The company that made the 30+ year old thermal paste replied to my email and quote: "We wish we could be of more assistance but we can not tell anything from picture. I will say it is very old as we do not use the green and white logos anymore." The mystery continues! Will continue asking around to see if we can pinpoint the exact manufacturing date!
If you can ask the 30year old thermal paste company when they switched over to their current container design then you will have a rough time period idea for the exact year like 1995-2000 so ask the company about how many times they have updated their product packaging in their company life cycle then you and the company could find a close enough Year estimate guess and pin the details in your comment.
One thing you didn't account for / didn't bring up is changes in formulation over the years. Even tubes manufactured in the same year may not have the exactly same formulation depending on where it was manufactured. Factures in different places will source similar / compatible raw material that are easier to get in their region.
Yea over the years advancements in formula's and increase in power draws of cpu's would likely cause them to reformulate the paste to have better thermal transfer properties.
Still as it stands there is hardly anything to worry about unless the thermal paste has turned to water, but yeah i dont doubt the forumals are differeing but most (quality) thermal pastes hit a wall many moons ago
As a former OEM assembler, I can tell you that we would often save TIM from the retail boxed coolers and use what we had on our desk. So you could have TIM from a cooler master cooler on a BeQuiet cooler for example. Typically, anything we used was name brand though.We would use the stock TIM on AMD and Intel coolers though.
I'm still using the same tube of UK-made RS Components white 'Heatsink compound' I've had since the 80's. It works great ! I only ever use the smallest amount needed to get a very thin film.
What your test determined was that your pc will throttle as much as it needs to do to keep the CPU from exceeding 90 degrees. We don't know how many degrees correspond with throttling points, so maybe put in that center fan that is missing so your cooling works better and we can see some real-world temperature differences between the thermal compounds. When you do this, you'll only know what temps you get from freshly-applied compound, which has never spent years degrading under heat stress. Degradation over time is what we really need to know. Also, compound formulation may have changed over time so newer compound might work better only because it has been improved. I find that thermal compound dries out very quickly after application, so being dried out isn't a big problem for performance, it's more a problem with application of the compound. That's the only reason I can think of that "old" compound would need to expire. Compound that is applied to a working CPU should work forever unless it has a tendancy to develop insulating bubbles, or the stress of use degrades it chemically so much that it really stops conducting heat as well.
The real test is how this old paste holds up in use. I've used 10+ old unbranded thermal paste in a server and wasn't able to get more than a year of 24/7 duty before it tripped the temp alarm and spins up the fans to max.
@@jimtaylor201 dunno, stuff you say is quite relative, for example if you put alarm low, and it might be 80 to 81°c difference, which is nothing. i mean, spinning fans to max on occasional redlining is nothing out of ordinary. do you aim not to spin fans? i have some old machine that one can hear when it's at 100% cpu, it spins, so what? i never renew paste in anything i've built. no issues. did a lot of video encoding.
@@ivok9846 the motherboard is set to thermal defaults and will spin up at full blast when it reaches those HP settings. No other CPUs have that issue in that server or other identical servers in the same rack.
Searched the internet so long trying to find a comprehensive answer to this question. You have saved my ever growing forgotten thermal paste collection from obsolescence. Thank you sir!
I've been buying new thermal paste every day for the past 32 years. Glad to know that hasn't gone to waste completely! My house is 90% thermal paste tubes
I seem to remember reading somewhere a few years ago Arctic changing up their MX-4 formula a lil bit to make it more thermal efficient. That could be why they look different tbh.
how so? i have some white silicon paste made exactly for transistor coolers that i was putting on CPUs and it's not hard at all. all "cpu specific" thermal pastes i ever saw are drier than that one, and from a get go.
Nice! I had a 15 year old tube of Arctic Silver 5 sitting around. I used it on a spare PC, worked just fine. Finally threw it away because was like 10% left.
The "active ingredient" in thermal paste is the solid filler. The oil is only there to facilitate application and help particles stay put. The "expiration" has more to do with optimal ease and repeatability of application than performance. Once applied, powder-based pastes can be 50+ years old and still perform perfectly fine. Most of the times where people run into thermal performance issues with a 5+ years old paste application is because they nudged their HSF while cleaning their PC and broke the paste after it had fully settled down. I'm still using a tube of 20+ years old GCChemicals ZnO thermal paste in my PCs today, got about 1/5th of the tube left. The tube has lost quite a bit of oil through the aluminum squeeze tube having ripped in a few folds.
Interesting comparison. The thermal throttling behavior of the CPU might have skewed the results a bit though. If the CPU throttles at 90°, it could mean that the temperature difference between new and old thermal paste would be much higher without the CPU throttling, as with throttling the temperature is kept at 90° max. irrespective of the thermal conductivity of the paste used.
That MX-4 seems to have changed recipe cause I don't remember having a brand new one looking like that newer one. All the MX-4 I had was more like the old ones you have even when brand new if I remember well...
@@chlonez It doesn't matter cause mine has many years and looks like his older ones and since day one it looked very similar so it seems different to his newer one...
we got a new tube of MX4 a while back and it looked just like the older mx4. Ours came in a different packaging so I'm also thinking they changed the formula. I did a little googling and found a 2022 reddit post stating arctic changed the formula, as the new MSDS for MX4 shows a totally different specific gravity.
I have Arctic Silver tubes in variety in year gotta say the larger tubes regardless of year are hard to push. I recommend storing the tubes themselves inside a sealed container tubes stored loose eventually dry out compared to the ones in a semi tight container
Thermal paste is basically made up of two components: liquid (mostly silicon oils) and solid (mostly metal-oxides) . Role of the paste is to get in various cavities, nooks and crannies between two metal surfaces. Thermal paste is NOT better than pure metal-to-metal contact, but is better thermal conductor than air. Liquid part of the paste is mostly to give viscosity and get it into aforementioned cavities, solid part (metal-oxides) actually conducts heat. Once you get paste in place, liquid part is no longer essential, and it could evaporate. Solid parts remain stable for years (centuries) and will continue to function as long as you do not move your cooler and displace them from cavities.
@@HanmaHeiro It does not unless you remove the cooler. Usually fan on the cooler fails (mechanical part, wear and tear ). So you remove the cooler and then you have to reapply the paste, of course. Interestingly enough, for cheaper processors with low TDP (laptop Celerons for example) you do not need to re-apply paste because they are passively cooled :)
I have a large tube of Arctic Silver 3 from around 2000, still has good consistency and works fine. I also have a smaller tube of Arctic Silver 5 I purchased a few years later (as soon as it came out) which I applied last year to the CPU I'm using to write this comment - it also works great and I have noticed no difference except for a very slight increase in viscosity.
I mean logically it's more sealed in the tube than on a cpu. It also makes sense for it not to break down much over time as it would otherwise do that while in a computer as well and that would lead to thermal problems. So that it has a decent shelf life and tends to not massively degrade while in use are a key point, especially considering most people don't want to apply new paste once a year either.
06:40 That is not thermal throttling. That is reaching a different boost speed. If you were using different processors to test each paste, even if they are the same model, your tests are completely invalid. You have to use the same exact processor, on the same exact board, and do at least three separate mounts for each paste.
Has he done a test where hes left the thermal paste in the pc for different amounts of time? for example: One thermal paste applied 10 years ago and not changed, one applied 5 years ago and not changed and compared to only just applied. All being used as regular pc's during the time periods of course
Usually in lower power CPUs (65-95W), I didn't saw any meaningful temperature increase with older thernal paste, often staying below throttling threshold. As a rule if the thumb, if you are not throttling and running at reasonable temperatures, the system is fine and you can delay a bit the purchase of that brand new thermal paste.
I found a tube of thermal paste in a motherboard box that supported Intel Pentiums that still had pins on their CPUs (before their LGA standard) at the start of 2020. It handled my 4790 for about 4 years back on my old Z97 build before I finally upgraded. Probably not the best of ideas but.. it worked! :D
It dries out, I still have an ancient tube of Arctic silver II, yes 2, I don't use it now but I keep it around for memories and the paste that comes out of it is like super dry and starting to become powder like.
I used some 12 year old MX-4 a few weeks ago. I seriously bought it when I built my Sandy bridge. It worked as well as it ever did I figure. I did keep it in the pencil cup stood on end along with old vapes. I think laying it on it's side would not be a good thing. It would separate.
I *finally* ran out a syringe of Arctic Silver late last year (it's not marked which # it is). It's older than 2005 since I had it when I moved at the time. I didn't use it on any new modern processors, just older stuff I was upgrading. No particular issues. Not to be used is some old copper paste from the pre-IHS days when they made copper shims for Pentium 3s. That was the paste for them.
I would assume that the tubes will last much longer than a tub as the tubes look like they have a good seal on them with the small tight endcap while that tub has a poor seal with a large lid and will be more susceptible to drying out.
I thought the 90c ceiling was indicative of the pastes working until I saw that 30 year old paste "work", and it suddenly became clear that something is off. When that paste was new, 50c was considered overheating for CPUs. There's no oil dry up in that paste, there was never any oil at all, because that wasn't necessary for cooling a 30 year old Pentium. I don't think the CPU going to 90c is a sign of the paste working correctly, because there is no way it should be working effectively at all. I think you should test again with a fan.
I just threw out a 14 year old tube of MX4 that I got back when I was doing hardware reviews. I just got a brand new tube of mx4 to replace it with. But I used the old stuff up until last week and it worked just fine.
I'm surprised how well that 30 year old thermal paste performed. Don't forget 30 years ago, processors generally used less than 1/2 the power and ran at 1/3 the clock speed. Also, generic thermal paste back then also performed significantly worse than the stuff available today.
Before you said the answer here's my guess: They say that thermal paste expires but i think that newer thermal paste is better at dissipating heat than older thermal paste, because technology evolves.
Thermal paste is based on silicone. If that expires, then you should have the exact same worry about the very chip (CPU, GPU etc.) you're applying it on, since chip wafers are also made out of silicone. But since you never worry about that piece of silicone "losing its properties", why would you worry about the other? Not to mention the silicone from the chips has waaaayyy higher requirements for purity and charge separation via its depletion layers, so next time you see a Pentium I from 1996 with the same thermal paste since 30 years ago, don't worry, it still runs fine.
I wonder if you emptied out one of the thermal pastes out of the tube and 're-mixed' it if it would preform any better? Also: Is there manufactured improvements between old and new 🤔
that one made me fun to watch and was really answering 20 years old questions...my conclusion - not really big differences...only 1 C and few Mhz....nothing to worry about and this also could be measurment differences and application... anyhow - good to know...amazing also how good AC Silver still performce
I think that it is important to realize that thermal pastes are getting better over time so it is sure that older tgermal pastes are worse than new ones but they don't have to be expired. Good video.
How does this test relevant as these pastes are basically brand new. For this test to be relevant the paste has to be using for the specified time like 5 years or so.
Does the 5600x3d really run that hot during a stress test? I mean, that's a pretty decent air cooler, and hitting 87 degrees, isn't that a bit too much?
a few weeks ago i opened on well over 15 years old notebook (hp compaq i think) to clean the heatsink and replace the termalpaste, but the termalpaste was like new and the computer has not ben used or opened for over 7 years, it was exactly the same as in my gaming pc i build 2 years ago, besides the color (white, gray). I replaced it anyway couse i already opened it, but realy impressing that termalpaste works fos so long, nice vidio btw
Whatever you do with old thermal pastes. Around 200-2005 or so there where some Pastes on the market that where conductive. What happens if you take the old separated stuff an give it a good mix? Squeeze the whole thing in a little jar and mix it, test it again.
Also curious, would using the 30-year-old paste in a low heat CPU be fine? If so, how long would it last compared to the life span of using new thermal paste?
this actually reminded me of something stupid I did back when i first got into pc building in around 2015-16, I found a tube of arctic silver 3 in my basement that I assume my dad had used when building his old Athlon XP system in the early 2000s and was just like yeah it'll be fine to use this on my system, thankfully I was running a relatively low wattage Athlon X2 (all I could afford) dunno how the temps were but it survived lol, I remember that I took it apart when I upgraded that it was all separated there was streaks of just oil and spots that looked like just tiny metal filings it was scuffed
great video! this experiment shows a lot of important stuff when it comes to thermal laste aging. I was most surprised how the two brands had opposite consistencies from newest to oldest
Have wondered myself how they hold up. I have some tube of some thermal paste from 2014, visually looks alright. Though I didn't dare to use it and bought some new.
Mabey you should mix different thermal pastes and then test them. 😀 When you mentioned how the first brand of paste was almost the opposite in how it aged versus the second paste got me to think "what if you mix the two oldest pastes from the MX and Arctic Silver?"
Extemly interesting love your effort ,recetly used mx4 paste 3 years old had no new paste temps wore hi idle 50 ,gaming 90, new paste a few days later ,idle 35,gaming 77 so in me case old past sucks
i haven't watched the video yet, my prediction is yes. I will edit this comment when I finish the video. I was kind of right considering the temperatures were extremely high. Great Video BTW.
I mean yeah kind of but the temps were not “extremely high” since even the thermal pastes that were brand new were near the “extremely high” so are you saying brand new thermal paste is not worth it?
how exactly were you right? how exactly paste expires when even the oldest paste, not spread at all, has just one degree worse result than new paste? how do you judge things? if something is "expired", shouldn't it be "unusable"? and can you imagine wouid this 30 years old compound affect something mostly idling, ie recent machine not encoding video or gaming non -stop?
@@ivok9846 honestly no idea what he was going on about the only thong that was probably expired seemed like the 30+ year old compound everything else was completely fine
my 5 year old mx-5 thermal paste has still held it's form and there is no heat bottleneck, maybe environment conditions also play a part in it as I have just used mine to refresh two of my old gpu's and it still comes out as if it was brand new. given i am almost out, i should buy another tube of the stuff, but it performs quite well atm...
Did you really have to run the tests at the thermal throttling limit? This makes it harder to really see whatever difference there is. Your cooler looks substantial but are you even undervolting that X3D chip?
It's a mistake to assume the formulation is consistent across all samples. Any differences could be the result of changes to that rather than aging. I've never seen any evidence that thermal paste actually expires.
Tbh it would have been much better if you had a good CPU and cooler for this that hit like 80C instead. I just don't like comparing clocks instead of temperature.
1-3° Seems within margin of error. Too bad this CPU basically hit thermal limits regardless which paste is used, maybe fixing clockrates would yield clearer results. Anyway i take old but decent paste over cheap paste any day.
Guys , ive got a question , i got a brand new xbox series x back in december 2020 , i have never used it , its been sitting in my basement for the past 3.5 years , should i replace the thermal paste if i want to use it again ?
U should have went down with PPT, TDC and EDC, so the CPU isnt that close to throttle temp. In my opinion that would have made a comparsion easier/better. Other than that a very cool video. Just bought new paste and threw my 8 year old away :(
The company that made the 30+ year old thermal paste replied to my email and quote:
"We wish we could be of more assistance but we can not tell anything from picture. I will say it is very old as we do not use the green and white logos anymore."
The mystery continues! Will continue asking around to see if we can pinpoint the exact manufacturing date!
If you can ask the 30year old thermal paste company when they switched over to their current container design then you will have a rough time period idea for the exact year like 1995-2000 so ask the company about how many times they have updated their product packaging in their company life cycle then you and the company could find a close enough Year estimate guess and pin the details in your comment.
could they at least provide the date they stopped using that logo to help narrow the search window down?
damn
that Thermal Paste came out in 1994
5th reply
One thing you didn't account for / didn't bring up is changes in formulation over the years.
Even tubes manufactured in the same year may not have the exactly same formulation depending on where it was manufactured. Factures in different places will source similar / compatible raw material that are easier to get in their region.
Yea over the years advancements in formula's and increase in power draws of cpu's would likely cause them to reformulate the paste to have better thermal transfer properties.
Still as it stands there is hardly anything to worry about unless the thermal paste has turned to water, but yeah i dont doubt the forumals are differeing but most (quality) thermal pastes hit a wall many moons ago
100000%
The saddest part of this video was that 2019 was 5 years ago
Wow, time really goes by fast. :(
Make that 8 years if you count Covid.
@@paulperry7091oh сука how is that POSSIBLE AAAAA TIME
literally yesterday i was literally on that flight to turkey? apparently 2019
@@BroironxAviation which flight ?
That 30 year old thermal compound is probably still better than what most computer OEMs use
and the Xbox thermal paste especially
@@aerospherology2001Xbox uses toothpaste
As a former OEM assembler, I can tell you that we would often save TIM from the retail boxed coolers and use what we had on our desk. So you could have TIM from a cooler master cooler on a BeQuiet cooler for example. Typically, anything we used was name brand though.We would use the stock TIM on AMD and Intel coolers though.
@@AlamedaFumbles meant were companies like dell , hp and more , those big ones
I'm still using the same tube of UK-made RS Components white 'Heatsink compound' I've had since the 80's. It works great ! I only ever use the smallest amount needed to get a very thin film.
What your test determined was that your pc will throttle as much as it needs to do to keep the CPU from exceeding 90 degrees. We don't know how many degrees correspond with throttling points, so maybe put in that center fan that is missing so your cooling works better and we can see some real-world temperature differences between the thermal compounds.
When you do this, you'll only know what temps you get from freshly-applied compound, which has never spent years degrading under heat stress. Degradation over time is what we really need to know.
Also, compound formulation may have changed over time so newer compound might work better only because it has been improved.
I find that thermal compound dries out very quickly after application, so being dried out isn't a big problem for performance, it's more a problem with application of the compound. That's the only reason I can think of that "old" compound would need to expire. Compound that is applied to a working CPU should work forever unless it has a tendancy to develop insulating bubbles, or the stress of use degrades it chemically so much that it really stops conducting heat as well.
The real test is how this old paste holds up in use. I've used 10+ old unbranded thermal paste in a server and wasn't able to get more than a year of 24/7 duty before it tripped the temp alarm and spins up the fans to max.
are you sure it wasnt the dust build up and not the thermal paste?
24/7 100% cpu load?
@@ivok9846 not running 100% all the time.
@@jimtaylor201 dunno, stuff you say is quite relative, for example if you put alarm low, and it might be 80 to 81°c difference, which is nothing.
i mean, spinning fans to max on occasional redlining is nothing out of ordinary.
do you aim not to spin fans?
i have some old machine that one can hear when it's at 100% cpu, it spins, so what?
i never renew paste in anything i've built. no issues.
did a lot of video encoding.
@@ivok9846 the motherboard is set to thermal defaults and will spin up at full blast when it reaches those HP settings. No other CPUs have that issue in that server or other identical servers in the same rack.
Searched the internet so long trying to find a comprehensive answer to this question. You have saved my ever growing forgotten thermal paste collection from obsolescence. Thank you sir!
Okay, but how do they differ in taste??!
🤣
im speechless....
no ur not you said something@@kadenonyt3
Tf💀😨😱
the later half will give you tongue cancer
I've been buying new thermal paste every day for the past 32 years. Glad to know that hasn't gone to waste completely! My house is 90% thermal paste tubes
What is bro doing with that much thermal paste
@@clyded_thysilly Use your creative imagination :P
that’s more than 90k for tubes 💀
the mx4's formula has changed over the years
Marketing myth.
@@kevinerbs2778 Ya for real
There could be differences in "recipes"
I seem to remember reading somewhere a few years ago Arctic changing up their MX-4 formula a lil bit to make it more thermal efficient. That could be why they look different tbh.
thermal compound that they used to put on transistors in amplifiers used to be drier than CPU thermal pastes
like why? its so bad over time
how so? i have some white silicon paste made exactly for transistor coolers that i was putting on CPUs and it's not hard at all. all "cpu specific" thermal pastes i ever saw are drier than that one, and from a get go.
Nice! I had a 15 year old tube of Arctic Silver 5 sitting around. I used it on a spare PC, worked just fine. Finally threw it away because was like 10% left.
Next vid idea: how does old thermal paste compare to the best thermal paste replacement?
The "active ingredient" in thermal paste is the solid filler. The oil is only there to facilitate application and help particles stay put. The "expiration" has more to do with optimal ease and repeatability of application than performance. Once applied, powder-based pastes can be 50+ years old and still perform perfectly fine. Most of the times where people run into thermal performance issues with a 5+ years old paste application is because they nudged their HSF while cleaning their PC and broke the paste after it had fully settled down.
I'm still using a tube of 20+ years old GCChemicals ZnO thermal paste in my PCs today, got about 1/5th of the tube left. The tube has lost quite a bit of oil through the aluminum squeeze tube having ripped in a few folds.
That's not an aging difference that's just the generation difference
Interesting comparison. The thermal throttling behavior of the CPU might have skewed the results a bit though. If the CPU throttles at 90°, it could mean that the temperature difference between new and old thermal paste would be much higher without the CPU throttling, as with throttling the temperature is kept at 90° max. irrespective of the thermal conductivity of the paste used.
That MX-4 seems to have changed recipe cause I don't remember having a brand new one looking like that newer one. All the MX-4 I had was more like the old ones you have even when brand new if I remember well...
what if your "new" paste already on store shelf for a year
@@chlonez It doesn't matter cause mine has many years and looks like his older ones and since day one it looked very similar so it seems different to his newer one...
we got a new tube of MX4 a while back and it looked just like the older mx4. Ours came in a different packaging so I'm also thinking they changed the formula. I did a little googling and found a 2022 reddit post stating arctic changed the formula, as the new MSDS for MX4 shows a totally different specific gravity.
@@kasuraga I'm still using MX4 from 2014, if they did change it, Im guessing its for the better, after all, technology improves over time
I have Arctic Silver tubes in variety in year gotta say the larger tubes regardless of year are hard to push. I recommend storing the tubes themselves inside a sealed container tubes stored loose eventually dry out compared to the ones in a semi tight container
Thermal paste is basically made up of two components: liquid (mostly silicon oils) and solid (mostly metal-oxides) . Role of the paste is to get in various cavities, nooks and crannies between two metal surfaces. Thermal paste is NOT better than pure metal-to-metal contact, but is better thermal conductor than air. Liquid part of the paste is mostly to give viscosity and get it into aforementioned cavities, solid part (metal-oxides) actually conducts heat. Once you get paste in place, liquid part is no longer essential, and it could evaporate. Solid parts remain stable for years (centuries) and will continue to function as long as you do not move your cooler and displace them from cavities.
So you're saying thermal paste applied on a regularly used computer over 5 years does not need to be changed?
@@HanmaHeiro It does not unless you remove the cooler. Usually fan on the cooler fails (mechanical part, wear and tear ). So you remove the cooler and then you have to reapply the paste, of course. Interestingly enough, for cheaper processors with low TDP (laptop Celerons for example) you do not need to re-apply paste because they are passively cooled :)
I have a large tube of Arctic Silver 3 from around 2000, still has good consistency and works fine. I also have a smaller tube of Arctic Silver 5 I purchased a few years later (as soon as it came out) which I applied last year to the CPU I'm using to write this comment - it also works great and I have noticed no difference except for a very slight increase in viscosity.
I bought yeesterpaste and it's great fr! props to mryeester for making something unique and high quality!
"There is a pc in my thermal paste"
I mean logically it's more sealed in the tube than on a cpu. It also makes sense for it not to break down much over time as it would otherwise do that while in a computer as well and that would lead to thermal problems. So that it has a decent shelf life and tends to not massively degrade while in use are a key point, especially considering most people don't want to apply new paste once a year either.
06:40 That is not thermal throttling. That is reaching a different boost speed. If you were using different processors to test each paste, even if they are the same model, your tests are completely invalid.
You have to use the same exact processor, on the same exact board, and do at least three separate mounts for each paste.
Has he done a test where hes left the thermal paste in the pc for different amounts of time? for example: One thermal paste applied 10 years ago and not changed, one applied 5 years ago and not changed and compared to only just applied. All being used as regular pc's during the time periods of course
Thanks for the the test!
Usually in lower power CPUs (65-95W), I didn't saw any meaningful temperature increase with older thernal paste, often staying below throttling threshold. As a rule if the thumb, if you are not throttling and running at reasonable temperatures, the system is fine and you can delay a bit the purchase of that brand new thermal paste.
Great vid but a shame you didnt use a CPU/Cooler combo that doesnt max out so easily.
ikr? really leaning on the the thermal throttling when the brand new paste is within licking distance of the thermal ceiling anyway
I found a tube of thermal paste in a motherboard box that supported Intel Pentiums that still had pins on their CPUs (before their LGA standard) at the start of 2020. It handled my 4790 for about 4 years back on my old Z97 build before I finally upgraded. Probably not the best of ideas but.. it worked! :D
It dries out, I still have an ancient tube of Arctic silver II, yes 2, I don't use it now but I keep it around for memories and the paste that comes out of it is like super dry and starting to become powder like.
I'd say in a tube like that yes. The oils seperate. If you get a little bottle of it. Just premix it again it should be ok to use
I used some 12 year old MX-4 a few weeks ago. I seriously bought it when I built my Sandy bridge. It worked as well as it ever did I figure. I did keep it in the pencil cup stood on end along with old vapes. I think laying it on it's side would not be a good thing. It would separate.
Does removing a cooler from a CPU give thermal paste separation anxiety? 😛
I *finally* ran out a syringe of Arctic Silver late last year (it's not marked which # it is). It's older than 2005 since I had it when I moved at the time. I didn't use it on any new modern processors, just older stuff I was upgrading. No particular issues. Not to be used is some old copper paste from the pre-IHS days when they made copper shims for Pentium 3s. That was the paste for them.
I would assume that the tubes will last much longer than a tub as the tubes look like they have a good seal on them with the small tight endcap while that tub has a poor seal with a large lid and will be more susceptible to drying out.
Have you tried car radiator fluid as computer watercooling fluid? Wondering how good it'd be.
I thought the 90c ceiling was indicative of the pastes working until I saw that 30 year old paste "work", and it suddenly became clear that something is off.
When that paste was new, 50c was considered overheating for CPUs. There's no oil dry up in that paste, there was never any oil at all, because that wasn't necessary for cooling a 30 year old Pentium. I don't think the CPU going to 90c is a sign of the paste working correctly, because there is no way it should be working effectively at all.
I think you should test again with a fan.
Companies usually tweak ingredients over the years as well and why they might look differently.
I just threw out a 14 year old tube of MX4 that I got back when I was doing hardware reviews. I just got a brand new tube of mx4 to replace it with. But I used the old stuff up until last week and it worked just fine.
Thermal paste is just to make sure that the CPU surface contacts well with the heatsink, so the quality and age isn't that critical
I'm surprised how well that 30 year old thermal paste performed.
Don't forget 30 years ago, processors generally used less than 1/2 the power and ran at 1/3 the clock speed.
Also, generic thermal paste back then also performed significantly worse than the stuff available today.
Before you said the answer here's my guess: They say that thermal paste expires but i think that newer thermal paste is better at dissipating heat than older thermal paste, because technology evolves.
Thermal paste is based on silicone. If that expires, then you should have the exact same worry about the very chip (CPU, GPU etc.) you're applying it on, since chip wafers are also made out of silicone. But since you never worry about that piece of silicone "losing its properties", why would you worry about the other? Not to mention the silicone from the chips has waaaayyy higher requirements for purity and charge separation via its depletion layers, so next time you see a Pentium I from 1996 with the same thermal paste since 30 years ago, don't worry, it still runs fine.
I wonder if you emptied out one of the thermal pastes out of the tube and 're-mixed' it if it would preform any better?
Also: Is there manufactured improvements between old and new
🤔
that one made me fun to watch and was really answering 20 years old questions...my conclusion - not really big differences...only 1 C and few Mhz....nothing to worry about and this also could be measurment differences and application...
anyhow - good to know...amazing also how good AC Silver still performce
Yes it does because the components separate. How much and how fast depends on the mix and temperature stored at.
I think that it is important to realize that thermal pastes are getting better over time so it is sure that older tgermal pastes are worse than new ones but they don't have to be expired.
Good video.
exactly!
How does this test relevant as these pastes are basically brand new. For this test to be relevant the paste has to be using for the specified time like 5 years or so.
Like my Arctic Silver 5 I been using with my rig since 2013. 40c for the CPU on idle and around 60 on load
Does the 5600x3d really run that hot during a stress test? I mean, that's a pretty decent air cooler, and hitting 87 degrees, isn't that a bit too much?
Mine doesn't even go to 70 maybe hes living in a hell hole lol
I personaly applied a toothpaste to Athlon Barton, and it lasts almost 2 years to start overheating
does the fan being in the middle work better for any reason or just show off your RAM?
Thanks exactly what I needed to know 😊,👍
a few weeks ago i opened on well over 15 years old notebook (hp compaq i think) to clean the heatsink and replace the termalpaste, but the termalpaste was like new and the computer has not ben used or opened for over 7 years, it was exactly the same as in my gaming pc i build 2 years ago, besides the color (white, gray). I replaced it anyway couse i already opened it, but realy impressing that termalpaste works fos so long, nice vidio btw
have you considered using different cpus as well? difference in ihs flatness and the thermal compound between the ihs and the die can also effects it
I think the main difference might have been, that thermal paste technology has improved a little over the years.
Whatever you do with old thermal pastes. Around 200-2005 or so there where some Pastes on the market that where conductive.
What happens if you take the old separated stuff an give it a good mix? Squeeze the whole thing in a little jar and mix it, test it again.
I wonder if you could stir the old thermal paste and make it almost like new.
They upgraded these thermal pastes throughout the years that could be the cause of temp difference.
You should try heating up the paste a little bit in the oven maybe and shaking it. Maybe its good as new??
Also curious, would using the 30-year-old paste in a low heat CPU be fine? If so, how long would it last compared to the life span of using new thermal paste?
this actually reminded me of something stupid I did back when i first got into pc building in around 2015-16, I found a tube of arctic silver 3 in my basement that I assume my dad had used when building his old Athlon XP system in the early 2000s and was just like yeah it'll be fine to use this on my system, thankfully I was running a relatively low wattage Athlon X2 (all I could afford) dunno how the temps were but it survived lol, I remember that I took it apart when I upgraded that it was all separated there was streaks of just oil and spots that looked like just tiny metal filings it was scuffed
great video! this experiment shows a lot of important stuff when it comes to thermal laste aging. I was most surprised how the two brands had opposite consistencies from newest to oldest
Applying thermal paste on a CPU while watching this video, using the same Arctic one. Nice.
older paste degrades faster than new paste, even if they both perform well at first (at least in my experience.)
Have wondered myself how they hold up. I have some tube of some thermal paste from 2014, visually looks alright. Though I didn't dare to use it and bought some new.
Mabey you should mix different thermal pastes and then test them. 😀 When you mentioned how the first brand of paste was almost the opposite in how it aged versus the second paste got me to think "what if you mix the two oldest pastes from the MX and Arctic Silver?"
Good idea for a short
i use a tube of cvs diaper rash cream works great been on my %800x for a year and a half and is still giving me cool temps
Great test!!!
What software did you use for testing?
Extemly interesting love your effort ,recetly used mx4 paste 3 years old had no new paste temps wore hi idle 50 ,gaming 90, new paste a few days later ,idle 35,gaming 77 so in me case old past sucks
Missed put on what happens when you mix em all into a new hyper evolution of thermal paste... with diamond powder also ❤
i haven't watched the video yet, my prediction is yes. I will edit this comment when I finish the video.
I was kind of right considering the temperatures were extremely high. Great Video BTW.
Ok
Of course it still worked. Duh.
I mean yeah kind of but the temps were not “extremely high” since even the thermal pastes that were brand new were near the “extremely high” so are you saying brand new thermal paste is not worth it?
how exactly were you right? how exactly paste expires when even the oldest paste, not spread at all, has just one degree worse result than new paste?
how do you judge things?
if something is "expired", shouldn't it be "unusable"?
and can you imagine wouid this 30 years old compound affect something mostly idling, ie recent machine not encoding video or gaming non -stop?
@@ivok9846 honestly no idea what he was going on about the only thong that was probably expired seemed like the 30+ year old compound everything else was completely fine
one question im curious about what if pull out the contents and give it a good stir can it reblend like its brand new?
The Artic one is exactly how I remember using it and there is no difference in thickness
my 5 year old mx-5 thermal paste has still held it's form and there is no heat bottleneck, maybe environment conditions also play a part in it as I have just used mine to refresh two of my old gpu's and it still comes out as if it was brand new. given i am almost out, i should buy another tube of the stuff, but it performs quite well atm...
Did you really have to run the tests at the thermal throttling limit? This makes it harder to really see whatever difference there is. Your cooler looks substantial but are you even undervolting that X3D chip?
I swear this guy eats thermal paste on his toast
i think the new MX4 just looks different. i have some arctic thermal paste from 2009 which has always looked shiny, even when it was new.
It's a mistake to assume the formulation is consistent across all samples. Any differences could be the result of changes to that rather than aging.
I've never seen any evidence that thermal paste actually expires.
i can't remember the last time i replaced mine, 2012 perhaps? maybe a few years up
The very old white thermal paste likely contains beryllium oxide in silicon oil.
i have a problem recently i changed thermalpaste in my pc (I7-12700K) and it gets hotter than it was before i bought arctic MX-4
you need to buy mx-6...it's a better option
What happens if you mix different types of thermal paste?
Tbh it would have been much better if you had a good CPU and cooler for this that hit like 80C instead. I just don't like comparing clocks instead of temperature.
This is really cool and great video
Why do ypu even have thermal paste this old. Did you just forget to clean your inventory?
I know this is unrelated to the video but i was thinking.
When a poison expires is it still poisonous?
depends on the poison, some poisons will only give you stomach aches while others will make you expire faster than they did before expiring
Great video 👍
Finally i got a thermal paste for my pc
I have a old tube of arctic alumna works great
1-3° Seems within margin of error.
Too bad this CPU basically hit thermal limits regardless which paste is used, maybe fixing clockrates would yield clearer results.
Anyway i take old but decent paste over cheap paste any day.
Pls make a video where you use the thermal paste from the Vision Pro chips on a PC!!!
They DO expire because they dry out. Even if you manage to get them working at first, I don't think they will hold well long term.
i had some mx4 recently and was so hard out the tube wouldnt spread or anything,,, seemed like it went off to me
Guys , ive got a question , i got a brand new xbox series x back in december 2020 , i have never used it , its been sitting in my basement for the past 3.5 years , should i replace the thermal paste if i want to use it again ?
I change my paste out monthly for that sweet sweet 1fps gain in WarZone
U should have went down with PPT, TDC and EDC, so the CPU isnt that close to throttle temp. In my opinion that would have made a comparsion easier/better. Other than that a very cool video. Just bought new paste and threw my 8 year old away :(
new cpus run to hot run this with ryzen 3700x so you can see if there is a diff
id say there is a difference since most cpus 5900x -7000 series targets 90c you cant fully see a difference