|
shrike82 posted:we can argue about qualitative improvements but that CPU number jumping by 16% is pretty misleading interesting i had RAM running at 2400 and I didn't notice it while I had it, but changing it to 3200 was a noticeable difference anywhere I was CPU bound (which was quite a bit when I was on zen+ 2600) and huge in ARMA. Night and day? Not at all, but definitely noticeable on frame minimums (Arma used to drop sub 30, now sticks above). Possible I 'notice' it more because it's not kicking in the sub-VRR framerates. Not sure I'd ever notice in Apex Legends.
|
#
?
Feb 26, 2021 13:11
|
|
|
|
| # ? May 30, 2024 21:46 |
|
|
i'm on a 3960X with a A6000 and 64GB on quad channels  i took the opportunity to run the AC: Valhalla benchmark + TimeSpy + ran around a fixed city space in CP2077 before and after I flipped XMP back on
|
|
#
?
Feb 26, 2021 14:01
|
|
|
shrike82 posted:One of the biggest points to remember during all of this -- and any other memory testing published by other outlets -- is that motherboard matters almost more than the memory kit itself. Motherboards are responsible for most of the timings auto configured on memory kits, even when using XMP, as XMP can only store so much data per kit. The rest, including unsurfaced timings that the user never sees, are done during memory training by the motherboard. Motherboard manufacturers maintain a QVL (Qualified Vendor List) of kits tested and approved on each board, and we strongly encourage system builders to check these lists rather than just buying a random kit of memory. Motherboard makers will even tune timings for some kits, so there’s potentially a lot of performance lost by using mismatched boards and memory. There are games that will get benchmark-level fps/tps increases from tuning memory. Factorio is one of them. Minecraft is another. Starcraft 2 is likely another. That's my short list of stuff it has mattered in. Khorne fucked around with this message at 15:05 on Feb 26, 2021 |
|
#
?
Feb 26, 2021 14:22
|
|
|
Why would there be / are there hidden memory timings that don't have a setting stored in xmp?
GRINDCORE MEGGIDO fucked around with this message at 19:43 on Feb 26, 2021 |
|
#
?
Feb 26, 2021 19:37
|
|
|
It's mind-blowing how there's no 595X stock any-loving-where. I've been actively trying to buy for 8 weeks now. I don't have a discord command center set up or anything, but I am checking in about twice a day. First world problems and all, but this is some crazy poo poo. I feel there's the very real possibility of a new chip lineup being announced before I can get my hands on one of these.
AARP LARPer fucked around with this message at 19:42 on Feb 26, 2021 |
|
#
?
Feb 26, 2021 19:39
|
|
|
My local Microcenter has them in stock - don't suppose you've got one?
|
|
#
?
Feb 26, 2021 20:58
|
|
|
Unfortunately, no.
|
|
#
?
Feb 26, 2021 21:02
|
|
|
I know X570 uses more power than B550 under load, but how does it compare at idle and light workloads?
|
|
#
?
Feb 26, 2021 21:30
|
|
|
MaxxBot posted:Idk if you have Cyberpunk but while messing with stuff I tried JDEC RAM speeds and I got huge FPS drops running through crowds in the city, putting the memory back to 3800 I saw no drops running through crowds vs doing anything else. Also I've definitely seen at least like 5%+ FPS gains going from XMP to tuned timings testing at 1080 high. (*: It doesn't go above 4GB memory usage on my system with 32GB RAM.)
|
|
#
?
Feb 26, 2021 23:13
|
|
|
Ever since I upgraded from 3700x to 5900x, I keep getting random resets when idle. I dont even see a blue screen. any ideas how to solve?
|
|
#
?
Feb 26, 2021 23:45
|
|
|
teraflame posted:Ever since I upgraded from 3700x to 5900x, I keep getting random resets when idle. I dont even see a blue screen. any ideas how to solve? What does event viewer say
|
|
#
?
Feb 26, 2021 23:47
|
|
|
B-1.1.7 Bomber posted:There seems to be a fair amount of magic and/or voodoo involved in over-clocking these days, so yeah, it would be real nice to have some robust tuning software like that. Maybe there's just too much turnover in mobos/cpus to make such a thing worthwhile, although the AM4 has had a nice run. I post this guy every 20 pages or so.. no one ever comments though. It exists. Quaint Quail Quilt posted:Sure it is, for the CPU and GPU.
|
|
#
?
Feb 27, 2021 00:15
|
|
|
teraflame posted:Ever since I upgraded from 3700x to 5900x, I keep getting random resets when idle. I dont even see a blue screen. any ideas how to solve? Did you tweak any of the CPU related settings in the BIOS?
|
|
#
?
Feb 27, 2021 00:31
|
|
|
teraflame posted:Ever since I upgraded from 3700x to 5900x, I keep getting random resets when idle. I dont even see a blue screen. any ideas how to solve?
|
|
#
?
Feb 27, 2021 03:09
|
|
|
Munkeymon posted:My local Microcenter has them in stock - don't suppose you've got one? I've never seen them in stock at Microcenter here (Denver).  edit: I went through every Microcenter location, and 4 each had 1 5950x in stock. One place in Kansas, 2 in Ohio, and one in Houston. Kilson fucked around with this message at 03:20 on Feb 27, 2021 |
|
#
?
Feb 27, 2021 03:14
|
|
|
teraflame posted:Ever since I upgraded from 3700x to 5900x, I keep getting random resets when idle. I dont even see a blue screen. any ideas how to solve? Up the LLC to medium for starts. if that doesnt work you gotta apply a voltage offset... before you ask. I have the same loving problems and so far this worked.
|
|
#
?
Feb 27, 2021 04:40
|
|
|
Quaint Quail Quilt posted:I post this guy every 20 pages or so.. no one ever comments though. It exists. Right on. I bookmarked it. Thank you and see you in another 20 or so.
|
|
#
?
Feb 27, 2021 04:57
|
|
|
terrorist ambulance posted:What does event viewer say Khorne posted:I set vsoc to 1.125v (reads just under 1.1v vsoc in hwinfo64) to stop this with my 5950x. No reboot in over a week. It drove me insane before then coming from a stable 3900x setup. redeyes posted:Up the LLC to medium for starts. if that doesnt work you gotta apply a voltage offset... before you ask. I have the same loving problems and so far this worked.
|
|
#
?
Feb 27, 2021 07:45
|
|
|
teraflame posted:Level is critical, source is kernel power, event ID 41. load line calibration Commonly taken to mean the setting that affects how much the voltage drops under heavy load
|
|
#
?
Feb 27, 2021 15:16
|
|
|
teraflame posted:Level is critical, source is kernel power, event ID 41. Sorry, Load Line Calibration. Put it at medium, or one notch above standard. Then for voltage offset, no you gotta do a positive (+) offset. Try maybe +.05v for vcore if the LLC doesnt work.
|
|
#
?
Feb 27, 2021 16:10
|
|
|
Do goons have any advice regarding the B550-E Strix vs the X570-E version? I want a motherboard with x8/x8 PCIE4 so I can run both a GPU and a host card full of m.2 drives, and 2.5Gig as an option for the future. The X570’s seven USB3.1 ports is very impressive next to the mostly 2.0 ports on the B550, but I’m thinking I’d add a 3.1 card to the PCI3 slot on B550 to compensate. Basically, is it worth the ~$60 difference to go X570 here?
|
|
#
?
Feb 28, 2021 01:34
|
|
|
Craptacular! posted:Do goons have any advice regarding the B550-E Strix vs the X570-E version? I want a motherboard with x8/x8 PCIE4 so I can run both a GPU and a host card full of m.2 drives, and 2.5Gig as an option for the future. The X570’s seven USB3.1 ports is very impressive next to the mostly 2.0 ports on the B550, but I’m thinking I’d add a 3.1 card to the PCI3 slot on B550 to compensate. I have a Gigabyte board, but I ended up going X570 because it felt like I was keeping more options open than going with a Bxxx. I'm not sure the B550 boards were out at the time, but I would have gone X570 anyway. I think X570 generally has better power-handling characteristics than B550, but I could definitely be wrong about that.
|
|
#
?
Feb 28, 2021 01:58
|
|
|
Craptacular! posted:Do goons have any advice regarding the B550-E Strix vs the X570-E version? I want a motherboard with x8/x8 PCIE4 so I can run both a GPU and a host card full of m.2 drives, and 2.5Gig as an option for the future. The X570’s seven USB3.1 ports is very impressive next to the mostly 2.0 ports on the B550, but I’m thinking I’d add a 3.1 card to the PCI3 slot on B550 to compensate. This maybe not the answer that you were looking for but with the need for all the pcie lanes can your budget stretch to threadripper?
|
|
#
?
Feb 28, 2021 17:28
|
|
|
HERAK posted:This maybe not the answer that you were looking for but with the need for all the pcie lanes can your budget stretch to threadripper? I'm not that insane. I just dream of a future where all my storage is fast m.2, therefore I'd like to put a GPU in one slot and leave the other to potentially have a bunch of nvme drives. In the case of the B550 board, the remaining PCIE gen3 slot would be used making up for all the lost USB3 ports, because I currently have around four of them on my ancient rear end machine and don't want less. X570 is ~$60 more, but it's almost worth it since the USB card would nearly cost that much on it's own from my research.
|
|
#
?
Feb 28, 2021 17:47
|
|
|
Craptacular! posted:Do goons have any advice regarding the B550-E Strix vs the X570-E version? I want a motherboard with x8/x8 PCIE4 so I can run both a GPU and a host card full of m.2 drives, and 2.5Gig as an option for the future. The X570’s seven USB3.1 ports is very impressive next to the mostly 2.0 ports on the B550, but I’m thinking I’d add a 3.1 card to the PCI3 slot on B550 to compensate. The key word you're looking for is pcie bifurcation and you defacto need x570 boards for that to get 2 working. You need threadripper for 4. edit: Oh and be careful to see which PCIE slots actually support bifurcation, often times its only the primary graphics card slot so you'd have to put the m.2 card in that slot and the primary graphics card in the now 8x other pcie slot. hobbesmaster fucked around with this message at 17:59 on Feb 28, 2021 |
|
#
?
Feb 28, 2021 17:53
|
|
|
hobbesmaster posted:The key word you're looking for is pcie bifurcation and you defacto need x570 boards for that to get 2 working. B550-E supports it. B550-F does not. ASRock's Taichi does, too. Don't know about other brands, I never considered them.
|
|
#
?
Feb 28, 2021 18:40
|
|
|
https://twitter.com/ExecuFix/status/1365981401808580614 Supposedly Zen4 CCDs are staying at 8 cores, so thats a lotta dies and a lotta heat. 400w in that small a space gonna be somethin to cool...
|
|
#
?
Feb 28, 2021 21:28
|
|
|
Cygni posted:Supposedly Zen4 CCDs are staying at 8 cores, so thats a lotta dies and a lotta heat. 400w in that small a space gonna be somethin to cool... First thought: "That's insane power usage!" Second thought: "Wait, that's 4.17W per core. Yessssssssss." Edit: I run my 3950Xs at 95W PPT, which is 5.93W/core. Stock (142W) would be 8.88W/core, and Eco mode (87W) is 5.43W/core. So if there's any legitimacy to this, and they're managing current-gen Eco Mode-equivalent performance, with 96 physical cores, at 4.17W/core, then Zen 4 would be looking amazing. That's several "if"s in a row, though, so don't go hog wild yet. I will be looking forward to the incredibly boring reveal, where the three slides of technical info are presented after 45 minutes of corporate fellatio to/from HP and other channel partners. mdxi fucked around with this message at 22:10 on Feb 28, 2021 |
|
#
?
Feb 28, 2021 21:53
|
|
|
6096  What’s the ETA on Genoa?
|
|
#
?
Feb 28, 2021 21:55
|
|
|
I completed an upgrade from an Intel Z170 build w/Intel I7-6700K to an X570 based 5800X build about a week ago, but kept my RAM/GPU/PSU (2x 16GB Corsair DDR4 3600, EVGA 2080ti FTW3 Ultra, Corsair 750W Gold) from the previous build that had been fairly rock steady for about 2 years. I am having two issues that just will not go away, no matter what combination of BIOS tweaks and OS tweaks I throw at this thing: 1) USB Devices stuttering/disconnecting fairly frequently. I just finished locking down all my PCI Express devices to default to 3 instead of 4, and also downgraded all my USB ports from 3.0 to 2.0 in an attempt to fix this stuttering. If this helps I will report back. 2) In mainly Unreal Engine games, I will get anywhere from 1 to 4 crashes per day, all with the same (or similar sounding) error message: D3DDEVICERESET/D3DDEVICEHUNT/etc. I've tried DDU, I've tried adjusting power settings on the GPU, underclocking it by 100Mhz, rolling back to stock Windows 10 NVIDIA driver version (this actually helped limit the crashes to 1 or 2 every couple of days), but nothing has really helped with the stability of the system. I've seen the 2nd issue been mentioned a handful of times on Reddit and Google, but I haven't found much elsewhere. I've thrown Intel Burn test and Furtmark at this build for 3-4 hours at a time and it's been rock steady during those stress tests, so I don't feel like this is a hardware issue with my Mobo/CPU/GPU. Has anyone run into similar issues with AMD X750 based builds and RTX GPU's before? Edit: Forgot to mention, the D3DDEVICERESET errors are always accompanied by a WIndows Event Log System Error: The description for Event ID 14 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.
|
|
#
?
Feb 28, 2021 23:45
|
|
|
I've also had issues with nvidia drivers crashing since getting my 5800X on my X470 board. I always suspected it might have been my RAM since I had a few issues with it before. I tried a bunch of things like completely uninstalling the the graphics driver and reinstalling but it kept happening. Eventually I got new RAM because I wanted to upgrade anyway and new drivers came out and it seems to have stopped, but no idea if I've fixed it. Mine was specifically that the driver crashed.I searched what I had into google and there were all kinds of threads spanning months of people having this issue with no real resolution.
|
|
#
?
Mar 1, 2021 00:26
|
|
|
So I disabled AMD SMT (multithreading) and I haven't had a crash since (3+ hours), hopefully that's the "fix"
|
|
#
?
Mar 1, 2021 01:03
|
|
|
Wicaeed posted:So I disabled AMD SMT (multithreading) and I haven't had a crash since (3+ hours), hopefully that's the "fix" I'm surprised nobody has chimed in yet to advise really looking at your memory settings. One of the first things I think would be suggested is to look at whether you are running XMP settings or not, and if you are running XMP to turn it off and see if the problem goes away. If it does then you probably need to manually configure your memory settings instead of using XMP. I'd suggest turning SMT back on and turn XMP off and see what the results are. I'd also put your USB settings back as that is most likely unrelated to your issues - the PCIe setting is much more probable.
|
|
#
?
Mar 1, 2021 01:07
|
|
|
Well 3+ hours later had the same issue, so doubt SMT is it. I disabled SMP (DOCP on AMD) and manually set the freq to 3200MHz on the memory. poo poo's frustrating =\
|
|
#
?
Mar 1, 2021 03:25
|
|
|
I do not want to give the impression that this is a widespread / common issue, but did Zen 2 have anything even close to having this kind of USB disconnection issue that we keep hearing about on the Zen 3?
|
|
#
?
Mar 1, 2021 03:34
|
|
|
Wicaeed posted:Well 3+ hours later had the same issue, so doubt SMT is it. No, like, people are saying stop overclocking your memory and run at 2400 or 2133 and eliminate memory as a potential problem here. Or just put your board back to full defaults and start over not overclocking anything. gradenko_2000 posted:I do not want to give the impression that this is a widespread / common issue, but did Zen 2 have anything even close to having this kind of USB disconnection issue that we keep hearing about on the Zen 3? Zen2 is actually affected as well. It seems to be an X570 problem rather than a Zen3 problem from what I’ve read.
|
|
#
?
Mar 1, 2021 03:34
|
|
|
Paul MaudDib posted:Zen2 is actually affected as well. It seems to be an X570 problem rather than a Zen3 problem from what I’ve read. ohhhhhhhhhhhhh okay. that's interesting.
|
|
#
?
Mar 1, 2021 03:38
|
|
|
Paul MaudDib posted:No, like, people are saying stop overclocking your memory and run at 2400 or 2133 and eliminate memory as a potential problem here. Don't see how running the memory at it's advertised speeds (CMW32GX4M2C3200C16) counts as "overclocking" but I'll try anything at this point =\ Edit: And from what I recall this problem actually existed out of the box, which is why I started tweaking the defaults to see if it was some unoptimized default setting that was causing this Wicaeed fucked around with this message at 03:49 on Mar 1, 2021 |
|
#
?
Mar 1, 2021 03:46
|
|
|
yeah, I guess 3200 is the supported speed on Zen3 now. Still probably better to just eliminate it as a variable and then keep going from there.
|
|
#
?
Mar 1, 2021 03:57
|
|
|
|
| # ? May 30, 2024 21:46 |
|
|
Wicaeed posted:Don't see how running the memory at it's advertised speeds (CMW32GX4M2C3200C16) counts as "overclocking" but I'll try anything at this point =\ Anything other than JEDEC settings is technically overclocking, iirc. Just turn off XMP and let your motherboard pick the RAM speed as it should default to the fastest JEDEC speed offered for your RAM. That is the only thing that is "guaranteed" to be stable.
|
|
#
?
Mar 1, 2021 03:58
|
|
