My computer keeps freezing right when I play games, starting two weeks after buying it.
My computer keeps freezing right when I play games, starting two weeks after buying it.
with a simpler, more casual tone while keeping your first-person perspective intact:
That, i do not believe. Well, assembly wise, perhaps. But not by component selection wise. Since if you did, you would not have bought this: Because IF you did the background checks for PSU, you've learned that, at best, it is mediocre quality unit. More like low quality (or even crap quality since Gigabyte lies about it's specifications). On top of that, it comes only 5 years of warranty. Which, at today's age, is low for proper PSU. Official specs: https://www.gigabyte.com/Power-Supply/GP-UD750GM#kf Review 1: Review 2: https://www.techpowerup.com/review/gigab...0gm-750-w/ PSU reliability can be seen from the warranty length it has been given. In a nutshell: up to 2 years - terrible reliability 3 years - poor reliability (e.g Corsair VS/CS) 5 years - mediocre reliability (e.g Be Quiet! Straight Power 11, Seasonic G12, Corsair CX/CXF) 7 years - good reliability (e.g Seasonic Core/Focus GM, Corsair TX/AX) 10 years - great reliability (e.g Seasonic Focus GX/PX, Corsair RMx/HX/HXi/AXi) 12 years - superb reliability (e.g Seasonic Vertex/PRIME) So, aim for a 10 year warranty PSU. 7 year warranty would be bare minimum. (E.g my 3x PCs are powered by Seasonic. I have 2x PRIME TX-650 units and one Focus PX-550 unit. Full specs with pics in my sig.) Random reboots or shutdowns are mostly caused by 2 issues: 1. CPU/GPU overheats and to prevent any damage, system shuts down. 2. PSU fails to deliver enough power to the GPU or fails to keep smooth enough voltage for PC's operation. First check your CPU/GPU temps, both at idle and under load. If temps are within reason then it's safe to assume that it's the PSU who is acting up. PSU issue can be easily validated with 2nd, known to work, good (or great) quality PSU. If you do end up replacing your PSU, i suggest you do your homework 2nd time around. Since PSU powers everything, it is the most important component inside the PC. Hence why NEVER cheap out on PSU! Also, never buy used PSU either. For your build 750W unit is enough. Here are best ATX 3.x units to choose from: link: https://hwbusters.com/best_picks/best-at...busters/3/ Kernel Power event ID 41 usually shows up when PC looses power suddenly. Usually to do with PSU issue. Rarely, MoBo VRM issue. Can also be issue of the main power grid (e.g wiring in your home). UPS fixes main power grid issues (well, it won't "fix" them, but offers stellar protection against brownouts, surges and blackouts. Given that you buy proper UPS, that is.).
I'm going to grab some screenshots showing the temperatures at different power levels so I can share them here soon. From what I see, things are actually good. I can run my GPU and CPU at full speed without any problems. My fans never get loud or hot. When picking out parts earlier, I always thought about keeping cooling in mind. Back when RDR2 first loaded into the game a little while after starting, it crashed with an error saying "Attempted execute of NoExecute Memory." That's different from my other crashes because this was the only one that happened before the computer shut down completely. Also, later on, the game had trouble rendering graphics; some models showed up as just black eyes and clothes instead of full characters, and I hit spots in the map where the scenery turned black too. After restarting the PC, I've been playing for a couple hours without any issues now. Will post the temp screenshots right after.
Okay so I'm just starting out with this stuff, so if I'm messing something up please tell me. I realize that I didn't run those GPU tests for very long at all - but I feel like the switch test would catch PSU problems really fast anyway. And I've done the constant load test a few times now and the numbers have stayed pretty much the same for ages. Here is an Imgur link to my screenshots, or you can find one here on postimg
Your GPU gets pretty hot but is fine for its job. The RX 7600XT usually runs around 60 to 70 degrees when it's busy, and the specs say that 58 degrees is actually okay. However, your CPU is way too hot. A good Ryzen 5 7600X with a proper cooler should hit about 88 degrees under heavy load, even lower for gaming loads. Yours is hitting up to 95 degrees, which is the maximum safe limit for that chip. You need to check what type of CPU cooler you have, but I can't see it listed in your specs. Overheating the CPU makes everything go haywire and could cause crashes or random power-offs. Also, how many fans are there? How big are they? Do they blow air from the front or the back? And what kind of fan do they use? Are those stock fans included with the case, or did you change them out yourself? Since your CPU is clearly not cooling down properly, you won't see temperatures like these coming out of that chip.
I own a Thermalright Phantom Spirit 120 SE along with an Antec Flux case and stock fans. Even when I play games, the air inside feels cool to touch because those fans are great for keeping things cold. This is just how Zen 4 CPUs work; they push themselves up to their max heat limits and try to keep running while gaming. In fact, my CPU never even gets near its highest temp limit when I play, I've never seen it hit close to 80 degrees Celsius. You can see where Linus talks about this feature of AMD's Zen4 thermal management strategy as well. Also check this direct quote from AMD themselves!
Yes, but if your CPU hits Tjmax (the max safe temperature), it will cut performance. The two links I showed actually used very powerful chips that could handle high temps easily, while Linus's video also mentioned thermal throttling. However, for the R5 7600X, you shouldn't see temperatures at Tjmax during normal gaming or testing. At the point shown in the link (about a minute and fifty seconds into Cinebench), your chip is still well below that limit. The King of Air Coolers usually keeps temps under 90C here. Let's double-check this with HWinfo64. Just make sure all your case fans are at full speed and run the test again. It will be noisy, but it only happens while testing. If you see temps drop below 95C, that means your cooling is working fine. But if they stay high, maybe just one or two of your fans aren't spinning fast enough. You could try turning all fans up to 100% and test again to see if that helps or causes any crashes. Once you figure out what's wrong (thermal issues or a bad power supply), you might not need to buy a new PSU. It could just be the time to upgrade your fan settings or add more exhaust air. Since these cases come with five fans already installed, it's simple to get better airflow than some other models that only have two or three fans. Speaking of airflow, here are two pictures comparing a standard setup with upgraded ones: In the first one, you see the Corsair 760T in white with stock fans. It has two front intake fans and one rear exhaust fan. Or the second one shows the same model in black with user-installed fans. This one has more powerful intake and exhaust fans to move better air around. How do you think that makes a real difference?
All this talk about Zen4 architecture running hot smells fake. It just looks like AMD tried to get PR from people who had some real thermal problems and put that "by design" excuse on them. The CPU actually slows down when it gets too hot (thermal throttling), so if you're always at the limit, your chip is underperforming constantly.
Also, the tests in your video only show that the flagship 7950x hits the max temp, while the 7600x never does even with coolers. But none of this matters much because games or computers crash anyway if you get too hot. And according to my experience, the main reason for crashes is usually RAM instability. Especially if you try to run a 6000 MT/s CL30 kit on an AM5 board at 1.35 volts. Try playing your games without enabling expo first. If you turn it on and crank up the voltage to 1.4 volts, see if things still crash?
You could also try running the AIDA64 stress test with memory included (that's important). If you have RAM issues, the test will probably stop halfway through because it detected a hardware problem. If AIDA runs without errors for 15-20 minutes, your system is likely fine.
It's not only AMD but Intel too, since 12th gen. From 12th gen, Intel top-end chips operated in 90C area, review: https://www.techpowerup.com/review/intel...en/21.html While starting from 13th gen, the 95C/100C operating temp became the new "norm", review: https://www.techpowerup.com/review/intel...0k/23.html And when chip reaches Tjmax on stock clocks, it means there's 0 OC headroom. Hence why today, CPU OC is dead. Back in the old days, e.g Intel 6th gen, i7-6700K peaked at 73C with stock clocks (Tjmax 100C), giving plenty of OC headroom, review: But what we see today, all comes down to where CPU makers have to release better and better performing CPUs (compared to previous gen), while the R&D in efficiency is lacking way behind. Latest top-end chips are already operating at Tjmax and ceiling is hit. Now, CPU makes have to invest into efficiency, since there is no way to increase performance. That, actually, Intel has begun to do, with their latest Core Ultra 200 series. Latest chips are more power efficient but since there's almost no performance uplift over previous (14th gen), almost no-one is buying the latest Intel chips. Especially since those cost considerably more than previous gen. And current AMD offering (Ryzen 9000-series) is both cheaper and at times, better performing as well. Though times for CPU makers currently. In terms of RAM, download and run memtest86, link: https://www.memtest86.com/ Guide to install and use it: https://www.memtest86.com/tech_creating-window.html 1 full pass (all 15 tests) is bare minimum. 2 full passes are better while 4 full passes is considered acceptable. Since it takes a while, best to let it run overnight. 1x 8GB one full pass takes ~1h. 2x 8GB ~2.5h. So, ~5h for your 2x 16GB per 1 pass. Or ~20h for 4 full passes. If there are no errors - RAM is sound.
Yeah, skipping the 7000 series was a reason for me, but I jumped on the 9800X3D right away when it came out.