|
Yeah, trying to determine the best value for thermal paste makes no sense to me. Just buy the best poo poo that's on the market. It's a negligible addition to your costs. edit: within the consumer-grade realm, i mean
Dr. Video Games 0031 fucked around with this message at 10:31 on Jun 15, 2021 |
#
?
Jun 15, 2021 10:15
|
|
|
|
| # ? May 15, 2024 19:09 |
|
|
OK, what's the best thermal paste?
|
|
#
?
Jun 15, 2021 10:25
|
|
|
since we're on the subject, has anyone tried using those graphite pads? as far as I know the thermal performance is strictly worse than thermal paste by a couple of degrees, but is supposed to be made up for by the endurance (as in they never have to be replaced) and convenience (as in if you ever want to change, there's no paste clean-up and reapplication involved)
|
|
#
?
Jun 15, 2021 10:28
|
|
|
LRADIKAL posted:OK, what's the best thermal paste? probably hydronaut
|
|
#
?
Jun 15, 2021 10:28
|
|
|
LRADIKAL posted:OK, what's the best thermal paste? so with that being all about the same price for the most part, might as well get any of them and are paying about the same. the actual #1 top spot to some tenth of a celsius doesn't matter. the important part is not to cheap out on some alibaba-tier poo poo over a few bucks. but as a go-to there's really no reason not to buy say NT-H2 or TF8 or Kryonaut as an easy simple decision. I mean i think we're all saying the same poo poo, just some semantics lost in translation. Xaris fucked around with this message at 10:41 on Jun 15, 2021 |
|
#
?
Jun 15, 2021 10:38
|
|
|
gradenko_2000 posted:since we're on the subject, has anyone tried using those graphite pads? as far as I know the thermal performance is strictly worse than thermal paste by a couple of degrees, but is supposed to be made up for by the endurance (as in they never have to be replaced) and convenience (as in if you ever want to change, there's no paste clean-up and reapplication involved) https://www.youtube.com/watch?v=niAQs8dZohE
|
|
#
?
Jun 15, 2021 10:42
|
|
|
LRADIKAL posted:OK, what's the best thermal paste? Thermal grizzly kryonaut extreme. At 14.2 W/mK it's over fifty percent more conductive than most other high performance pastes on the market (8-9 W/mK). You can actually see 2-6c improvements over other stuff. If you want to go crazy there is always liquid metal too, lol.
|
|
#
?
Jun 15, 2021 15:50
|
|
|
I think the main difference is the viscosity or thickness of the paste. Noctua's is very easy to spread, but people don't have as much ease of use with certain other prominent pastes.
|
|
#
?
Jun 15, 2021 15:59
|
|
|
I think I'm going to jump out of the window. The be quiet dark rock slim arrived. It needs the am4 stock backplate. I threw out the stock am4 backplate during spring cleaning. Any old am4 backplate from Amazon should work right?
|
|
#
?
Jun 15, 2021 18:56
|
|
|
I'm considering ASRock ROMED8-2T for my next virtual machine host / NAS / whatever hobby box and I'm stumped on what cooler to get. That board has the socket in different orientation than "mainstream" boards (TR, TRX etc) so most coolers (Noctua, Coolermaster) would be pointing upwards which is something I don't want. What are my options re: coolers?
|
|
#
?
Jun 15, 2021 19:04
|
|
|
gradenko_2000 posted:since we're on the subject, has anyone tried using those graphite pads? as far as I know the thermal performance is strictly worse than thermal paste by a couple of degrees, but is supposed to be made up for by the endurance (as in they never have to be replaced) and convenience (as in if you ever want to change, there's no paste clean-up and reapplication involved) I used one on a computer I built for someone else that is pretty much going to stay as is for it's entire life. Worked fine, no mess and no bothering to wonder if I actually applied a correct amount. A+ would use it again but I also rarely if ever upgrade my CPU without a new mobo anyway
|
|
#
?
Jun 15, 2021 19:29
|
|
|
gradenko_2000 posted:since we're on the subject, has anyone tried using those graphite pads? as far as I know the thermal performance is strictly worse than thermal paste by a couple of degrees, but is supposed to be made up for by the endurance (as in they never have to be replaced) and convenience (as in if you ever want to change, there's no paste clean-up and reapplication involved) GN has done testing on them. They're not good. Sure, they will work on a low-power system - but you're spending more money for worse performance, in order to gain what? More convenient repeated removal and reinstallation of HSF on a low-power system where you have no reason to be doing that? I can't see a reason to ever use them unless it's for some really bizarre application where environmental factors would cause thermal paste to age extremely prematurely.
|
|
#
?
Jun 15, 2021 19:53
|
|
|
I've got one, and there a good number of situations where I think it works well. I've currently got it in my HTPC, which is a 65w TDP part with no overclocking, and its great in that use. I've also trimmed them down and used them in a few laptop applications. If you do a lot of parts testing in a shop or you are sorting through like a box of CPUs off ebay or somethin, they are probably a godsend. For a gaming desktop though, I wouldnt use one. Good paste is cheaper and much better performing.
|
|
#
?
Jun 15, 2021 20:16
|
|
|
K8.0 posted:GN has done testing on them. They're not good. Sure, they will work on a low-power system - but you're spending more money for worse performance, in order to gain what? More convenient repeated removal and reinstallation of HSF on a low-power system where you have no reason to be doing that? I can't see a reason to ever use them unless it's for some really bizarre application where environmental factors would cause thermal paste to age extremely prematurely.
|
|
#
?
Jun 15, 2021 21:05
|
|
|
I think my tube of arctic silver II can legally drink now.
|
|
#
?
Jun 15, 2021 22:01
|
|
|
Lord Stimperor posted:I think I'm going to jump out of the window. The be quiet dark rock slim arrived. I don't see why not. I still had my original backplate so it wasn't an issue for me, but it should be a stock part.
|
|
#
?
Jun 16, 2021 00:16
|
|
|
K8.0 posted:GN has done testing on them. They're not good. Sure, they will work on a low-power system - but you're spending more money for worse performance, in order to gain what? More convenient repeated removal and reinstallation of HSF on a low-power system where you have no reason to be doing that? I can't see a reason to ever use them unless it's for some really bizarre application where environmental factors would cause thermal paste to age extremely prematurely. You're misrepresenting GN's verdict. They decided that thermal pads were just fine, actually. There's no problem with using them unless you're a hardcore overclocker. The flip side to this is that there's no obvious reason to use them for your home builds, either. You may as well just use the paste your cooler came with, because it's not like the pad will be any better. The main reason to use pads is for things you expect to have a long life and wish to service as little as possible. Laptops, grandma's computer, etc. Or in professional environments where you're swapping out components regularly and having perfect thermals doesn't matter.
|
|
#
?
Jun 16, 2021 00:24
|
|
|
Good thermal paste lasts drat near forever. Comparing it to the specialist OC record chasing products that don't have long lifespans is silly, since they aren't what you'd use in those applications. I'm not misinterpreting the data, it's just bad unless you have extremely abnormal conditions (i.e. environmental or a CPU you expect to go untouched for 20+ years for some reason) or you specifically need easy frequent re-application without cleanup and don't mind paying much more for slightly less performance.
|
|
#
?
Jun 16, 2021 00:40
|
|
|
I don't know, I notice I have to repaste every like 2-5 years and I have given serious thought to pads because I actually am that lazy. They are conductive though, so make sure they don't hang over the side or something.
|
|
#
?
Jun 16, 2021 02:02
|
|
|
laptops are definitely a case where paste will age abnormally fast. Once you "cook" it once by letting the laptop get super hot sitting on a blanket in bed or something, the paste dries up and it will regularly overheat with much less provocation. thermal pads actually sound extremely good in that situation
|
|
#
?
Jun 16, 2021 02:55
|
|
|
The best use for thermal pads are situations where you're not going to be opening the thing up ever, like a laptop or industrial PC. If you're sensitive enough to temps that you're repasting more than once a decade you probably won't be happy with their baseline performance.
|
|
#
?
Jun 16, 2021 03:15
|
|
|
I'm using the stock 3600x thermal pad because I'm lazy and somehow I'm having lots of fun with this thing and it hasn't blown up yet
|
|
#
?
Jun 16, 2021 07:18
|
|
|
B-1.1.7 Bomber posted:I'm using the stock 3600x thermal pad because I'm lazy and somehow I'm having lots of fun with this thing and it hasn't blown up yet I was doing that. Then, when I decided to upgrade my cooler, I discovered that pre-installed thermal paste had turned into a curiously dusty kind of clay. I've never seen thermal paste degrade that badly before.
|
|
#
?
Jun 16, 2021 07:39
|
|
|
Status update from hot crash hell! Wanna give a quick update because I'm honestly super happy how much everyone is helping me with my dumb computer. You're really going deep with your responses and I sincerely appreciate that. TL;DR: Temperatures fine, crashes still bad, testing memory now. Temperatures fine I threw out the borked M22 AIO. I had originally ordered a be quiet dark rock slim to replace it. However, that needed the stock backplate which I no longer had and which wouldn't arrive before Saturday. So I very carefully put it back in the box, replacing the opened plastic baggies with new ones so it can be sold to a new customer. I then installed a Cooler Master Hyper 2012. This has pretty much resolved the temperature issues. The BIOS says that my CPU idles at 38C. That's not a bad value considered outside temps peaked at 31C today and I'm sitting here shirt- and pantless. After boooting, argues reports 55C which drop down to 40C really quickly. Temps were never better. The only mild regret is that I ordered Kryonaut before learning that it'd need a repaste after a year or so, but that's not an issue now. Stability bad, testing memory now, suggestions welcome However, system stability is still poor. - Bluescreen with random error in 5-20 minute intervals - Activity doesn't matter, curiously Chrome tabs keep crashing as well - Resetting BIOS as per to suggestion did not help yet - Now A/B testing different memory configurations to see whether it's a specific stick or slot that causes issues - Suggestions welcome! I know that on the software side, memtest and that some prime95 modes can spot memory issues, but not really anything further
|
|
#
?
Jun 16, 2021 21:40
|
|
|
Are your memory timings set manually or using XMP?
|
|
#
?
Jun 16, 2021 21:49
|
|
|
CaptainSarcastic posted:Are your memory timings set manually or using XMP? BIOS default, XMP should be off (will double check)
|
|
#
?
Jun 16, 2021 22:08
|
|
|
Lord Stimperor posted:Status update from hot crash hell! What does Windows Event Viewer say? Assuming XMP is off, if it's certain WHEA errors, it could be bad ram. Certainly trying only 1 stick at a time and making sure they're in the right slot(s) is a good step. You can actually try CoreCycler as a stability test for CPU. It'll do a test 1-by-1 by locking each core, only using 1, using SSE by default of light-loading them to max and cycling through them. You can also edit config.ini to increase it to AVX/AVX2 and different sizes, or have it run y-cruncher. https://github.com/sp00n/corecycler/releases TestMem5 with anita extreme profile should be very good at memory testing. https://www.overclock.net/threads/memory-testing-with-testmem5-tm5-with-custom-configs.1751608/ Also make sure to run "sfc /scannow" in Command Prompt (ran as admin) to check for any windows integrity issues I guess there's a possibility of a bad motherboard
|
|
#
?
Jun 16, 2021 23:03
|
|
|
I mean I knew my recent ECC DIMMs had it, but it's the first time since I have them that I look at sensors. Kinda fun, I suppose they're another source to observe chassis temperature changes:
|
|
#
?
Jun 17, 2021 16:35
|
|
|
Jim Keller has an interview on Anandtech, talking about a whole bunch of poo poo including his role wrt Zen. Tldr he was in management.
|
|
#
?
Jun 17, 2021 17:46
|
|
|
Lord Stimperor posted:Previously in Computer Hell Xaris posted:I presume the BIOS is updated to the latest version? Do you have the latest AMD Chipset drivers? BIOS is (now) updated to the most recent version. All chipset drivers were already the newest. In fact, the windows installation was about a day old when I made the first post, couldn't have been much cleaner. Windows Event viewer always says something along the lines of error 6008 -> error 41. To my knowledge that means something along the lines of "There was a crash, had to reboot". I've made a bit more progress but it hasn't been super elucidating. Creshes also occur with every possible combination of RAM sticks and slots. On the software side, I've used memtest pro, testmem, and prime 95 to stress memory and CPU, respectively. The computer can these for hours on end without problems. But when I interact with it, it's a matter of moments to minutes until it crashes. And with 'interact' it is really something as simple as writing in Notepad. You leave it, it runs. You touch it, it breaks. That doesn't rule out a hardware problem of course, but as a result I'm looking at Windows itself again. I'm now on a Ubuntu live disk and lo and behold, I can type messages, browse YouTube and dick around without anything happening. Windows itself was complaining already during the installation procedure, and crashed from the start. If Ubuntu works the rest of the evening, I believe I'll just wipe the C:\ drive and start over.
|
|
#
?
Jun 17, 2021 20:30
|
|
|
What are you using for your C: drive? Is it an SSD or is it an actual HDD?
|
|
#
?
Jun 17, 2021 21:52
|
|
|
Lord Stimperor posted:BIOS is (now) updated to the most recent version. All chipset drivers were already the newest. In fact, the windows installation was about a day old when I made the first post, couldn't have been much cleaner. I'd also consider the possibility of not just a corrupted install, but a drive failure. I don't remember, have you run any drive scanning software yet?
|
|
#
?
Jun 17, 2021 21:59
|
|
|
CaptainSarcastic posted:What are you using for your C: drive? Is it an SSD or is it an actual HDD? Dr. Video Games 0031 posted:I don't remember, have you run any drive scanning software yet? It's an SSD. The SMART status is okay. I've run a short smartctl test and that passed as well. I'll try if I can run the long one while my windows USB stick finishes being prepared. e: smartctl says that test will complete in 85 minutes so I'm probably not completing it tonight lol
|
|
#
?
Jun 17, 2021 22:59
|
|
|
Xaris posted:I presume the BIOS is updated to the latest version? Do you have the latest AMD Chipset drivers? AAAnnnd there is a possibility of a bad CPU.. it happens with AMD. Increase the voltage offset a hair. Ive had to do it on some AMD cpus for no loving reason at all other than they crash.
|
|
#
?
Jun 17, 2021 23:42
|
|
|
Lord Stimperor posted:It's an SSD. The SMART status is okay. I've run a short smartctl test and that passed as well. I'll try if I can run the long one while my windows USB stick finishes being prepared. Is it a SATA drive? I'd consider swapping cables and/or ports on it. Assuming your Windows install media is okay, it would be interesting to see if a Linux install on the same drive acted as badly. If Ubuntu is running happily off a USB drive then it would be worthwhile to see if also ran happily if installed on the same target disk you've been using for Windows.
|
|
#
?
Jun 17, 2021 23:52
|
|
|
Lord Stimperor posted:BIOS is (now) updated to the most recent version. All chipset drivers were already the newest. In fact, the windows installation was about a day old when I made the first post, couldn't have been much cleaner. I had a similar issue with my system, a Ryzen 5600x, where my ram xmp profile caused RAM failure errors, and had to stabilize the SOC voltage to 1.1 and the DIMM voltage to 1.38 to get the errors to disappear. To be fair, I didn't check the QVL before buying the RAM because the RAM was on sale and I wasn't gonna pass it up (2x16gb Corsair Vengeance LPX for $150). But I was getting critical error ID 41 like you about once a week for a couple months, I'd be at idle, or watching a movie, or playing an indie game and it would hard lock, and I would hard reboot to see the 41 error. I did a lot of searching about this, I reset my OC back to stock, dropped the RAM back to JEDEC spec, and turned off my GPU OC to see if it would still happen, and it did. After extensive testing of each part in a test bed of different combinations, I concluded that it wasn't a hardware issue per se, and went to Tom's Hardware to get my problem sorted. Turns out on Ryzen motherboards there's a setting in the BIOS called "Power Supply Idle Control", and the default/auto setting sets this option to "Low Current Idle", but the other option, "Typical Current Idle" is sometimes the solution to Ryzen kernel-power failures. I've been running my system for a couple weeks like this, and it solved the hard reboots afaik. I don't know what electrical wizardry makes this work for me, but it hasn't crashed yet. Not saying this is your issue, but it might be, and it took me a couple months to find the solution.
|
|
#
?
Jun 18, 2021 00:02
|
|
|
that's a good point because he says it's stable under a constant load but crashes rapidly when he interacts with it. But interacting with it means that it's going between boost and idle states, maybe it's crashing on the power-state transition. could either be core or memory, on the core side you could try disabling power c-states or even seeing if there's a way to disable whatever the AMD version of speedstep is so that it just runs max clocks all the time. again, not a long-term solution, but maybe it'd help you figure out what's going wrong.
|
|
#
?
Jun 18, 2021 00:31
|
|
|
Try upping voltage a hair.
|
|
#
?
Jun 18, 2021 01:02
|
|
|
SwissArmyDruid posted:I'm starting to think that being an early adopter on Zen 1 was an increasingly raw deal, since I have the exact same problems and keep having to drop memory timings to even boot, but I can't justify spending dosh to get off of this 1600 with how the market is, and I won't get onto a dead-end socket.  Well, turns out both my mobo and my RAM were wack. The 1600 marches onward. SwissArmyDruid fucked around with this message at 01:44 on Jun 18, 2021 |
|
#
?
Jun 18, 2021 01:41
|
|
|
|
| # ? May 15, 2024 19:09 |
|
|
I'm actually impressed by this thing, those esports titles are doing low end magic, I guess: https://youtu.be/KycNI1FxIPc
|
|
#
?
Jun 18, 2021 02:00
|
|
