|
runaway dog posted:like for instance, I keep getting this problem where my computer wont restart anymore, I right click start and hit restart and it just goes black like it's going to restart but everything in my pc just stays on and the only way to "restart it" is to do a hard shutdown by holding the power button for 5 seconds. it happened before and I couldn't figure it out so I reinstalled windows 11 and that fixed it but now it's back, This sounds like software / drivers / OS to me. I think the OS is getting wedged on the last bit of shutdown and not sending the power-off message to the bios. Do you have latest chipset drivers, direct from AMD?
|
# ? May 14, 2023 12:53 |
|
|
# ? May 31, 2024 03:37 |
|
Gigabyte F8d bios has been out for a week or so now without getting pulled down, so I went ahead and updated from F8c this morning. I had EXPO off before, but turned it back on for testing. Booted up with no issues, gonna play some Diablo 4 and run a few 3dmark benches and see how it holds up. So far vsoc hasn't gone above 1.245.
|
# ? May 14, 2023 15:12 |
|
Klyith posted:This sounds like software / drivers / OS to me. I think the OS is getting wedged on the last bit of shutdown and not sending the power-off message to the bios. I was thinking it had something to do with onboard audio reinstalling itself after I cleared cmos and forgot to disable it again, but then I was messing with some memory checking setting that was supposed to fix the longer boot times I get whenever I turn on expo, but as soon as i turned it on and booted and rebooted it hung up on the reboot again. so I changed it back to default and it still stalled a reboot, so then I disabled the onboard audio thing and rebooted and it worked fine. so I rebooted again and went into bios and turned back on that memory checking option and booted and rebooted and it hung again, so then I rebooted and turned it off again but didn't reboot until today before work and it seemed to reboot fine, and now I'm wondering if it was just because I was trying to reboot too quickly after a fresh boot and it was loving things up? who the hell knows but I can't really check until I get home, cant remember the exact name of the option either, it was something like "turning on x will make memory not do a full check on restart if possible". also when it was doing this the first time before I reinstalled windows I hadn't touched that option yet so it makes me feel like its a unrelated problem.
|
# ? May 14, 2023 16:11 |
|
runaway dog posted:I was thinking it had something to do with onboard audio reinstalling itself after I cleared cmos and forgot to disable it again, but then I was messing with some memory checking setting that was supposed to fix the longer boot times I get whenever I turn on expo, but as soon as i turned it on and booted and rebooted it hung up on the reboot again. so I changed it back to default and it still stalled a reboot, so then I disabled the onboard audio thing and rebooted and it worked fine. so I rebooted again and went into bios and turned back on that memory checking option and booted and rebooted and it hung again, so then I rebooted and turned it off again but didn't reboot until today before work and it seemed to reboot fine, and now I'm wondering if it was just because I was trying to reboot too quickly after a fresh boot and it was loving things up? who the hell knows but I can't really check until I get home, cant remember the exact name of the option either, it was something like "turning on x will make memory not do a full check on restart if possible". also when it was doing this the first time before I reinstalled windows I hadn't touched that option yet so it makes me feel like its a unrelated problem. Memory context restore
|
# ? May 14, 2023 16:23 |
|
yeayeayea that was it
|
# ? May 14, 2023 17:27 |
|
Well, I'm sure glad I waltzed through the BIOS after yet another update, after noticing the IGP was enabled. Because what the gently caress is the point of an auto option, if it always defaults to the same setting?!
|
# ? May 17, 2023 20:42 |
|
Subjunctive posted:Not feeling super good about my AM5 ASUS motherboard right now, but the voltages all look OK with the latest BIOS (even on EXPO) and I’m not going to turn on any of the overclocking stuff so I’m sure(?) it’ll be OK. Hmm, it happened again overnight so I guess I’m replacing the PSU after all.
|
# ? May 17, 2023 21:12 |
|
I was wondering how AMD was going to make sure the 7800x3d didn't just ruin the entire AMD CPU stack for consumers and I guess I didn't realize AMD had quite the trick up their sleeve. Do any of the MBs work reliably yet?
|
# ? May 17, 2023 21:31 |
Lockback posted:I was wondering how AMD was going to make sure the 7800x3d didn't just ruin the entire AMD CPU stack for consumers and I guess I didn't realize AMD had quite the trick up their sleeve.
|
|
# ? May 17, 2023 21:41 |
|
Lockback posted:I was wondering how AMD was going to make sure the 7800x3d didn't just ruin the entire AMD CPU stack for consumers and I guess I didn't realize AMD had quite the trick up their sleeve. My gigabyte board (B650 Aorus Pro) had some weird voltage behaviors, but it was never unstable. I set a manual SoC voltage of 1.2 and it's been rock solid for me, with overclocked memory.
|
# ? May 17, 2023 21:44 |
|
Lockback posted:I was wondering how AMD was going to make sure the 7800x3d didn't just ruin the entire AMD CPU stack for consumers and I guess I didn't realize AMD had quite the trick up their sleeve. Gigabyte B650i, no issues with most recent bios, voltage has yet to go over 1.245 since updating. I did see some peeps had front audio issues with that bios, but I don't have any front ports, so can't comment on that.
|
# ? May 17, 2023 22:15 |
|
I'm on a Gigabyte X670 and haven't had any issues on F8c or F8d bios with expo enabled, voltage stays at 1.245
|
# ? May 17, 2023 23:47 |
|
Asus B650E-F and my SoC voltage hasn't gone above 1.3v pre or post BIOS update but I'm also running a 7900X.
|
# ? May 18, 2023 00:08 |
|
My local microcenter has a huge stack of asus am5 returns. I was going to sit out the first generation of am5, but this is tempting.
|
# ? May 18, 2023 00:08 |
|
Yudo posted:My local microcenter has a huge stack of asus am5 returns. I was going to sit out the first generation of am5, but this is tempting. whats the deal with these? smoked or??? heavily discounted?
|
# ? May 18, 2023 01:45 |
|
My Asus X670E-PLUS has always been stable. Memory is at 6000 with that DHOC option (or whatever is called, it's not EXPO, but does the same thing). I do use 7950X though. I upgraded to the latest bios out of caution, not need.
|
# ? May 18, 2023 02:03 |
|
yummycheese posted:whats the deal with these? smoked or??? heavily discounted? They are open box being resold: a return. They work fine, ostensibly, though they may be missing sata cables or whatnot. Some returns are due to annoying poo poo like coil whine (though even newegg will occasionally rma a whining product), or because someone decided they wanted the xtx over the xt etc. I would never buy a newegg open box just due to how scummy they are. Microcenter and Amazon will accept returns on open box purchases and don't usually gently caress around.
|
# ? May 18, 2023 02:24 |
|
Yudo posted:
hard agree. I have a microcenter local to me and generally speaking they know whats up and dont mind taking back a open box item. so might take the gamble here. I have everything else. a mobo and cpu would get me going
|
# ? May 18, 2023 02:49 |
|
Lockback posted:Do any of the MBs work reliably yet? I've got a GB X670E Master.
|
# ? May 18, 2023 05:24 |
|
Lockback posted:I was wondering how AMD was going to make sure the 7800x3d didn't just ruin the entire AMD CPU stack for consumers and I guess I didn't realize AMD had quite the trick up their sleeve. 78x3d with a gigabyte x670 running F8c and 6000 ram, no problems.
|
# ? May 18, 2023 08:24 |
|
I'm feeling pretty mixed about my 7800X3D too but at least it turned out the problems I were having was related to broken Wow addons and not some issue with the chip or mobo. That being said it does some pretty bonkers stuff, this probably means nothing to most of yall but you can get 4K/70-80fps in crowded Valdrakken in World of Warcraft. Which is insane because the game uses like... 5% of one core to run. Some other things I've liked a lot about AM5, for balance: 1) the build quality on the mobo that I got (TUF 670E Wifi) has really good components for a relatively decent price (paid $250 on Amazon warehouse), the antenna is great for a solid wifi 6E signal. Everything has been pretty turn key at the end of the day despite all the politics afoot. 2) My AM4 motherboard had tons of issues with XMP profiles and it was a constant battle getting things to work stable. The board never worked with its own XMP profile which is obviously a terrible look. Meanwhile the AM5 motherboard is rock solid on the DDR 6000 expo it shipped with, though I guess it's hard to pin down what is attributed between the ram sticks and mobo. 3) Armoury-Crate doubled my ram for free? One weird trick Lisa Su hates! You won't believe what happens next Seriously though do we have a good grasp on how many users are being affected by all of these issues? I am skeptical on some level because everyone and their dog was freaking the gently caress out about the 4090 catching fire and that turned out to be the biggest non-thing ever. I'm open to the idea that my cpu/mobo sucks but what is the actual failure rate here...
|
# ? May 18, 2023 11:42 |
|
Subjunctive posted:Hmm, it happened again overnight so I guess I’m replacing the PSU after all. FWIW I am super proactive about swapping out a PSU in circumstances like yours, because a bad PSU might not be limited to just powering off your system or whatever. A bad PSU can damage things. Friend of mine had his decade old PSU go bad with random occasional power-off behavior. He waited over a month before he bothered me for tech support. Shortly after replacing the PSU his GPU also died, which I can't help but find suspicious. Taima posted:
Incredibly low. Like, the number of reported CPUs burned up is even fewer than the 4090 power cables. But the valid reason for the general freak out is that 1.4v will damage the CPU over time, even if it doesn't immediately go up in smoke. It's actually kinda lucky (for both users, AMD, and Asus) that a few people did have dead CPUs. If this had gone unnoticed for a year and then CPUs started being unstable and dying from long-term overvolting, it would have been much worse.
|
# ? May 18, 2023 14:01 |
|
Just in case anyone cares my Asus Z790 intel box is unstable doing one thing.. can crash it at will and although I was sure it was the CPU I just had the chance to try another Asrock Z790 mobo and the cpu isn't the problem. Im NOT overclocking either.
|
# ? May 18, 2023 15:34 |
|
looks like the new asus bios is finally up and it's not a beta: v1416, at least for my specific mobo anyway. 1. Update AGESA version to Combo AM5 PI 1.0.0.7.a 2. Support 48/24GB high-density DDR5 memory module. 3. Memory QVL amended to account for AMD 1.3V SoC voltage limit. 4. EXPO/XMP prompt notice removed. curious about what number 4 means, haven't had a chance to install it yet. edit: seems good so far, all the numbers look good expo 1 booted no problems, don't think memory context restore was working properly on v1303 at least for me it didn't seem to do anything, but seems to be working now, boot times are definitely better, hopefully it stays that way. runaway dog fucked around with this message at 17:17 on May 18, 2023 |
# ? May 18, 2023 16:38 |
|
https://www.youtube.com/watch?v=3mFb1cloPJA It seems some Ryzen 7000 chips needed a VSOC of 1.35 to be stable with a 2r2dpc configuration (4x32GB), even at sub-JEDEC speeds, so the new bios versions that limit you to 1.3V are essentially now locking those chips out of that memory configuration entirely. But at least 96GB (2x48) configurations seem doable and quite fast. edit: Granted, VSOC of 1.35 isn't great for long-term use due to degradation concerns, but I guess this is more about how shoddy AMD's support is for four-dimm configurations in general. Dr. Video Games 0031 fucked around with this message at 17:26 on May 18, 2023 |
# ? May 18, 2023 17:21 |
|
Dr. Video Games 0031 posted:https://www.youtube.com/watch?v=3mFb1cloPJA I think that is what has happened with my setup since I am at sub jdec speeds.
|
# ? May 18, 2023 17:25 |
|
Also that Gigabyte board Wendell gave the SOC a voltage in excess of 1.4V out of the box during the initial memory training, which is insane, lmao. He speculates this is probably due to the bug that caused gigabyte boards to "remember" previous VSOC settings, even when the BIOS settings have been reset to default.
|
# ? May 18, 2023 17:30 |
|
runaway dog posted:looks like the new asus bios is finally up and it's not a beta: v1416, at least for my specific mobo anyway. I'm assuming when they say the QVL has been updated, that means the memory listed on this link? https://rog.asus.com/us/motherboards/rog-strix/rog-strix-b650e-i-gaming-wifi-model/helpdesk_qvl_memory/ Or is there somewhere else I should be looking? I don't see a QVL list in the manual and it just says to check the website. If that link I posted is the right place to look, then it appears that my memory is still on the QVL list, so, yay!
|
# ? May 18, 2023 17:31 |
|
Dr. Video Games 0031 posted:I guess this is more about how shoddy AMD's support is for four-dimm configurations in general. There are some pretty severe technical limitations that come into play to support 2 dpc, and the performance gap between 1 and 2 is only going to grow. The whole industry would be better served by acknowledging that modern ddr has line-of-sight to signaling rates > pcie g3 and 1 dpc is the right physical topology to support that.
|
# ? May 18, 2023 17:49 |
|
SlapActionJackson posted:There are some pretty severe technical limitations that come into play to support 2 dpc, and the performance gap between 1 and 2 is only going to grow. The whole industry would be better served by acknowledging that modern ddr has line-of-sight to signaling rates > pcie g3 and 1 dpc is the right physical topology to support that. Isn’t it the same as it’s always been and the real issue is ranks per channel, not DIMMs per channel per se? 2r2dpc has always been slow.
|
# ? May 18, 2023 17:56 |
|
I see that the non-beta version of the Gigabyte F8 bios is out as of yesterday, but still on the old 1.0.0.6 AGESA version. I'm on F8d currently with no problems, so don't see any reason to update again.
|
# ? May 18, 2023 17:59 |
|
hobbesmaster posted:Isn’t it the same as it’s always been and the real issue is ranks per channel, not DIMMs per channel per se? Multirank dimms pose some of the same challenges, but at a smaller scale. It's easier to make the interface work for one 2r dimm than two 1r dimms. It's a good compromise for enabling high memory density on desktops until the ecosystem is ready to move to rdimm.
|
# ? May 18, 2023 18:56 |
|
Updated my Asus BIOS to the newest version and my SOC voltage is 1.264 and I am still able to run my memory at the 6000mhz docp profile with no stability issues, so far.
|
# ? May 19, 2023 01:37 |
|
Kibner posted:I'm assuming when they say the QVL has been updated, that means the memory listed on this link? https://rog.asus.com/us/motherboards/rog-strix/rog-strix-b650e-i-gaming-wifi-model/helpdesk_qvl_memory/ Or is there somewhere else I should be looking? I don't see a QVL list in the manual and it just says to check the website. seems like a good deduction, I actually don't know but based on what you posted I'd say you nailed it.
|
# ? May 19, 2023 02:29 |
|
Updated the BIOS on my Asus B650E-F to 1616 which I think is the first non-beta since the VSOC drama began. No issues updating the BIOS from a USB 2.0 flash drive plugged into non-flashback port. 3DMark CPU score is slightly higher than the previous 1602 beta BIOS, though only 1.5% so within run-to-run variance. It's also the highest score since building this 7900X system, though I'm manually recording them in my notes app so I haven't tracked all the runs. HWinfo showed a max VSOC of 1.240V. Strangely, CPU (Tctl/Tdie) hit 90.8C even though none of the individual cores went above 80C and the CCDs maxed at 85.5C. 1616 feels like a good update. Josh Lyman fucked around with this message at 03:43 on May 19, 2023 |
# ? May 19, 2023 03:35 |
|
Josh Lyman posted:
Tctl/Tdie has always been hotter than Tdie I've noticed on my 78, usually a good 10c higher though it can get closer to Tdie if the cpu is mostly idle. Before this I didn't pay too much attention to the difference between the two values with my 39x so this might be normal for all I know.
|
# ? May 19, 2023 04:31 |
|
The CPU Tdie does indeed tend to be a little hotter than the CCD Tdie on my 7800X3D. Which I guess means the hot spot is somewhere on the I/O die? That seems odd to me, but I guess that's normal?
|
# ? May 19, 2023 05:25 |
|
I wonder if this is part of what people were saying about the 5800x3D and the hotspot on that moving away from the center and out towards the edge of the die making it harder for coolers to work optimally. At least, that's what I recall seeing back around when it came out.
|
# ? May 19, 2023 05:58 |
|
Ardryn posted:I wonder if this is part of what people were saying about the 5800x3D and the hotspot on that moving away from the center and out towards the edge of the die making it harder for coolers to work optimally. At least, that's what I recall seeing back around when it came out. The hotspots on chiplet ryzen have always been off center. It’s just that with the 5800x3d cooling was less effective due to the 3D cache so everyone was looking for every optimization possible.
|
# ? May 19, 2023 17:36 |
|
|
# ? May 31, 2024 03:37 |
|
I remember seeing people say to not use the direct heatpipe contact coolers with chiplet Ryzens because one heatpipe will get all the heat. That seemed logical enough to me.
|
# ? May 19, 2023 17:43 |