Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
alex314
Nov 22, 2007

My friend wants to build PC for light workstation loads. So far he likes B650 and 7700X. Is there 64GB set that works with Ryzen at 6000MT? I've tried looking at MB QVLs but I can't find anything. Is 5600MT good enough for general work stuff?

Adbot
ADBOT LOVES YOU

Klyith
Aug 3, 2007

GBS Pledge Week

Fabulousity posted:

I got a half-built computer sitting on my dining room table that will be a 7800X3D on a MSI B650 Tomahawk. Still waiting on the M2 for the main drive so I have it just sitting there partially wired up but all these posts have me just going :catstare:

I figured I'd have to do a BIOS update before attempting to enable EXPO anything but MSI's latest update is dated April 14th and is marked as beta. AMD's official AGESA update to deal with the X3D problem was released on 4/27. Is MSI not having problems or just not caring so far?

So far I don't think anyone's shown a MSI board that was throwing unsafe voltage around, just Asus & Gigabyte.

OTOH MSI are often pretty slow about putting out bios updates.

alex314 posted:

Is 5600MT good enough for general work stuff?

yes

FuturePastNow
May 19, 2014



That's less then they spent on some of the lovely prebuilts they've reviewed

orcane
Jun 13, 2012

Fun Shoe
I want to see the video already :getin:

slidebite
Nov 6, 2005

Good egg
:colbert:

drat, I want a 7800X3D but I think I'm just going to sit back for a bit and see how this settles out.

Saturnine Aberrance
Sep 6, 2010

Creator.

Please make me flesh.


On the Asus side of things, the most recent BIOS 1413 claims to limit the voltage to 1.3 volts, isn't listed as a beta unlike the last several, and even fixes the long-standing problem I've had where 4 of the rear I/O USB ports weren't working.

Dr. Video Games 0031
Jul 17, 2004

Just be aware that none of the recent beta bioses and poo poo actually have the new updates from AMD. That will come in the form of AGESA 1.0.0.7, while everyone's still using 1.0.0.6. Motherboard companies are doing quick and dirty VSOC limits of 1.3 while they get bioses with AGESA 1.0.0.7 in order.

Fabulousity
Dec 29, 2008

Number One I order you to take a number two.

Is VSOC something that can be directly limited until the newest AGESA goes out? Does leaving EXPO turned off guarantee that the issue won't happen?

Klyith
Aug 3, 2007

GBS Pledge Week

Fabulousity posted:

Is VSOC something that can be directly limited until the newest AGESA goes out?

Yes, in the advanced / overclocking section of your bios you can manually set a voltage for VSOC. That should override whatever voltage is being picked by the board to support higher memory speed.

Fabulousity posted:

Does leaving EXPO turned off guarantee that the issue won't happen?

As far as we know, yes.

hobbesmaster
Jan 28, 2008

Fabulousity posted:

Is VSOC something that can be directly limited until the newest AGESA goes out? Does leaving EXPO turned off guarantee that the issue won't happen?

Hopefully GN Steve posts a video with his findings soon, tweets indicate it should be good and thorough on all this complete with destructive testing:

https://twitter.com/gamersnexus/status/1651308183745331202?s=46&t=TBi_iSImUmzjTxXAKsMEHw

https://twitter.com/gamersnexus/status/1651854854862635008?s=46&t=TBi_iSImUmzjTxXAKsMEHw

https://twitter.com/gamersnexus/status/1651856839384920065?s=46&t=TBi_iSImUmzjTxXAKsMEHw

Unless they report something off the wall setting everything manually should work. Silly settings with voltages when translating XMP was a problem on am4 too but never quite so dramatically destructive. The solution on AM4 was always just to set the primary timings and frequency to match the XMP/EXPO profile by hand.

runaway dog
Dec 11, 2005

I rarely go into the field, motherfucker.

Dr. Video Games 0031 posted:

Just be aware that none of the recent beta bioses and poo poo actually have the new updates from AMD. That will come in the form of AGESA 1.0.0.7, while everyone's still using 1.0.0.6. Motherboard companies are doing quick and dirty VSOC limits of 1.3 while they get bioses with AGESA 1.0.0.7 in order.

screw it then, just gonna take it back to stock. again heh

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!
Igor's been yapping a short bit about the release notes of the newest AGESA. In the screenshot of them, this stood out:



PROCHOT means Processor Hot. Seems some thermal management was FUBAR, too.

karoshi
Nov 4, 2008

"Can somebody mspaint eyes on the steaming packages? TIA" yeah well fuck you too buddy, this is the best you're gonna get. Is this even "work-safe"? Let's find out!

Combat Pretzel posted:

Igor's been yapping a short bit about the release notes of the newest AGESA. In the screenshot of them, this stood out:



PROCHOT means Processor Hot. Seems some thermal management was FUBAR, too.

Isn't PROCHOT the hottest point where the chip should start non-negotiable throttling? Before it destroys itself like in those 20 year old videos comparing AMD and intel?

Dr. Video Games 0031
Jul 17, 2004

love when my critical overheating protections have no effect

BlankSystemDaemon
Mar 13, 2009



Fabulousity posted:

Is VSOC something that can be directly limited until the newest AGESA goes out? Does leaving EXPO turned off guarantee that the issue won't happen?
From what I understand, the answers are yes and yes.

EDIT: Beaten because i left the tap open and didn't refresh.

runaway dog posted:

screw it then, just gonna take it back to stock. again heh
Yeah, that's my plan if it hasn't been fixed by the time I get my GPU, SSD and monitor.

They should get here by the 10th of May, according to the retailer.

BlankSystemDaemon fucked around with this message at 19:09 on Apr 28, 2023

sauer kraut
Oct 2, 2004

Fabulousity posted:

I got a half-built computer sitting on my dining room table that will be a 7800X3D on a MSI B650 Tomahawk. Still waiting on the M2 for the main drive so I have it just sitting there partially wired up but all these posts have me just going :catstare:

I figured I'd have to do a BIOS update before attempting to enable EXPO anything but MSI's latest update is dated April 14th and is marked as beta. AMD's official AGESA update to deal with the X3D problem was released on 4/27. Is MSI not having problems or just not caring so far?

What memory did you buy?
If it's 6000 I'd consider returning it for a pack of 5600s that are 100% supported by your mobo.

sauer kraut fucked around with this message at 19:45 on Apr 28, 2023

BlankSystemDaemon
Mar 13, 2009



sauer kraut posted:

What memory did you buy?
If it's 6000 I'd consider returning it for a pack of 5600s that are 100% supported by your mobo.
Nah:
https://twitter.com/SquashBionic/status/1565012130746998784

sauer kraut
Oct 2, 2004
I have no clue what that those twitter pics from last August are supposed to contribute here.

hobbesmaster
Jan 28, 2008

sauer kraut posted:

I have no clue what that those twitter pics from last August are supposed to contribute here.

The reason you’d want to crank VSoC is so you can run a higher FCLK. If FCLK has no great impact then leave it on auto and turn down the VSoC voltage.

sauer kraut
Oct 2, 2004
That is stuff for people who tinker with expert settings. The goon with the half-done build presumably just wants to switch on Expo and save&exit, like many other people.

Klyith
Aug 3, 2007

GBS Pledge Week
I still wouldn't return the 6000 ram even if you didn't want to do anything manually. By the time your RMA is processed and your new ram arrives, there will probably be bios updates to fix things.

Just don't press the button and run it stock for a few days.

Fabulousity
Dec 29, 2008

Number One I order you to take a number two.

sauer kraut posted:

That is stuff for people who tinker with expert settings. The goon with the half-done build presumably just wants to switch on Expo and save&exit, like many other people.

G.Skill Flare X5 Series (AMD Expo) 32GB (2 x 16GB) 288-Pin SDRAM DDR5 6000 CL30-38-38-96

Yes I would like to PRES BUTAN GET PERFORMZ but I'm okay sitting around for a bit until a good BIOS gets shat out by MSI. Worst case I run it and get my hands in there and mess with everything manually.

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!

hobbesmaster posted:

The reason you’d want to crank VSoC is so you can run a higher FCLK. If FCLK has no great impact then leave it on auto and turn down the VSoC voltage.
I have no idea what they're talking about. 2000MHz FCLK supposedly hard to achieve or something?

hobbesmaster
Jan 28, 2008

Combat Pretzel posted:

I have no idea what they're talking about. 2000MHz FCLK supposedly hard to achieve or something?

Im just saying that’s what the tweet says :shrug:

EoRaptor
Sep 13, 2003

by Fluffdaddy
Just an FYI about Gigabyte: They publish beta BIOS's as lettered, eg: F7b or F7g are considered 'beta'. However, when they are about to roll out a final version, they will pull all the beta versions from the website. Usually, in 5 to 10 days, the final will arrive. That may not fully apply right now, as they may pull BIOS's for safety reasons, but it's been the historical pattern when beta BIOS's suddenly vanish.

Cygni
Nov 12, 2005

raring to post

As a quick aside, someone put together a roadmap based on the latest DigiTimes article:

https://twitter.com/harukaze5719/status/1651816556874317824

If this ends up happening, there is a theoretical world where Intel actually regains the node advantage here in the CPU market. Wild times.

hobbesmaster
Jan 28, 2008

Cygni posted:

As a quick aside, someone put together a roadmap based on the latest DigiTimes article:

https://twitter.com/harukaze5719/status/1651816556874317824

If this ends up happening, there is a theoretical world where Intel actually regains the node advantage here in the CPU market. Wild times.

The rumors on the specs for xeons that would go on 20A would appear to indicate that Intel won’t be doing anything with that advantage.

Desuwa
Jun 2, 2011

I'm telling my mommy. That pubbie doesn't do video games right!
Asus at least are rolling out another round of BIOS updates that extend the 1.3v limit to non -3d cpus.

BlankSystemDaemon
Mar 13, 2009



sauer kraut posted:

I have no clue what that those twitter pics from last August are supposed to contribute here.
If it's not clear, the tweet is talking about the FCLK to UCLK to MCLK ratio; ie. frequency of Infinity Fabric Clock to Unified Memory Controller Clock to Memory Clock.

Leaving FCLK to auto means it'll default to 1733MHz, and having the others be 1:1 effectively means that since DDR5-6000 runs at 3000MHz, you need to set the UCLK to 3000MHz too.

We don't yet have any real idea of how many chips that come out of the factory being able to run at 2000MHz FCLK.
From what I've seen of some synthetic benchmarks, it does give slightly better performance - though the difference in gaming was much less pronounced, and might well fall within the margin of error for individual run variance.

Most of this is also covered by the Raphael overclocking guide that Skatterbench did.

The take-away is that it's dumb to return the memory, because even if you don't want to enable EXPO until it's been known to work, the memory runs fine at lower speeds.

Combat Pretzel posted:

I have no idea what they're talking about. 2000MHz FCLK supposedly hard to achieve or something?

hobbesmaster posted:

Im just saying that’s what the tweet says :shrug:
The tweet is from a fair while ago, but I think it's still too early to say - but whether it's going to mean something for gaming is another matter entirely.
And if you have a 7800X3D, you're probably primarily interested in gaming, so it's probably not going to matter.

Dr. Video Games 0031
Jul 17, 2004

My motherboard set my fclk to 2000 automatically when i enabled expo and it just worked :shrug:

BlankSystemDaemon
Mar 13, 2009



Dr. Video Games 0031 posted:

My motherboard set my fclk to 2000 automatically when i enabled expo and it just worked :shrug:
Well, if Robert turns out to have been wrong, that's great - but it's probably also a question of the specific motherboard and/or revision and/or firmware.

PC LOAD LETTER
May 23, 2005
WTF?!
That guys twitter comments are from near launch late last year. Getting FCLK to 2000 seems pretty easy right now. Going much over that is hard. Mine won't even get to 2100. 2000 is zero effort though. I think BZ has said he struggles to get to 2133.

He's also said though that going over 2033 usually isn't worth it unless you're trying to absolutely squeeze every last little bit out of the CPU so I wouldn't worry about it for daily driver use. Particularly if you're just wanting to do a quick XMP/EXPO set it and forget it OC.

Palladium
May 8, 2012

Very Good
✔️✔️✔️✔️

hobbesmaster posted:

Unless they report something off the wall setting everything manually should work. Silly settings with voltages when translating XMP was a problem on am4 too but never quite so dramatically destructive. The solution on AM4 was always just to set the primary timings and frequency to match the XMP/EXPO profile by hand.

i never used xmp at all when pushing ddr4 OCs

auto ocing "standards" pushing unsafe voltages was a problem at least since the Intel 8th gen days, but the vendors just cant stop doing it

hobbesmaster
Jan 28, 2008

hobbesmaster posted:

Hopefully GN Steve posts a video with his findings soon, tweets indicate it should be good and thorough on all this complete with destructive testing:

https://twitter.com/gamersnexus/status/1651308183745331202?s=46&t=TBi_iSImUmzjTxXAKsMEHw

https://twitter.com/gamersnexus/status/1651854854862635008?s=46&t=TBi_iSImUmzjTxXAKsMEHw

https://twitter.com/gamersnexus/status/1651856839384920065?s=46&t=TBi_iSImUmzjTxXAKsMEHw

Unless they report something off the wall setting everything manually should work. Silly settings with voltages when translating XMP was a problem on am4 too but never quite so dramatically destructive. The solution on AM4 was always just to set the primary timings and frequency to match the XMP/EXPO profile by hand.

https://twitter.com/gamersnexus/status/1652098512706838530?s=46&t=TBi_iSImUmzjTxXAKsMEHw

https://twitter.com/gamersnexus/status/1652098996188377093?s=46&t=TBi_iSImUmzjTxXAKsMEHw

Well, i guess keeping it at stock speeds is the way to go for the moment

wargames
Mar 16, 2008

official yospos cat censor

Fabulousity posted:

G.Skill Flare X5 Series (AMD Expo) 32GB (2 x 16GB) 288-Pin SDRAM DDR5 6000 CL30-38-38-96

Yes I would like to PRES BUTAN GET PERFORMZ but I'm okay sitting around for a bit until a good BIOS gets shat out by MSI. Worst case I run it and get my hands in there and mess with everything manually.

what bios are you running, because with f5b in the gigabyte b650 my memory speeds with the 36cl kits is wayyyyyyyy down.

Fabulousity
Dec 29, 2008

Number One I order you to take a number two.

wargames posted:

what bios are you running, because with f5b in the gigabyte b650 my memory speeds with the 36cl kits is wayyyyyyyy down.

Not running anything yet: the system is sitting on a table with cables splayed everywhere while I wait to see what GN says.

In the meantime I can say the Lian Li 216 case is very nice to work with and feels like cheat mode building.

Edit: MSI just popped 7D75v162(Beta version) for the Tomahawk board and they say:

MSI posted:

Description:
- By following the mitigation from AMD, fix CPU SoC voltage upper limit for Ryzen 7000X3D and non-X3D series CPU,
which might affect the performance of certain EXPO memory modules.
- Support 48/24GB high density DDR5 memory module.

:thunk:

Fabulousity fucked around with this message at 06:50 on Apr 29, 2023

Palladium
May 8, 2012

Very Good
✔️✔️✔️✔️

Fabulousity posted:

In the meantime I can say the Lian Li 216 case is very nice to work with and feels like cheat mode building

i see u are new to this ODD-less form factor

Arzachel
May 12, 2012

Cygni posted:

As a quick aside, someone put together a roadmap based on the latest DigiTimes article:

https://twitter.com/harukaze5719/status/1651816556874317824

If this ends up happening, there is a theoretical world where Intel actually regains the node advantage here in the CPU market. Wild times.

Oh, they're finally doing a proper chiplet APU?

kliras
Mar 27, 2021
for all the old goons who don't see community updates through the mobile youtube app:

Gamers Nexus posted:

We have been able to reproduce a catastrophic failure resulting in the motherboard self-immolating while we were running external current logging, thermography, and direct VSOC leads to a DMM. The issue involves incompetence on many levels. Video script being finalized now. This video has taken us all week to do. We stepped off a plane from AMD HQ (unrelated, but convenient timing) on Tuesday and immediately set to work. It has been a time-consuming, expensive, physically exhausting process and we've loved every minute of it. The team is thrilled to have a real result and successful root-cause analysis that goes layers beyond "it's SOC voltage," because it's not that simple. We appreciate your support in viewing the video when it launches, because that's the real reward for our tired team. Keep an eye out and share it wide. Thank you!

Adbot
ADBOT LOVES YOU

sauer kraut
Oct 2, 2004

Arzachel posted:

Oh, they're finally doing a proper chiplet APU?

The monolithic APUs are the 'proper' Zen processors (sadly crippled by low # of PCIe lanes), why would they make a crappy chiplet one that sucks 25W+ in idle spinning plates?
For bottom shelf prebuilt office PCs maybe? No way any laptop vendor would touch that.

sauer kraut fucked around with this message at 15:46 on Apr 29, 2023

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply