesty and knowledge. It’s clear you care and take time with each response and explanation. The pc community owes you for sharing your passion with us
Hi Steve. Would you ever consider building a case yourself? You're so good with spotting flaws immediately - I suspect with a day with a metalworker you could make something rivalling the HAF-X in performance
Minds you, a good food critic doesn't make a good chef, a good movie reviewer might not be a good director.
Steven Zephyc they’re subjective opinion based fields, best cooling case would be engineered scientific results driven (aesthetics aside)
Hi Steve@GN, the gauge and length of the wire causes the resistance. A small gauge wire say 24 and 12" long will have way more resistance than a larger 6 gauge wire 2" long. Also m (milli) ohms is .001 ohms and M (Mega) ohms is 1,000,000 ohms, that is a huge difference. Just in case anyone wants to know. Very informative video, thanks...
On the topic of side intake fans I think a little better idea for case manufacturers would be a rear bracket under for the GPU that would change out the un-used expansion slots and replace them with fan mounting holes. That would allow for an intake/exhaust fan for the GPU.
I remember when huge 250mm fans on the side was the trend especially for cheaper cases. I never saw a high end case with them though.
RE the wires and breadboard question: you could compensate for the resistance of the wires by just using a slightly smaller shunt resistor on the breadboard. the real problem is the inductive loop you'd create with the wires going off to the breadboard, which would create all sorts of weird frequency-dependent effects
Thanks Steve! I'll check these places when I visit Taiwan. As you can imagine, my trip is postponed. Right now, you would need to provide documents of granted entry + stay in quarantine for 14 days. Nevermind the fact that most flights are cancelled.
For Ask GN: Apart from the motor or the bearing, what improves case fan performance more? Is it the number of blades, the angle of the blades, the design of the blades, etc.? Is static pressure vs airflow still a thing, or are there fans that have a good balance between the two? Can buyers trust the airflow, static pressure, and noise numbers listed by fan makers and vendors? Thanks GN!
Oh, I'm interested in this one too.
I have a question for you Tech Jesus: How does one learn BIOS F/W editing and creating F/W for Laptops and desktops? Didn't really find too many resources that give a fundamental based idea on how one can go about it. This has to do with making OEM locked hardware work anywhere (MXM gpus for example) and bypassing voltage and power limits on laptops for example.
Would be cool to know, mostly it's a matter of buying a popular laptop, I've the Op 17+ and was able to get an unlocked Bios to mod it like a desktop, was able to fine tune the voltage, clock, fan speed, to dramatically lower the temps
nVidia: MXM? What's that?
I suspect you might be able to get help with swapping GPU packages from Louis Rossmann because he has a BGA machine, but I have no idea how he would respond on the personal side.
I would love colab between GN and Louis so much. I thing they would be pretty compatible in hardware content.
this!!
5:36 onward for a few seconds was the most anxiety-inducing thing I've seen in weeks!
I felt pain from just looking at it... Pavlov reflex of tech enthusiasts
Ask GN: Aside from isolating from the internet and dealing with update, do you guys do anything else to try and limit background processes on Windows before testing? Do you run scripts or de-bloat / lock down tools? Or do you go in and disable certain things manually? Also, do you guys try and clean up Windows on you’re personal rigs or try to limit telemetry etc?
Thanks, keep up the awesome work.
29:48 Thermaltake did that with their Spedo case. It had pad contacts on the bottom inside facing of the side panel and another contact pad on the bottom of the case where the side panel touched and it would power the 200mm fan that was on the side panel. No Wire from the board to the fan. That's one part of the design I wish Thermaltake kept for their other cases. But Thermaltake these days chases trends.
33:27
I love how, in the middle of all of these signs in Chinese and Taiwanese, there's a random "OPEN" sign XD Aside from the Intel logos in the top right and the "FULL HD" above one of the entryways, it's the only English thing in the entire frame.
28:16 Top mounted PSU and bottom intake would be a good asthetic and cooling.... well I wouldn't even call it a compromise really; it's just better.
Why doesn't AiO coolers just use standard car coolant? I've used some extremely old new stuff (from Woolco to help date it) to restore a cooler that was completely gunked up. I've left that sitting for around 2 years because I didn't have a use for it/didn't have the proper mounting hardware (only had the LGA 2011 screw plate) but I recently took it and ziptied it to a 3770 and it worked perfectly. Absolutely 0 issues except maybe I underfilled it as it was noisy in some orientations, but that didn't affect performance. Also tried some experiments by cooling the radiator in a bucket of dry iced coolant, which also worked fine.
i was wonderingm first some context. i am a over the road truck driver. i like to video edit, photo edit, and video game. even without doing my hobbys i require a working computer at all times. in the past ive had to buy new laptops when my main laptop breaks. which is a problem as i have to reserve a ssd slot. i also carry multiple external harddrives up to 10 and may need to have multiple plugged in at once. so my two questions. Is there a portable desktop solution that allows me to reach 64 gigs of ram and still be portable enough in all weather conditions whether it is covered or not? and since i can experience a couple hundred thousand miles a year with all the vibrations, are expensive ssds my only options?
Great question. We'll try to ask some storage experts.
I remember there is something like this:
www.silverstonetek.com/product.php?pid=644&area=en
I first learn about it from LTT, when Linus was having a blast taking the piss out of Silverstone's exhibits (just jokingly):
th-cam.com/video/Bvl5-t9gvR0/w-d-xo.html
I believe server grade HDD's are alot lore vibration and heat resistant than regular Seagate Barracudas and WD Reds
Working in storage for a couple of years. Enterprise grade HDDs can take a bit more vibrations than desktop HDDs, as they are designed to run in chassis with 60-100+ drives at once. However, the shocks on the road are something completely different. There are special server cases available for extended mobile use even in rough terrain, mainly used by the military and expeditions. That will get more expensive than a bunch of enterprise grade SSDs, though.
Regarding weather conditions, there are components and complete portable systems available that support extended environments from -20°C to +70°C, high humidity and dust. Thats mostly industrial stuff, passively cooled and/or with extensive filtering which usually ends up with lower performance and regular maintenance intervals. Something in between this and a regular consumer notebook would be a Panasonic Toughbook or something similar. Those usually last for years even under the hardest conditions, but should be replaced after the warranty ends to avoid any trouble.
@@cb139 The problem with laptops is that it is difficult to self service. If a component in a desktop breaks i can order it at a city ahead of me, go and pick up the product at the store. if a laptop does that I have to go and buy another laptop while sending the first one in for RMA. That could take months to send it and then receive it. And the cheap tempory laptop i bought would break pretty quickly. Like i just recently got rid of 25 cheap laptops. Ideally i dont want to buy them, but i cant work without one. and buying two laptops does not serve me well since i would still be have a issue with performance. while i am waiting to see how the amd lineup holds for laptops coming up, in the past laptops took a really long time to put a video together. While a amd desktop could take way less of a time if i got one right now. And on storage the temperature range be whats inside the semi truck. Most of the time it should not get too hot or too cold as the humans inside it would do what it can to adjust. The range it usually gets would be -10 degrees f to 160 f degrees if no one is inside the vehicle. that wont be operating temperatures and temperature under the bed would be more 85 degrees f. In fact usually keep room temperature items under the bed during the summer. although during the winter or in cold times you want to keep room temperature items ideally on the top bunk as that usually is the hottest part of the truck. it varies.
Regarding side intake fans, I wouldn’t have an air cooled card without a fan there to get fresh air to it. I’ve modded cases to put one on a case that didn’t come with one. Worst case scenario, take the side cover completely off and cool the graphics card with an external fan lol
I did that for years before I got into liquid cooling. I had multiple multi GPU setups of 2, 3, and 4 GPUs. I had three side intake, top blowing, pin wheel style air filters that moved a lot of air on high or moved moderate air on medium while remaining silent. My systems never got dusty lol and even with 3 graphics cards sandwiched together, temps remained decent. If you know anything about blower cards when you sandwich 3 together, the middle one ordinarily gets very hot compared to the other two. Although, to be fair, I did have two GTX 280s die on me once out of a 3 way SLI setup. But XFX replaced them for free, so who knows what happened there?
Yup i have a Fractal Design Define R5 with a side intake fan!! Easily lowers my GPU temps about 6-8C compared with no side intake. I also recently bought a thermal probe sensor which is supported on my Asus Maximus X Hero motherboard and stuck it to the back of my GPU. I use the thermal probe temp to control the side fan using the Fan control in the BIOS. Works amazing the side intake fan hits about 50% fan speed at 60C and will only hit 100% fan speed at 75C which pretty much never happens anyway.
did you also have the XFX 790 i mobo...I had that setup ...3 GTX 280's sucked a lot of juice....I took the grill plate off the backs of them, the bit that screws the cards to the chassis, those things were quite restrictive..certainly helped...
psycronizer actually I had an ASUS Striker II Extreme 790i, that was like the God of mother boards back then. Back when $300 was considered crazy for a mother board. It was such a great board. I had an EVGA 780i before that and the chipset used to get really hot, around 70c with 3 way SLI. That’s when I started liquid cooling, hence the Striker II, it had a water block to cool the chipset. Although, after I upgraded, I removed my cooling system and added this funny little fan I made. It was a tiny 40mm fan blowing through a little funnel and a tube so I could get air through the water block lol, surprisingly it worked well enough to keep temps in check, but I wasn’t running 3 way SLI anymore at that point, I was running 4 way Crossfire with HD 5970s on my next system.
Actually what’s great about that system? I sold that ASUS board years later for $350 and the Q9650 that was good for 4.05 GHz at 1.325v on air for $300.
I have a inwin 101c with air cooling intake at back, 2 intakes at bottom yeeting air at the gpu and 2 side exhausts near the front. Gpu and CPU temps are good because they both get fresh air
@@Shadowdane I have the now retired Fractal Arc Midi. I have a slim 180mm Silverstone fan mounted on the side as an exhaust which seems unorthodox but I have crossfired GPU's so helps my CPU. I see the R5 is still listed on Fractals site so still an option.
Yep. It's a part of the reason why I have refused to ditch my Cooler Master HAF X. The airflow in it is amazing and glass can go get bent if I can have good temperatures instead.
Also they're Binned cards in top for HOF, Kingpin, Lightening.
Nvidia bins them before sending them to any of those people, and keep the best for themselves. So those FE GPU cores are almost always going to be the best.
A few questions on your schedule in one, I hope you can answer at least a few of them.
Why did you stop uploading frequent Ask GN episodes? I think in the last two years there were only 16 or so, while we got almost weekly episodes before that.
When is the popular fans roundup coming that was promised around 2 years ago?
With all of the expensive PSU testing gear you aquired, are we going to get more PSU reviews?
so cool to see other inputs to the questions too .. love this format
The GPU can be done, I am a Rework engineer who worked to install and do research on the first BGA machine. It is NOT for the average hobbyist. It takes a stereo microscope a set of hot micro tweezers a ton of patience and a bunch of time. In the defense and aerospace industry I have done them to increase memory capacity of obsolete systems on the B-52 avionics, I did it in the early 2000s to update the encoder boards in the Hubble telescope (It was cool because I built the originals at Sperry univac when I first started out) I also did them to update the U2 for NASA as part of a avionics update to allow it to operate on civilian networks. BGA processors are not easy but it can be done as a once in a lifetime experience. Btw, if you overheat the board, its toast forever.
For the iridium sheet, you need exceptional pressure to cause the foil to deform enough to get good contact
Indium Corp sells indium alloy sheets. The sheets need a certain amount of pressure to "liquify" them. I have been considering getting some but it is a little bit expensive (they only sell them in lots).
27:37 my thermaltake Gandalf case was brilliant at side cooling! It had a massive fan on one side of the case!
I used to own an Aerocool VX-E pro as my first case. AMD Phenom II X6-1090T BE, HD 6870, 4GB ram. Man, that thing had one hell of a side intake. One hell of a front one too!
I was wondering about the side intake fans as well. My first PC build ever (back in like 2004) had a side intake fan that was pretty much essential for keeping my GPU stable. I should have known that form over function was the culprit. Although the glass issue is probably the bigger culprit, because I'm sure someone could figure out how to do it in an aesthetically pleasing way.
tbh, Steve, I always thought the whole GN team got to read these questions and give input. Impressive!!!
Somebody ain't sleepin. Question is - let's see liquid helium overclocking at GN livesteam :P So you can answer once and for all, those question that popup every OC livestream "pff, LN2 OC, why not LHe instead?!". There ;)
We might need some help with that one...
@@GamersNexus It's probably expensive as all hell with the price of helium
what's the advantage of helium? wouldn't cold bug make it pointless?
@@NovemberOrWhatever Some CPUs can still work at temperature below -230...-240 °C. But yes, using helium to do overclocking is a big waste of non-recoverable resource paramount for scientific and technological advance of humankind.
@@KirstyWales $6k+ USD for one dewar. LN2 cost is
Thank you for taking my question! The issue you had with the "liquid metal pads" might be addressed by the micro-spring indium TIMs, as they are designed to adapt (somewhat) to the surfaces that they are adapting. The ones I was looking at were from Indium Corporation here www.indium.com/products/thermal-interface-materials/#heat-spring.
Ah, I actually address this specific product in a long-winded answer in a main comment somewhere, but the tl;dr here is that the HeatSpring (and non-spring indium foil) require such extreme mounting pressure to achieve their ridiculous performance potential that it isn't practical (and usually impossible) on consumer hardware.
>Also god bless those things are expensive vs. liquid metal.
Oh cool, you have a Blender guy on staff! I guess that explains what your logo is likely made in. 👍 I have just started watching some of your videos since they show up in my Google News feed. Neat stuff, although I'm not so much into the hardware as you, it's good info! I actually deal with some of the Blender administration for the data center and studio, so always nice to hear it used! Anyway, cat wants in, cheers! 🍻
I remember owning a side intake fan case cut into acrylic back in the day. The fan sat right over my AMD Athlon 2800+ Barton. It was so ugly and made cable management a complete pain; though with P-ATA/IDE cables it was damn near impossible like in today's rigs. Kids like Steve don't know what it is like to have had to walk to the computer store uphill both ways in the snow while barefoot just to get the parts to build your computer. Oh, how I miss the days of using a Dremel to cut a hole in my case for a window.
My Corsair Carbide Series 400R case from 2015 has side panel air vents instead of glass. I mounted 2 Noctua 120mm static pressure fans and a filter that blows directly on my RTX 2080 Super. Between that, my top mounted 2 120mm fan Corsair Hydro Series H100i radiator and the 2 120mm front intake fans I can run most games at 1440p 60 ~140 fps with epic settings and stay in the 60 ~ 70c range.
I had an Nvidia 9800 GTX a long time ago. Then I modded my case by cutting a hole to the bottom part of the side panel and installed an intake fan there, blowing air at the GPU. My GPU thermals went up. So it doesn't always help. The 9800 GTX was a blower design, though, and I think the extra intake fan somehow messed up the intake airflow of the radial blower fan.
I feel like Steve should find a way to yeet his phone across the room during all his advertisements.
.... why didn't you just say throw? >_
@@Lindquist-24 Thank you.
@@Lindquist-24 I wanted it to imply a much more ridiculous throw. Not just some average toss.
Toss is calm, throw is aggressive... Right?
@@Lindquist-24 but yeet is like an aggressive throw. I dunno, seems like multiple levels: toss -> throw -> yeet -> shot put.
On the indium TIM, I wonder how the mounting pressure comes into play? Computer hardware is probably less tolerant of higher mounting pressure than other (Industrial?) applications. I suspect that a sheet of indium would require higher mounting pressure than more conventional TIM.
Excellent question and point!
@@GamersNexus Indium would likely work best if you can get it to melt slightly. A tough task for sure. The other option would be to soften it a bit with a different eutectic mixture where it flows at a more reasonable temperature.
@@MrMackievelli You could probably use ultrasonic welding techniques to input a relatively low amount of heat and still flow the indium. There are ultrasonic soldering irons that can make indium solder flow - a modification to the head should be able to get it to flow and spread across a wide area.
www.cheersonic.com/portfolio-items/industry-ultrasonic-soldering-iron-for-indium/
when i worked for an electronics manufacturer that did their own refurbs and repairs in house, we used pads that required "excessive heat" to flow, but once initial flow happens, you dont have to wait for temps to hit "holly hell that coolers boiling" for it to start working properly... it was mostly used in situations where extreme heat and cycling was expected, where reflow was also expected, often the units fans would only kick on after reflow temps were hit, then the fans would kick in and cool the whole unit down drastically...
not a normal common use situation, less so now then back then even since electronics are more efficient and the units we built where some sort of industrial inverter with both stepup and step down, only thing i ever saw them used for was when tested off a huge bank of batteries and running various test equipment ac and dc, or the same units after repair used to re-charge batteries after testing, funny part was, we did all the tech work, then, the units got shipped someplace to be put in a new shell... despite us having those shells on the other side of the complex... very strange way they handled that...but... it was a job.. and watching my boss sit on a live, and amazingly hot power supply board, that was on test stool...that he was always telling people not to sit on or sat things not being worked on/tested on them... burned his ass threw his levis... so bad he ended up staying home for the rest of the week... LOL.. he was such a snotty shithead....
This is what the perforations or micro-springs are for, concentrating the clamping force down onto a smaller amount of indium or using spring forces from the indium itself. These products are actually marketed specifically for computer hardware TIM applications, which was why I brought them up.
14:53 Maybe two years ago I found guy on some laptop upgrade/mod forum that transplanted and showed the whole process of putting desktop GTX980 chip into GTX980m MXM 3.0b board and it worked, so I think It would be worth trying transplanting for example Titan RTX chip into Kingpin's 2080Ti PCB and others...
I stick to my Define R5 just because it's got a side panel where i can mount a fan. It's a good case very quiet and good temps if you configure the fans in a good way, I hate tempered glass and if i wanted a glass case I'd buy a fish Tank or a see through plastic bucket. as for RGB I couldn't give a toss about it. I only have a RGB Keyboard and Mouse just so I can see the keys etc in the dark.
One thing I want to add to the question why cases don´t have side intakes anymore: I saw one case, the Geeekcase A50, it´s a 2 chamber ITX case (similar to stuff like the DAN case) but with windows on both sides. I´m pretty sure it´s not tempered glass, but it still looks pretty good. It actually has holes in the windowed side panels! Technically it doesn´t have side intakes but the GPU fans and the CPU fan are close enough to the holes I think. I would be really interested in the performance of that case.
I've just gotten the XFX 5700xt GN recommended heavily against. It was on sale. How do I salvage it? What's the best aftermarket GPU cooler?
Hello Guys hello, my friend bought one, it’s not that bad of a card, heating is an issue though. We got the NZXT G12 bracket and a NZXT AIO (240mm liquid cooler) and it’s great. Look up the NZXT G12 bracket for a 5700xt. Best after market solution. The bracket is like $20, and the AIO, depending on which one you get, make sure it’s compatible, can be around $130. Super good results. Very quiet too.
You could also opt for an aftermarket air cooler, the arctic accelero appears to be a good option but I have little experience with aftermarket GPU solutions.
@DANK sure, you can buy the whole kit, but for someone who can't fit/doesn't want to do a full custom pump it's a good idea. Full custom is better tho.
I am currently using this exact "LIQUID MetalPad" Which I am pretty sure is the product you were talking about. and I can agree, these things are so incredibly thin (feels as thin as gold foil) that there's a good chance that, depending on how you tightened the screws on your cooler, there's noticeable imperfections. You're supposed to switch off your cooling and let the CPU "burn in" for 30+Minutes at at least 70°c for the metal to liquefy and spread out properly. This does not perform any better than standard MX-4 thermal paste. In fact I might have trouble with hot spots on my Ryzen 3900x IHS so this stuff is definitely coming off the next time I change something on my build.
I really want side mounted fans, defintely a superb way to cool hot gpus.
Mesh VS CCX design all cores clock for clock, but mesh (cache pushed) and ccx pushed to fast speeds compared to stock!
I wonder, with the glass sides if they could lay some transparent tracks on the glass like in an lcd panel to get power to a fan, or if there would not be enough material to handle the current.
What are the criteria for reviewing stuff (e.g. AIO-Coolers)? I was kinda sad that the Arctic Freezer II was not review by you guys because it seemed pretty innovative (VRM-Fan [not technically new cause Cryorig], their own pump design, one header for all Motors). So that made me wonder what your process for selecting stuff for review is.
Thermaltake had great side intake fan case, i still have it! i love this case.Level 10 case i believe is name. it is great case.
Yeah side intake fans do a great job in keeping GPU thermals in check. Have an old Antec three hundred that I bought new years ago. My first case with a side intake
Question: Have you ever consider making a page on your web with full charts with comparisons of pc cases CPU temperatures, GPU temperatures, dB levels, coolers and others? I will love to see it there to compere every case you tested in one chart.
29:55 Steve's answer to this question should be: "I am the hardware Jesus FFS! No opinion is required!" ha ha ha
Side fans are amazing, when i changed my system i went from my old case with side fan to a typical stupid glass panel. Since then i tested to run my system in my old case: ryzen 2700 at 4.15 ghz cooled with 240mm AIO, X470 Strix and Palit 2080. I dont have proper thermal gun, but AIDA showed that temps went from: 69 to 61 on CPU, 84-85 to 80 on GPU and most importantly from 87 to 69 on VRM. If my old case had dust filters and wasn't so out of date i would rather keept using it. So when i will be looking for a new case, a side fan will be a big factor for my purchase. Ps: i would also love if manufacturers once again started making custom VRM heatsinks, this would benefit many users with boards below 250$/€.
For the Indium foil and Indium HeatSpring (and similar metal "pads") vs. liquid metal question: most of the "liquid metal pads" require excessively high mounting pressure (such as found in FPGAs and other less-consumer-oriented devices) to achieve winning performance. Where a traditional consumer GPU might have mounting pressure specs around 35PSI, a well-made micro-spring metal sheet will only begin to approach low-end thermal grease - best case. Most publicly available metal pads only begin to really shine around 60-80PSI and beyond (source from one of the materials I researched: chart 3 on www.indium.com/products/thermal-interface-materials/ ).
In my experiments, I too did the stacking that Steve mentioned. It was the only way on 80% of consumer platforms to achieve reasonable pressure, but this only brought it down to temps matching thermal paste. I even started with this approach with some of the lower mounting pressure FPGAs before opting to push limits on pressure specs instead. By recklessly ignoring pressure specs, I was able to reach some impressive performance.
With lapped or otherwise better-mated IHSsink surfaces AND higher mounting pressure, the metal pads were exceptional and great for the application I was targeting (two-stage immersion cooling of FPGAs).
Making an educated guess based on these findings, I believe liquid metal would've likely performed similarly (possibly better, as there would be less of a 'barrier' than with a non-displaceable pad), but it couldn't be used in our immersion baths (imagine liquid metal escaping and floating freely in another liquid surrounding your precious computer bits) so I didn't get around to testing it at the higher mount pressures :(.
Also, the micro-spring or latticed metal pads are ungodly expensive in most cases. Non-laticed/non-spring indium foil on its own was barely worth looking at, as it carried about half the performance of the micro-spring adaptation, was painfully easy to permanently deform, and not particularly reusable as it would carry an imprint of the imperfections of the surfaces it was mating. The micro-spring was somewhat more reusable, but did also carry some deformations/imprints.
---
Sorry for the text wall, but it's a neat question and subject. I rarely get to share :P
Edit: If Steve sees this I'll be so happy ;_;
Vikram : How difficult is it ?
Steve : YES
Pin-to-pad would be stupid-complex to engineer and implement. You're talking about an additional brace along one side with not just contacts for four-pin / three pin fans but also the RGB connectors, which then would make it a miniaturized 8-pin connector.The pads would have to be mounted to the side panel, and the pins would have to be mounted to an otherwise-useless support mechanism which would not permit easy access.
There could be slots in the case body to permit drop-in installation and some tabs with indents to pop the rail in place, giving it _some_ semblance of attachment, _or_ an indented tab on front with a thumbscrew in back ala old-style cases which allowed only eight screws to hold in an optical device (and usually, it was _only_ two of the eight holes anyway) which would provide both a temporary and permanent attachment. But the means to fasten it still has to be engineered into a case. _That's not ever going to happen.
What I would like to see manufacturers of cases do is to have a cutout for the GPU, entirely. We see this with sandwich-style boards where the GPU is side-mounted with riser, so why not purpose that with mesh on a steel panel to draw in cool air from the outside? The GPU heatsink willl still work as intended to spill its hot air out, but the case, allowing for air to come in from outside directly rather than a convoluted 90-degree rotation from top, back or side would be better than in the middle next to other components. It would make for a wider case, but that's what needs to be done to re-introduce side intake.
Sapphire's Pulse RX AMD AIB cards has a very ingenious fan system where only one screw holds it on, with the fan motor assembly on a thin plastic frame. For people who would rather side _exhaust_ with their GPU, Sapphire's partners could easily produce a flipped version which spits the air out directly from the top and draw in from the sides, which would make the GPU function similarly to side-mounted exhaust for the entire system should a user prefer. More video card manufacturers making their fan assemblies easily removable from cooler housings would permit more of an interest in cases with side ventilation, provided people want this, and people _do_ want this, but they're a vocal minority the case manufacturers are not going to heed because the idea is too expansive to manufacturer, and many users... simply don't care about performance.
We've seen that 2018, we've seen that 2019, we're seeing it today. Shame it came to that, but manufacturers need to satisfy the _majority_ of users, and avoid being too niche to remain profitable. You voting with your wallet is irrelevant, and the TH-cam tech scene is also partly to blame for this, showing off concepts which choke airflow back in 2018 and people just _eating that up. _Couldn't have enough!_ Also sadly, even though some people in the tech scene _eventually_ made their cries for airflow heard, and cases are being made with airflow in mind thanks to the efforts of every Steve in the scene (people who know will get the joke), the airflow trend is going to last for maybe a year or two before case manufactures cock it up again and block off _all_ of the holes for cost savings, passing it off as for looks. It's only a matter of time.
If I have a R5 1600 and a 1060, and I have to choose between an upgrade to zen 2 or Turing, which one would be better?
Follow up question : I know this will porbably dont work but will still ask anyway .What about a 1x2 Chip socket instead of a full breadboard , soldered paralel to the saunt resistor , then just attach legs from a jumper cable onto the saunt resistors you want to test out .
Bit of a followup to the side panel fans not looking good, where personally do you draw the line between form and function with a computer? Like how much in terms of looks or acoustics are you willing to sacrifice in the name of performance?
Tempered glass it not needed for the side panel. For example Fractal design should save up and purchase the Define R5 case and examine how they were able to add a side panel fan to their solid steel side panel, and use that info to design a side panel for their newer cases.
my 2010 Area 51 Alienware case that I still have and like to play around with it because it has an ASUS P6T V2 X58 Motherboard in it... This case has Pin to Pad fan power on both side panels. I've even thought about taking the hardware off of the case in order to use it on another case to do something similar.
Would love to see you guys use the EVC2 by Elmor to push the Voltage on a 2080 Ti or a Titan RTX. I would be interested to see Turing's scaling with more voltage, that 1.093 voltage limit is annoying on Pascal.
@29:55 Did anyone else get an image in there head of Steve sitting at a round table saying: "And who pays your pay check?"
What about those fans you used to get that you could fit into your spare pci cutouts on the back of the case. I'd be interested to know if lining up one or more of those under the GPU helped thermals at all. I think they were exhaust when you could get em, but i bet you could flip the fan.
indium sheets can perform very good as tim, but you need A LOT of mounting pressure, wich is not reasonably achieveble, especially between the die and IHS.
the reason you need that much mounting pressure is the same reason you need TIM in the first place, you want to fill in microscopic gaps. to achieve that, the mouning pressure has to overcome the point where the TIM starts to expierience plastic deformation. I'm not sure if you need to maintain that level of mounting pressure or if a one time "pressing", maybe at high temperatures would be enough.
I have a UV acrylic case and it has two side fan intake slots. It was made by Logisys
And what about Graphene thermal pads? I've seen some companies make them and prices do look similar to normal thermal pads. How is that performance-wise?
On another topic - GPU transplants were common in laptop industry because there was a series of faulty products back around 2009. What you generally need is a sieve for specific chip to place BGA's in correct positions and hotair soldering station. Until now I didn't seen anyone setting BGA's by hand.
would zombiemodding epyc make sense for cinebench?
Yeah Im using IC Diamond Paste on my gaming build, seem to cool good, my 3900x never go over 70% under full load. However,, not the easiest paste to work with its very thick, so ideally the tube under the sun 10 min before applying would help it spread.
BTW, don't try to drill through tempered glass, because in the tempering process the glass gets a ton of internal stresses, and if you drill into it the forces will be released, and the glass explodes.
Dear Steve and GN. I was wondering if turning CPU to real time or high etc really does help gaming or multithreaded operations in any way whether it be fps, latency, etc. I sweat that 1% lows or avg fps seem to be higher when I go to real time
9900k@5.0 in my case 2080ti 1080p low 240hz trying for 240fps+ in everything
Viewer; How diffucult to transplant gpu?
GN's Steve; Yes
I still use my old antec case that I got when I built my first computer in 2012-2013 and it has a fan mount on the side panel. That thing is fucking amazing for my GPU temps and I never want a case without it. Sure you can make arguments that if I had a bigger case with a better design I could also improve my GPU temps, but that side mounted fan REALLY helps my GPU in this case. I hate glass-sides on computer cases, I just want a black box that's functional and that's what I've got.
I think it would be really cool for case manufacturers to put a fan mount on the rear of the case, so it would blow air in through the PCI slots
Could we have some content covering silicon manufacturing technologies- we see clips of fabs, and mentions of different processes, but it'd be nice to understand what actually happens in a fab- some forms of lithography, apparently, but more detail(as far as processes being proprietary can take us) would be great?
The first point about contact being an issue, makes me think, would it be feasible for CPU manufacturers to make the die from a solid piece of metal that includes the heatsink? So like the die, IHS and heatsink are all 1 piece and the chip is built around that? considering the difference between some of intel's stock TIM and regular paste VS liquid metal between IHS and heatsing, you could probably get insane cooling on air alone with relatively small heatsinks. Maybe make instead of a heatsink make it 1 piece as a standard fitting waterblock that you can attach any pump/rad combo to. I know when I delidded and repasted my i7-7700k it dropped from like mid 90s to 65°C peaks from paste alone.
Awesome Ask GN, Steve! Would love to see u guys hit up Vince and revisit EVGA HQ when they're open again...
I had coolaboratory with H100i + i3570k. Two years. Then it let the exact same print on the block and CPU as the original Corsair paste. The OEM was on it for 5 years before.
You have to polish both surfaces as the instruction says. So warranty void. It had 5c better delta. In 550D full with HDDs and air cooled XFX RX 580 GTR-S uperhot. Low RPM intake and exhaust.
The alternatives to liquid metal video would be extremely interesting. If possible could you test the graphene pad as alternative to thermal paste. that LTT did a video on?
GPU solder mod question: How about soldering a socket onto the board that allows you to swap shunt resistors like op amps in a headphone amp?
@GN The Pins to Pads Idea for Side Mounted Fans had been done before, Thermaltake did it with their Level 10 GT Case, I still use my Level 10 GT Snow Edition Case and its Side Fan has Connectors at the Hinge. I don't know if any other manufacturers did or if Thermaltake had Patended the idea or not. I do know why they stopped doing side mounted fans is mostly due to the glass trend like you said however. I do think that they could look decent together with the right designs though, kind of like how the Level 10 looks except with a bigger glass panel possibly and maybe a smaller fan, but it definitely could work!
Indium solutions can be as good as 80 W/mK, but there are numerous issues with using them as thermal pads in DIY computer applications. You mentioned the issues with compression and flow at low pressure and temperature, but there are also issues with oxide formation, especially with high In content solutions. Mostly I've seen them in power semiconductor and photonic applications, as well as cryogenic applications. It might be interesting to use In for cryo cooling for extreme OCing, but applying it would be more like soldering the die to the pot than applying thermal grease. Another potential use case would be in high power modular systems, like those found in military and aerospace environments. Ultimately, as cool as the idea is I can't see it being practical outside of extreme applications.
I work in the architect/construction industry and deal a lot of with Revit workloads. Would you consider adding a Revit benchmark like RFO to your CPU testing. I think that that Revit is mostly single core frequency dependant workload, but I'd be interested to see where ryzen 3k stacks up.
so if wire has enough resistance couldn't you find the right gauge wire and just jump from one side of the shunt to the other? i mean a thick enough or long enough wire would cause resistance?
Hey Steve.
Are you down with the thickness??
Thermaltake lvl 10 gt snow edition full tower has a large side intake fan with pin to pad power connection without wires from the motherboard to the fan itself.