I am not interested in hardware. I stopped playing cs a long time ago. But I still watch all the kliksphilips. Been subbed since i was in year 8 and im now in uni. Klik empire is too good to give up
@@redwiltshire1816 Yeah, same here. I don't feel like clickbait titles/thumbnails on youtube respect me as a person tbh. If someone's so desperate for my attention I don't wish to engage with their video
I have an okay middle-ground solution: If something really interests me, I do two things: 1. Always follow the rule, "If it's too good to be true, it is". Even without knowing a subject very well, there's some things that just seem plain fanciful. 2. If possible, try to find friends or connections with people who do know the subject. These people are often a gold mine of information when you need to ask questions without wanting to put in all the time and effort to learn the subject. The best solution to counter misinformation and lies is to be omniscient, but that's not possible so you should always have multiple strategies in place to deal with it.
Like I tell everyone always get multiple sources or even try the product through a friend or even a demo I find products with no demos tend to have something to hide
ok, but you can, if you don't try hard, just get the same general narrative spun to you, from different sources. And of course, just going to someone counter to whatever narrative won't necessarily free you from Bs either. IOW - your post seems good at first blush, but really, it's not going to save you from BS, in fact, it might make your more susceptible to different kinds than the one you attempt to avoid, or you just fall harder into it.
Random friends probably aren't knowledgeable. They're just talking confidently, you have no idea where they got that information. "I read it," could mean a Twitter post.
@@friendofp.24 Okay, but random posts online also "probably aren't knowledgeable", ala user benchmark. The best you can do is receive a high volume of data and then sift through it, since if you only need that for a single decision, it's easier than learning that entire subject just to, say, buy a graphics card. And "random friends" do tend to have done more research than you have in the area, so they're a great starting place, much better than you seem to point them out to be. That, or you just have very stupid friends.
At first i laughed at people worried about AI misinformation, "Heheh why worry about that when sites like userbenchmark exists?" Then I realized AI is scraping userbenchmark data..... and that yes we must fight against AI misinformation by informing the AI what sites are to be trusted.
I love that little detail. If I was going to create 'hype train' art, it would be a hundred exhausts and one tiny little carriage with a single, rotten piece of fruit on it.
How come no one has managed to track down the owner of User benchmark for an interview yet. I sure as heck would watch that. I would love to hear what AMD did to him back in the day that spawned this long campaign of hate and how it became such an all consuming thing.
I've known people that are like UB's reviews, where they just _despise_ something for seemingly no reason? What probably happened was they kept saying 'Intel > AMD' for many years, and then AMD started getting good, and they didn't want their advice they'd spread for the last 5+ years to become incorrect So instead of updating it to fit with... reality... they doubled-down and kept trying to find reasons for why 'Intel > AMD', even when it wasn't Years of digging their own hole, and they're now at a point where any choice is a loosing scenario: continue being ridiculed for spreading misinformation, or change opinion and get ridiculed for spreading BS for several years lol
the amount of mixed feedback from over excited bearded youtube tech channels that repeat DLSSSFSRXESSAMDAMDAMDFSR79007800X3DX3DVRR999999999 take up most of the time in videos when all of it can be summarized in a bar graph
The problem with MLID is that you get really excited about the hardware, because there are nice new ideas, but ultimately it doesn't matter because they cost so much on release. The rx7x00 cards did have the architecture features he leaked, but it didn't matter because they had worse price/performance than the rx6x00 products. He doesn't get accurate price leaks until a few weeks before launch anyway, by which time you might as well wait until launch.
@@hoholu4650this is exactly what Phil criticizes in his video "rDNA 4 is cancelled" It's obviously not the case, but one specific design feature of rdna4 gets canned "See I was like almost half right anyways"
I liked the older AdoredTV videos when he analysed die sizes and things like that to see what might come next gen rather than straight up providing leaks, but ultimately that was only for interest, because once again, without pricing it's power was limited.
Man these thumbnails in the beginning make me furious lol. "7800x3d still enough?!".. Sitting here with a 5800x3d that absolutely slaps. I also have a 3070 ti (which had a good price when I got it, even if it's considered terrible value) oh so sad, it can barely run MS paint with its low VRAM, my whole system is in danger, oh no! Guess what.. everything runs fine, most games that are terribly optimized VRAM hogs are shitty anyways so why bother? I play niche, easy to run games and most banger games are well optimized too, so why bother with fear of missing out? Nvidia hasn't released a single good price-performance ration GPU in ages, so why should I buy anything now. I am comfy af rn with what I have. Meanwhile I know a dude who is still paying monthly amounts to pay off his 4090 NASA PC that nobody needs. The one thing I learned from this: don't by things with money you don't have!
Awesome advice, thanks Philip! There's just one small issue: the time someone spends watching leaks can only be reinvested elsewhere if they already suspect the foul play. If they don't, they'll probably see the leaks as the legitimate info they were looking for -- meaning people already have to be suspicious to discover the lie, and no one has the energy to be suspicious of everything all the time. Still a solid video though.
I guess the best advice is to be reasonable. Do not trust fanboys or hype factories. When things seem too good to be true then they usually are. Keep in mind Companies always want your money, giving you a small but steady increase sadly is way more financially wise and therefore more realistic...also never give a company a personality, do NOT fall for that, they aren't your friend and when they suck, they suck.
Here is a thing that I think you will find intresting: Test Horizon Forbidden West with a RTX 30 series card and the DLSSFG to FSR 3 in Nexus mods by nukem. Maybe compare the results with a performance wise similar RTX 40 series card. Note: be sure to turn off any kind of AA or Upscaler inside the game when using the mod. Have fun!.
I actually do not really care about hardware, I am just here for you. 6:05 To quote Sherlock Holmes: "It is a capital mistake to theorize before you have all the evidence. It biases the judgment."
First time using an AND (ryzen 6 6600H) and a proper graphics card (nvidia GTX3060). Having fun actually playing games. A shame that the AMD is so much worse than any intel.
lmao, getting MLID in suggested videos. Anyway, I feel there IS substance to have Intel-fanboys writing reviews that showcase best-case scenarios, AND have AMD shills writing why intel isn't even close. They'll bring the most points to the debate (as centrist channels will drown down focus from the best and worst qualities, unless it makes them a good title). I believe Userbenchmark isn't making stuff up about intel's product line straight on, but instead create scenarios that are just favourable (probably feeling injust with most coverage favouring AMD, when image is not as clearcut to make such simple verdict).
Great video with very good points. I've been following tech and PC hardware since the early 90s and I agree that a wide array of information sources is the best start. Over time you'll get a feel for who is reliable, accurate, consumer advocacy oriented, and least biased. Ditch the news sources who are inconsistent or show obvious bias, BUT be conscious of your own bias (we all have some). Don't fall for the tribalistic idea of "brand loyalty" like so many people do. Corporations are not your friend. Reward them when they take care of you, and go elsewhere when they won't.
I still think it would be nice if the steam hardware survey included a simple benchmark. A simple single core and multicore cpu benchmark a file copy and simulated game loading SSD benchmark and a simple short GPU benchmark with and whiteout RT. Right now the CPU survey is useless and there are just too many CPU models to get any info out of when designing a game. Say I want to make a game that can run on 50% of all steam users pc. Then it would be nice to get an approximate value of how fast a pc should be able to run the game. It would never be perfect and you should still look for reviews when buying hardware. But to get an idea about how fast users pcs are when designing games I think it could be a good tool. And It should only be a 5 min benchmark nothing overly complicated. And yes there are big differences from game engine to game engine, but if steam was open about how the test is conducted then the developers could take that into account. Maybe valve could collaborate with game developers. Epic could do the same thing but they would obviously use the Unreal Engine, might be a way they could stand out when compared to other game launchers.
User benchmark for experienced PC building community has been a joke for well over a decade and Tom's hardware has lost much credibility in the same amount of time for it's blatant and unfounded fanboyism.
"so much bias in the past" -nope, they're still doing it. read the latest test, that one guy is still blabbing on about AMD committing tax fraud [this is a joke]
I get that you really cant compare various hardware to yours, using this. I tend to use it to compare the your hardware against the same hardware for quick diagnostics or to verify everything is running okay after a rebuild or mujor OS update. Know that here are plenty of more accurate and relevant tests but this is easy to install, quick to run and supplies passable info when comparing to the same hardware.
A problem I find is that even if you have data, interpreting it can go very wrong. Two of the channels mentioned, I'd tolerate for raw data but they spew nonsense when they start giving their opinion on things. If I'm generous, they can't be that stupid and are playing their core audience to stir the drama pot. At worst, they are ignorant of how the real world works and are trying to impose their distorted reality.
I find it very cost effective to buy the top of the line hardware from one or two generations behind. Used prices are great, performance to cost ratio is the best, good research and thorough testing has already happened! Bugs fixed, etc. I don't really care about speculations and future releases.
7:52 "Whatever you choose to buy, at any time, will always be a gamble of sorts - and will only ever be fully judged with hindsight years after" I bought a i7-7700k not too long after it - and the first gen Ryzen CPU's - came out and have somewhat regretted it ever since having seen what an incredible thing AM4 has become. I previously had a i5-3570k that was starting to struggle in some games, and because I knew nothing of AMD CPU's at the time I wanted to keep it simple, I just went with the best Intel I could get. Had I waited but a few months longer, I could have gotten an i7-8700k (same CPU as my i7-7700k but with 6 cores instead of 4) or if I'd had a crystal ball and could see in to the future I could have gotten a Ryzen system and been able to upgrade to the 5800X3D years later.
I've seen many pages of Userbenchmark's CPU comparisons and I think even GPUs too, with them talking shit about AMD. I get that with the GPUs that AMD sometimes has disadvantages like RT performance and FSR is always a step behind DLSS, but they often have advantages like lower prices and more Vram. Just look up their review of the R7 5800X3D. They say it's a slightly slower clocked 5800X with 3x the cache, which is true. But then they say it performs the same in real world tasks, which is certainly does not, in gaming it's always faster if your GPU can make use of it, maybe at the time it was having firmware/driver issues or something, but they insulted people who were saying it was a good product.
I also like reviews from IgorsLab (videos are german only, but articles are in english as well) because he includes a lot of metrics, some that are rarely seen in other reviews, and der8auer because he often rather than doing a general review, focuses on a single aspect of the product and goes into greater detail about it, giving you a better understanding of it and it's importance.
I think UserBenchmark/3D Mark is usefull to compare the performance of your CPU/GPU to the same CPU/GPU, so you know how well your specific hardware is performing, as in looking up if you won or lost the silicon lottery. 3D Mark is great for dialing in an underclock (as overclocking is worth it with current hardware).
The problem with a lot of reviews is they no longer include RX480 8GB or Core i5 6600 or Star Wars: The Old Republic. It would be nice to know whether I can max out settings for that game on newer hardware, so I don't have to turn down settings at 4k.
Userbenchmark still up? wtf xd 7:40 This Remind me when I told my parents that I need Laptop many years ago, and they want to "surprise" me with scuffed laptop, I said to them "Don't do this again". xd
It would be an interesting video for one of these hardware reviewers to do to see how these benchmark softwares correlate to different game performance. IE you'd see how a set of different hardwares stack up in a few games and a few benchmark softwares and see how statistivally correlated the benchmark scores are for each game's framerate.
A long time ago, we could just tell which CPU was better by how many mhz or ghz was written on the box. You could do the maths right that the shop about price to performance. Then it seemed 3ghz seemed to be the hard ceiling, so it turned into number of cores. But two cores wasn't twice as good as one core. It was nebulous and comparing cpus with the same core number was impossible. So these benchmark websites seemed like a good way to surpass that limitation. Each bit of kit would crunch numbers and you could judge which thing was best by that extra step. Now 20 years later, we still have CPUs with 4 to 8 cores that are all clocked about 3ghz, but somehow they have massive benchmark scores compared to ones with similar setups from years before.
While I agree that of course a single benchmark doesn't tell the whole story, I don't know if it's fair to say that a single benchmark is basically useless if it's not for the specific task you'll regularly use. For example, a GPU that performs really well in one AAA game that you never play is also likely to perform well in almost all other games, including those you do actually play. The workloads tend to be similar enough that the data really can be expanded to a whole group of use cases. The same thing applies for all computing components (CPUs or AI accelerators) and for different types of workloads as well. If an AI accelerator performs very well at one CNN, it will likely perform reasonably well on all CNNs (though I'd argue this comes with more caveats than gaming does for consumer products). This really only starts to break down once the accelerators become too specialized. For most people's use cases, I think it may not be so bad to boil down performance to a single number, so long as the number is generally truly representative of the real use case of the user.
I've been actively avoiding user benchmark to not boost their click rate. I've been using TH-cam videos(sometimes inaccurate if by random channels), gamersnexus, ltt the lab, and technical city. Technical city is fairly good, but also inaccurate at times, still way more accurate than the meme userbenchmarks
3:20 - I just look at someone playing some games with the GPU I am thinking of. This does mean I already need to know which GPU I want..Or rather GPUs if I find out something is better suited for me using these videos.
There is a use for... User benchmark. It shows how your setup compares to others running the same components. It makes inadequate cooling or poor memory configurations stick out like a sore thumb. It's a quick sanity check, not something to make buying decisions on.
Yes, the fact that User Benchmark suddenly lowered the importance of higher core count and multitasking once AMD exceeded Intel in core counts a few years back left a bad taste in my mind. It does make sense depending on what you're doing with said computer. Cores matter in some cases. Most every game at the time did not and Intel wanted to make sure everyone knew that their smaller core count ran better for gaming. You could argue in their defense that for gaming and other single core processing, the Intel chips were usually better. But, if they are a good source for benchmark results, then they should allow you to weight features based on few questions, provide different comparison results based on use categories, and not assume that everyone is just gaming. And this type of issue could easily be attached to any component to skew the results for one side or another. This is the major reason why one test cannot evaluate a component or system. Thanks for sharing!
There is actually a similar site, i forgot how its called but you enter the cpu and gpu and it spits out fps for a specific game. It relies on user data but i crossreferenced it with my system and it seemed pretty acurate. Also tells you bottlenecks and stuff
2:02 - "You shouldn't just rely on one source", - that's right, that's why we needed Source 2.
Which is 2kliksphilip.
@@jacobshirley3457 And 3kliksphilip is Source 3.
I wish you went back to clickbaiting us with your handmade thumbnails sweety 😔💅💅
What
I mean it is handmade - he prompted an AI for the train, touched it up, put on the classic lens flare and the font.
@@darbyshiredanny97that's not handmade... That's touched up.
@@jimmyegg6892 ai bad because ai
Cry about it
so how do we trust philip
-burger40
Ask GN or HUB to benchmark kliks, 2kliks and 3kliks, see which one is best
thank u burger40
good question burger40
by reading TH-cam COMMENTS, of course
nothing more trustworthy, friendly and... "nice :)" like the good people here, down in the comments below.
What happened to the 6?
My first graphics card was a gt210 because I didn't do any research before buying it.
F
F
I thought my first card being a gt320 was bad, but it seems it's 300% better than the gt210 lmao
F
@@benperkins2929 Yep haha, In a way I got clickbaited by the product page
the amount of people talking about snapdragon like it's some holy grail made the bull****** sensors go off in my head long ago.
It's gonna be a M1 with shit support in the first year and full of bugs.
Snapdragon is just smartphone chip in laptop, weak as hell
@@GameBacardi Qualcomm acquired nuvia, who were ex-apple CPU engineers, making a new arm chip. That is what X Elite is based on.
Youre seriously underestimating arm... The surface pro x has a much older snapdragon and is still crazy fast for what it is
it can run mario bros at 12000000000000fps
I am not interested in hardware. I stopped playing cs a long time ago. But I still watch all the kliksphilips. Been subbed since i was in year 8 and im now in uni. Klik empire is too good to give up
REAL
De-arrow helps with clickbait titles and thumbnails
anyone reading this, check out the de-arrow extension. I would also highly recommend it.
On TH-cam
Honestly any thumbnail with over the top mr beast text i avoid
@@redwiltshire1816 Yeah, same here. I don't feel like clickbait titles/thumbnails on youtube respect me as a person tbh. If someone's so desperate for my attention I don't wish to engage with their video
Wasn't de_arrow removed from the competitive pool?
I really do like that you use subtitles.
Thanks! I run em through an AI program then manually review them afterwards. Saves a lot of time
0:50
Oh you're very wrong. This video doesnt seem to be about any of those from the title alone, and im just following 2kliksphilip.
I have an okay middle-ground solution: If something really interests me, I do two things:
1. Always follow the rule, "If it's too good to be true, it is". Even without knowing a subject very well, there's some things that just seem plain fanciful.
2. If possible, try to find friends or connections with people who do know the subject. These people are often a gold mine of information when you need to ask questions without wanting to put in all the time and effort to learn the subject.
The best solution to counter misinformation and lies is to be omniscient, but that's not possible so you should always have multiple strategies in place to deal with it.
@elliejohnson2786 Mejor Comentario 10/10
Like I tell everyone always get multiple sources or even try the product through a friend or even a demo I find products with no demos tend to have something to hide
ok, but you can, if you don't try hard, just get the same general narrative spun to you, from different sources. And of course, just going to someone counter to whatever narrative won't necessarily free you from Bs either.
IOW - your post seems good at first blush, but really, it's not going to save you from BS, in fact, it might make your more susceptible to different kinds than the one you attempt to avoid, or you just fall harder into it.
Random friends probably aren't knowledgeable. They're just talking confidently, you have no idea where they got that information. "I read it," could mean a Twitter post.
@@friendofp.24 Okay, but random posts online also "probably aren't knowledgeable", ala user benchmark. The best you can do is receive a high volume of data and then sift through it, since if you only need that for a single decision, it's easier than learning that entire subject just to, say, buy a graphics card.
And "random friends" do tend to have done more research than you have in the area, so they're a great starting place, much better than you seem to point them out to be. That, or you just have very stupid friends.
At first i laughed at people worried about AI misinformation, "Heheh why worry about that when sites like userbenchmark exists?" Then I realized AI is scraping userbenchmark data..... and that yes we must fight against AI misinformation by informing the AI what sites are to be trusted.
You should be worried about misinformation in news and politics, not hardware
☝️🤓
Userbenchmark's benchmarking software looks like a game from 20+ years ago. There's no chance that could ever be relevant to modern games.
Please make more userbenchmark videos in the future. The website is extremely damaging for beginner pc enthusiasts that are just learning.
Good video, bashing Userbenchmark is somethign I can always get behind too!
Hype train has 3 exhausts 😂😂
I love that little detail. If I was going to create 'hype train' art, it would be a hundred exhausts and one tiny little carriage with a single, rotten piece of fruit on it.
I used to like Philip but his cringe AI thumbnails really put me off.
@@overdark33Hahahaha
@@overdark33 would you rather want him to draw his own cringe thumbnails in ms paint? the effect is the same
@@okarowarrior yes actually
How come no one has managed to track down the owner of User benchmark for an interview yet. I sure as heck would watch that. I would love to hear what AMD did to him back in the day that spawned this long campaign of hate and how it became such an all consuming thing.
I've known people that are like UB's reviews, where they just _despise_ something for seemingly no reason?
What probably happened was they kept saying 'Intel > AMD' for many years, and then AMD started getting good, and they didn't want their advice they'd spread for the last 5+ years to become incorrect
So instead of updating it to fit with... reality... they doubled-down and kept trying to find reasons for why 'Intel > AMD', even when it wasn't
Years of digging their own hole, and they're now at a point where any choice is a loosing scenario: continue being ridiculed for spreading misinformation, or change opinion and get ridiculed for spreading BS for several years lol
LoserStenchmark
epically pwned
LOL
the amount of mixed feedback from over excited bearded youtube tech channels that repeat DLSSSFSRXESSAMDAMDAMDFSR79007800X3DX3DVRR999999999 take up most of the time in videos when all of it can be summarized in a bar graph
7:03 Bro casually attacked VideoCardz, Moore's Law is Dead, RedGamingTech, Gamer Meld and Graphically Challenged
Tbh dont understand his beef towards MLD, K the RDNA4 screen title was overblown but at the same time right, chiplet RDNA4 was cancelled..
The problem with MLID is that you get really excited about the hardware, because there are nice new ideas, but ultimately it doesn't matter because they cost so much on release.
The rx7x00 cards did have the architecture features he leaked, but it didn't matter because they had worse price/performance than the rx6x00 products.
He doesn't get accurate price leaks until a few weeks before launch anyway, by which time you might as well wait until launch.
@@hoholu4650this is exactly what Phil criticizes in his video
"rDNA 4 is cancelled"
It's obviously not the case, but one specific design feature of rdna4 gets canned
"See I was like almost half right anyways"
I liked the older AdoredTV videos when he analysed die sizes and things like that to see what might come next gen rather than straight up providing leaks, but ultimately that was only for interest, because once again, without pricing it's power was limited.
@@Daniel-ul8kr except MILD never claimed anywhere that the full stack was cancelled, its just a clickbait thumbnail
Well the previous video just happened to be by Hardware Unboxing 😂😂
your thumbnails kinda look way worse nowadays i wonder why
You're right, what am I doing here? I don't care about hardware.
0:44 im not into hardwate, but im not following rumor mills or userbenchmark. im just watching your videos
I think you'd be a great podcast host, but I'm glad you're not because podcasts are lame
LOL
I wonder why this bodcast trend have been forced a lot in this year everywhere :S it is crab
Lmao
Click on a video that says it's about clickbait and lies...
It's actually about computer hardware...
c'mon phillip
Man these thumbnails in the beginning make me furious lol. "7800x3d still enough?!".. Sitting here with a 5800x3d that absolutely slaps. I also have a 3070 ti (which had a good price when I got it, even if it's considered terrible value) oh so sad, it can barely run MS paint with its low VRAM, my whole system is in danger, oh no!
Guess what.. everything runs fine, most games that are terribly optimized VRAM hogs are shitty anyways so why bother? I play niche, easy to run games and most banger games are well optimized too, so why bother with fear of missing out? Nvidia hasn't released a single good price-performance ration GPU in ages, so why should I buy anything now. I am comfy af rn with what I have. Meanwhile I know a dude who is still paying monthly amounts to pay off his 4090 NASA PC that nobody needs. The one thing I learned from this: don't by things with money you don't have!
Phoronix is also a great resource
What's crazy is that people are so polarized on which site is good or bad. It's as bad as politics in the US
Is it making a channel based on doing really long farts?
My grandmother always gave me a very simple philosophy: If it seems too good to be true, it's likely false.
Userbenchmark benchmark software looks worse than Bungholio Marks!
Awesome advice, thanks Philip! There's just one small issue: the time someone spends watching leaks can only be reinvested elsewhere if they already suspect the foul play. If they don't, they'll probably see the leaks as the legitimate info they were looking for -- meaning people already have to be suspicious to discover the lie, and no one has the energy to be suspicious of everything all the time. Still a solid video though.
I guess the best advice is to be reasonable. Do not trust fanboys or hype factories. When things seem too good to be true then they usually are. Keep in mind Companies always want your money, giving you a small but steady increase sadly is way more financially wise and therefore more realistic...also never give a company a personality, do NOT fall for that, they aren't your friend and when they suck, they suck.
hi philip i must say i am very happy to see captions returning in force
I love your video “nightwalker” I think I’ve seen it 20+ times. You have a very nice perspective of the night. Thank you
Here is a thing that I think you will find intresting:
Test Horizon Forbidden West with a RTX 30 series card and the DLSSFG to FSR 3 in Nexus mods by nukem.
Maybe compare the results with a performance wise similar RTX 40 series card.
Note: be sure to turn off any kind of AA or Upscaler inside the game when using the mod.
Have fun!.
I actually do not really care about hardware, I am just here for you.
6:05
To quote Sherlock Holmes: "It is a capital mistake to theorize before you have all the evidence. It biases the judgment."
misinforming me philip???
lol hardware unboxed
Chad and unfathomable based philip shitting on userbench, all is right with the world.
passmark is a great alternative to userbenchmarks database of various hardware that is completely unbiased
Yeah I've used Passmark for a while now, it seems fairly accurate
can you compare shit on there? i want to see comparisons of synthetic performance, cant do that on userbenchmark with weird chips for whatever reason
@@balukrol yes you can actually
The only true benchmark is Aquamark. Everyone knows that.
First time using an AND (ryzen 6 6600H) and a proper graphics card (nvidia GTX3060). Having fun actually playing games. A shame that the AMD is so much worse than any intel.
lmao, getting MLID in suggested videos.
Anyway, I feel there IS substance to have Intel-fanboys writing reviews that showcase best-case scenarios, AND have AMD shills writing why intel isn't even close.
They'll bring the most points to the debate (as centrist channels will drown down focus from the best and worst qualities, unless it makes them a good title).
I believe Userbenchmark isn't making stuff up about intel's product line straight on, but instead create scenarios that are just favourable (probably feeling injust with most coverage favouring AMD, when image is not as clearcut to make such simple verdict).
Great video with very good points. I've been following tech and PC hardware since the early 90s and I agree that a wide array of information sources is the best start. Over time you'll get a feel for who is reliable, accurate, consumer advocacy oriented, and least biased. Ditch the news sources who are inconsistent or show obvious bias, BUT be conscious of your own bias (we all have some). Don't fall for the tribalistic idea of "brand loyalty" like so many people do. Corporations are not your friend. Reward them when they take care of you, and go elsewhere when they won't.
I still think it would be nice if the steam hardware survey included a simple benchmark. A simple single core and multicore cpu benchmark a file copy and simulated game loading SSD benchmark and a simple short GPU benchmark with and whiteout RT. Right now the CPU survey is useless and there are just too many CPU models to get any info out of when designing a game.
Say I want to make a game that can run on 50% of all steam users pc. Then it would be nice to get an approximate value of how fast a pc should be able to run the game. It would never be perfect and you should still look for reviews when buying hardware. But to get an idea about how fast users pcs are when designing games I think it could be a good tool. And It should only be a 5 min benchmark nothing overly complicated. And yes there are big differences from game engine to game engine, but if steam was open about how the test is conducted then the developers could take that into account. Maybe valve could collaborate with game developers.
Epic could do the same thing but they would obviously use the Unreal Engine, might be a way they could stand out when compared to other game launchers.
Nice video about protecting yourself against BsMeld or Graphically clikcbaits.
It's easy, just follow hardware unboxed. There is one rumour I trust though. Daniel Owens, I think he is very balanced
User benchmark for experienced PC building community has been a joke for well over a decade and Tom's hardware has lost much credibility in the same amount of time for it's blatant and unfounded fanboyism.
>GN and HUB
How many angry Linus stans did you end up up getting off this video, Philip?
Personally I’m in the “why would I look at user benchmark and runner mils in the first place” category
"so much bias in the past" -nope, they're still doing it. read the latest test, that one guy is still blabbing on about AMD committing tax fraud [this is a joke]
I get that you really cant compare various hardware to yours, using this.
I tend to use it to compare the your hardware against the same hardware for quick diagnostics or to verify everything is running okay after a rebuild or mujor OS update.
Know that here are plenty of more accurate and relevant tests but this is easy to install, quick to run and supplies passable info when comparing to the same hardware.
About the SD chip it seen to be great, but i hate how RISC compare with CISC, they don't work the same to measure performance with numbers
Why isn't there a website that compares actual game performance
A problem I find is that even if you have data, interpreting it can go very wrong. Two of the channels mentioned, I'd tolerate for raw data but they spew nonsense when they start giving their opinion on things. If I'm generous, they can't be that stupid and are playing their core audience to stir the drama pot. At worst, they are ignorant of how the real world works and are trying to impose their distorted reality.
people will always disagree against their Bias
like 4060 conversation is worse than 3060 Ti
it's not really the hardware
it's the future promise of escaping this hellhole called reality
and plug me in forever
I bought a razer laptop with a 1060 for full retail right before they dropped rtx gpus lol
TH-cam Clickbait Remover is fantastic on firefox
ok I've been out of the klikosphere for a bit but I now hear why some people think he's using ai voices now
I only care about hardware news when its time for upgrade.
YES. The day has come for a NooserNenchNark video.
i thought you only did counter strike videos until i discovered this channel.
i love' this
all my excitment for unreleased tech went with Vega
Please do the 9000 series
Linus Tech Tips is cringe.
Thanks for listening
ugh i hate bser uarchmark
I find it very cost effective to buy the top of the line hardware from one or two generations behind. Used prices are great, performance to cost ratio is the best, good research and thorough testing has already happened! Bugs fixed, etc. I don't really care about speculations and future releases.
Make another operation flashpoint video please
dont trust in the law.
well all of your points can be said about any topic irl
UserBenchmark is so trash XD
7:52 "Whatever you choose to buy, at any time, will always be a gamble of sorts - and will only ever be fully judged with hindsight years after"
I bought a i7-7700k not too long after it - and the first gen Ryzen CPU's - came out and have somewhat regretted it ever since having seen what an incredible thing AM4 has become.
I previously had a i5-3570k that was starting to struggle in some games, and because I knew nothing of AMD CPU's at the time I wanted to keep it simple, I just went with the best Intel I could get.
Had I waited but a few months longer, I could have gotten an i7-8700k (same CPU as my i7-7700k but with 6 cores instead of 4) or if I'd had a crystal ball and could see in to the future I could have gotten a Ryzen system and been able to upgrade to the 5800X3D years later.
I've seen many pages of Userbenchmark's CPU comparisons and I think even GPUs too, with them talking shit about AMD. I get that with the GPUs that AMD sometimes has disadvantages like RT performance and FSR is always a step behind DLSS, but they often have advantages like lower prices and more Vram.
Just look up their review of the R7 5800X3D. They say it's a slightly slower clocked 5800X with 3x the cache, which is true. But then they say it performs the same in real world tasks, which is certainly does not, in gaming it's always faster if your GPU can make use of it, maybe at the time it was having firmware/driver issues or something, but they insulted people who were saying it was a good product.
I also like reviews from IgorsLab (videos are german only, but articles are in english as well) because he includes a lot of metrics, some that are rarely seen in other reviews, and der8auer because he often rather than doing a general review, focuses on a single aspect of the product and goes into greater detail about it, giving you a better understanding of it and it's importance.
I think UserBenchmark/3D Mark is usefull to compare the performance of your CPU/GPU to the same CPU/GPU, so you know how well your specific hardware is performing, as in looking up if you won or lost the silicon lottery.
3D Mark is great for dialing in an underclock (as overclocking is worth it with current hardware).
Personally after I watched your first video I switched to looking at 3D mark scores since those are available for everyone online!
A Userbenchmark video? Philip! You know April 1st was LAST month, right?
The problem with a lot of reviews is they no longer include RX480 8GB or Core i5 6600 or Star Wars: The Old Republic. It would be nice to know whether I can max out settings for that game on newer hardware, so I don't have to turn down settings at 4k.
Userbenchmark still up? wtf
xd
7:40 This Remind me when I told my parents that I need Laptop many years ago, and they want to "surprise" me with scuffed laptop, I said to them "Don't do this again". xd
It would be an interesting video for one of these hardware reviewers to do to see how these benchmark softwares correlate to different game performance.
IE you'd see how a set of different hardwares stack up in a few games and a few benchmark softwares and see how statistivally correlated the benchmark scores are for each game's framerate.
A long time ago, we could just tell which CPU was better by how many mhz or ghz was written on the box. You could do the maths right that the shop about price to performance. Then it seemed 3ghz seemed to be the hard ceiling, so it turned into number of cores. But two cores wasn't twice as good as one core. It was nebulous and comparing cpus with the same core number was impossible. So these benchmark websites seemed like a good way to surpass that limitation. Each bit of kit would crunch numbers and you could judge which thing was best by that extra step. Now 20 years later, we still have CPUs with 4 to 8 cores that are all clocked about 3ghz, but somehow they have massive benchmark scores compared to ones with similar setups from years before.
While I agree that of course a single benchmark doesn't tell the whole story, I don't know if it's fair to say that a single benchmark is basically useless if it's not for the specific task you'll regularly use. For example, a GPU that performs really well in one AAA game that you never play is also likely to perform well in almost all other games, including those you do actually play. The workloads tend to be similar enough that the data really can be expanded to a whole group of use cases. The same thing applies for all computing components (CPUs or AI accelerators) and for different types of workloads as well. If an AI accelerator performs very well at one CNN, it will likely perform reasonably well on all CNNs (though I'd argue this comes with more caveats than gaming does for consumer products). This really only starts to break down once the accelerators become too specialized. For most people's use cases, I think it may not be so bad to boil down performance to a single number, so long as the number is generally truly representative of the real use case of the user.
I've been actively avoiding user benchmark to not boost their click rate.
I've been using TH-cam videos(sometimes inaccurate if by random channels), gamersnexus, ltt the lab, and technical city.
Technical city is fairly good, but also inaccurate at times, still way more accurate than the meme userbenchmarks
3:20 - I just look at someone playing some games with the GPU I am thinking of. This does mean I already need to know which GPU I want..Or rather GPUs if I find out something is better suited for me using these videos.
thanks for this video
Philip's GPU naming hierarchy comparison was instrumental to me when i was just starting my journey into understanding CPU components.
There is a use for... User benchmark. It shows how your setup compares to others running the same components. It makes inadequate cooling or poor memory configurations stick out like a sore thumb.
It's a quick sanity check, not something to make buying decisions on.
Yes, the fact that User Benchmark suddenly lowered the importance of higher core count and multitasking once AMD exceeded Intel in core counts a few years back left a bad taste in my mind. It does make sense depending on what you're doing with said computer. Cores matter in some cases. Most every game at the time did not and Intel wanted to make sure everyone knew that their smaller core count ran better for gaming. You could argue in their defense that for gaming and other single core processing, the Intel chips were usually better. But, if they are a good source for benchmark results, then they should allow you to weight features based on few questions, provide different comparison results based on use categories, and not assume that everyone is just gaming. And this type of issue could easily be attached to any component to skew the results for one side or another. This is the major reason why one test cannot evaluate a component or system. Thanks for sharing!
Weird looking lancia
thanks clickyphil
We have a Schrodinger's CPU - it is both good and bad, and the box is the price of the laptops.
I love listening to you shit on Uselessbenchmark
I love watching your videos on hardware, we are both growing in a similiar pace. Ive been getting more and more into it, just like you :>
There is actually a similar site, i forgot how its called but you enter the cpu and gpu and it spits out fps for a specific game. It relies on user data but i crossreferenced it with my system and it seemed pretty acurate. Also tells you bottlenecks and stuff
Ironic ...
Some guy in the comments just quoted himself haha
how would you edit your videos without a slow pan?
Any comments about CinebenchR23?