Question Zen 4 builders thread

[Markfw]

I know its early, but as someone pointed out, I better get my ram now before the scalpers are all over it.

To go with a 7950x, what do you think of this ram:

https://www.newegg.com/g-skill-32gb-288-pin-ddr5-sdram/p/N82E16820374369?Item=N82E16820374369

(edited for my best recommendation, 12/5/2023)

 

[Hail The Brain Slug]

I wouldn't rule out the negative CO as the cause.

Yeahhhhhhhhhhhhhhhhhhhhhhh
My 7950X3D won't take more than -10 or -15 before it will cause random reboots. The thing is run extremely thin on voltage out of the box.

I thought I could manage -30 and I would get random reboots gaming.

 

[Markfw]

I wouldn't rule out the negative CO as the cause.

I have FIVE 7950x. 4 of them use an ASRock motherboard. All FIVE use the same 6000 CL30 memory. ALL FIVE user PBO CO -25. Only the ASUS has been a problem until now.

 

[Hail The Brain Slug]

I have FIVE 7950x. 4 of them use an ASRock motherboard. All FIVE use the same 6000 CL30 memory. ALL FIVE user PBO CO -25. Only the ASUS has been a problem until now.

Yes because all silicon is identical when run outside of manufacturer spec.

 

[Markfw]

Yes because all silicon is identical when run outside of manufacturer spec.

EXPO was created by AMD to run outside JEDEC standards on AMD Zen 4 processors. I would not call that "outside of manufacturer spec".

I was trying to make a point that the BIOS is the first good one from ASUS since I got this motherboard.

 

[CakeMonster]

My 7950X can't handle any negative CO (well at least not -10) which made it pretty annoying when reading those overconfident OC guides 'saying all CPU's can do AT LEAST -X' (often a value well below -10)).

I just limit the temperature to 85C to reduce fan activity and its rock solid regardless so I'm not complaining.

 

[Markfw]

My 7950X can't handle any negative CO (well at least not -10) which made it pretty annoying when reading those overconfident OC guides 'saying all CPU's can do AT LEAST -X' (often a value well below -10)).

I just limit the temperature to 85C to reduce fan activity and its rock solid regardless so I'm not complaining.

What motherboard to you have ? Based on my experience the motherboard/BIOS has been the main factor with mine. I mean, all 5 of mine are running 135-142 watts at PBO CO= -25 as recommended by an article way up in this thread. 5 is not anything to prove that this should be fine most of the time, but it sure makes a case for it.

 

[CakeMonster]

I have the Asus Hero. It will 'run' with CO but crash eventually (more with bigger values) like Justinus describes above. Yeah I don't doubt that most CPU's might handle it, but I'm just slightly annoyed by several big sites and YT channels actually recommend starting with even bigger values, which might lead people to think something else is wrong.

 

[Markfw]

I have the Asus Hero. It will 'run' with CO but crash eventually (more with bigger values) like Justinus describes above. Yeah I don't doubt that most CPU's might handle it, but I'm just slightly annoyed by several big sites and YT channels actually recommend starting with even bigger values, which might lead people to think something else is wrong.

OK, I think you just proved my point. What bios do you have ?? the one a few days ago ? Thats the first one for me thats been stable.

Do you have this one ? It came out just yesterday, I recommend you try it.

ROG MAXIMUS Z690 HERO BIOS 2703

Version 2703
13.09 MB 2023/08/17

 

[CakeMonster]

I tried CO fairly early on, probably one of the 08xx BIOSes (I have the earlier ones I've used on a flash drive, so 0801 or 0805). Then I tried it again after one of the early fixes for the VSOC problem, so I'm guessing 1401, but might have been a later one. I haven't attempted since that. Running 1602 beta BIOS uploaded to the ROG forum right now. If I have the time during the weekend I might give it another attempt.

 

[Markfw]

I tried CO fairly early on, probably one of the 08xx BIOSes (I have the earlier ones I've used on a flash drive, so 0801 or 0805). Then I tried it again after one of the early fixes for the VSOC problem, so I'm guessing 1401, but might have been a later one. I haven't attempted since that. Running 1602 beta BIOS uploaded to the ROG forum right now. If I have the time during the weekend I might give it another attempt.

I would be interested in your results, as I am very happy with this latest. None before have been worth anything. ASUS=Late to the game.

 

[AdamK47]

I have the Asus Hero. It will 'run' with CO but crash eventually (more with bigger values) like Justinus describes above. Yeah I don't doubt that most CPU's might handle it, but I'm just slightly annoyed by several big sites and YT channels actually recommend starting with even bigger values, which might lead people to think something else is wrong.

Keep in mind that Mark runs one single continuous workload. Folding at Home. If he were using the same settings under gaming scenarios where load levels bounce around, I have no doubt instability from one or more of his CPUs would pop up.

 

[Markfw]

Keep in mind that Mark runs one single continuous workload. Folding at Home. If he were using the same settings under gaming scenarios where load levels bounce around, I have no doubt instability from one or more of his CPUs would pop up.

Actually... NO. I run many DC projects, and the three I run at the same time on this box, are Rosetta@home, WCG (both on CPU) and folding@home on GPU.

 

[Hotrod2go]

Yeah since ~1 year ago, how come ?

Except AM5 wasn't released then... Much discussion here but hardly any results at all. Besides its necessary to see results with latest bios & AGESA updates.

 

[gdansk]

My 7950X doesn't tolerate any negative offset either. Board is an ASRock X670E Steel Legend.

I think it's simply luck of the draw and sometimes you get a chip which shows the reason why AMD shipped with the voltage they did.

Even if >90% can handle at least -10 for limited workloads I'm sure it is less when a wide range of workloads need to be stable 365 days a year.

 

[AdamK47]

Actually... NO. I run many DC projects, and the three I run at the same time on this box, are Rosetta@home, WCG (both on CPU) and folding@home on GPU.

Similar continuous load. Unstable CO makes itself known with varied workloads. Put simply, all core -25 CO will not work on every CPU.

 

[Det0x]

Has anyone had success with 6600MHz RAM + FCLK=MCLK, stable in windows yet?

Edit: typo, mean't UCLK=MCLK.

Yeah since ~1 year ago, how come ?

Except AM5 wasn't released then... Much discussion here but hardly any results at all. Besides its necessary to see results with latest bios & AGESA updates.

Yeah seems i missed by a quarter, have to blame it on faulty memory (), but from this very thread 8 months ago i wrote the following:

Will need some time to fully stabilize this, but finally broke 100k mb/s writespeed
FCLK @ 2200mhz
MCLK @ 6600MT/s
View attachment 73142

If your new to the scene, i'm domdtxdissar from overclock.net forum your linking to..

Haven't shared any results lately because i dont want people to copy settings before the HWBOT Team CUP 2023 - AMD ends in 25 days, but i can give a sneak peak on what i'm currently working on: (stable settings which completed 22k karhu + few hours y-cruncher all tests enabled)

Some of my previous high memoryspeed results can be found here:

• 8000MT/s 7950X3D

Domdtxdissar 7950x3d 8000MT/s 2200mhz FCLK CL 34 T1 GDM disabled 02.08.2023

Discover the magic of the internet at Imgur, a community powered entertainment destination. Lift your spirits with funny jokes, trending memes, entertaining gifs, inspiring stories, viral videos, and so much more from users.

imgur.com

• 6600MT/s 7950X

https://imgur.com/a/J74HIN0

• 6600MT/s 7950X3D

https://imgur.com/a/gULmCbU

Spoiler: 6600MT/s 7800X3D

Is there anything in particular you are wondering about in regards to running 6600MT/s on Zen4 ?

 

[Hotrod2go]

Yeah seems i missed by a quarter, have to blame it on faulty memory (), but from this very thread 8 months ago i wrote the following:

If your new to the scene, i'm domdtxdissar from overclock.net forum your linking to..

Haven't shared any results lately because i dont want people to copy settings before the HWBOT Team CUP 2023 - AMD ends in 25 days, but i can give a sneak peak on what i'm currently working on: (stable settings which completed 22k karhu + few hours y-cruncher all tests enabled)
View attachment 84637
Some of my previous high memoryspeed results can be found here:

• 8000MT/s 7950X3D

Domdtxdissar 7950x3d 8000MT/s 2200mhz FCLK CL 34 T1 GDM disabled 02.08.2023

Discover the magic of the internet at Imgur, a community powered entertainment destination. Lift your spirits with funny jokes, trending memes, entertaining gifs, inspiring stories, viral videos, and so much more from users.

imgur.com

• 6600MT/s 7950X

https://imgur.com/a/J74HIN0

• 6600MT/s 7950X3D

https://imgur.com/a/gULmCbU

Spoiler: 6600MT/s 7800X3D

View attachment 84639

Is there anything in particular you are wondering about in regards to running 6600MT/s on Zen4 ?

Thanks, appreciate the effort you went to with all this. But 1.375v on the VSOC? is that a reliable reading from Zen timings? Only asking cause' AMD's recommendation to stay under 1.30v.

 

[JoeRambo]

1.375v on the VSOC? is that a reliable reading from Zen timings?

It's in HWInfo64 as well in MB sensor section, so it's just what it takes to run mem this tight on Z4.

 

[Markfw]

OK, to all of you that said many CPUs can not do -25, you are right, and I was wrong. This sucker started rebooting every hour or so. I got so mad I reset everything to bios default. Now here is the interesting part. Its running faster (5.2 ghz) but using less power (115 watts or so). The memory speed sucks, but with the added cpu speed, its fine. For now I am just going to let it run. Below is what it is doing at 100% load on all 32 threads:

 

[Hotrod2go]

It's in HWInfo64 as well in MB sensor section, so it's just what it takes to run mem this tight on Z4.

Silicon lottery plays a part too.

 

[Det0x]

Thanks, appreciate the effort you went to with all this. But 1.375v on the VSOC? is that a reliable reading from Zen timings? Only asking cause' AMD's recommendation to stay under 1.30v.

Your looking at the oldest 6600MT/s results?
Those screenshot were taken before it was known higher than 1.3vsoc were killing some X3D cpus

It's in HWInfo64 as well in MB sensor section, so it's just what it takes to run mem this tight on Z4.

My later 6600MT/s + 2200FCLK and 8000MT/s + 2200FCLK results have been done at 1.3vsoc, so no, 1.4vsoc is not needed to run mem this tight on Zen4. (when it get colder in weather i even plan to run 6800MT/s 1:1 1.3vsoc for some benches)

[Official] Zen 4 X3D Owner's Club (7800x3D /...

This thread was created to serve as a meeting place to share thoughts, speculate, and theory-craft on the upcoming AMD Zen 4 Raphael-X 7XXXx3D chips: The Ryzen 7 7800X3D, Ryzen 9 7900X3D, and Ryzen 9 7950X3d. After launch, this thread will serve as a place for owners to share experience...

www.overclock.net

If anything, 2:1 results last few months have shown lower is better in regard to some voltages, especially CLDO VDDP

 

[Det0x]

Silicon lottery plays a part too.

100% correct, your average part will probably not do 6600/2200, but with latest agesa i think 6400/2133 can be a realistic target for most.
(SMU v84.79.22x increased the MT/s limit by almost 200 on average i would say)

 

[JoeRambo]

Its running faster (5.2 ghz) but using less power (115 watts or so). The memory speed sucks, but with the added cpu speed, its fine.

Depends on workload, but assuming You are running FP heavy DC stuff, they scale real well with memory speed. Poster child is Linpack type workload where one can vary Gflops by 25% just be changing memory speed and timings.
So even if wattage is less, it simply might mean less work done. Less work = less wattage = less amps = more clock.

 

[CP5670]

Similar continuous load. Unstable CO makes itself known with varied workloads. Put simply, all core -25 CO will not work on every CPU.

I find that -20 CO works fine on any game or Prime95, but literally any CO offset crashes in idle eventually when the PC is not doing anything. I just keep it at stock, the CPU does 4.75ghz or so in most games.

 

[Markfw]

Well, another dead Corsair PSU. RM1000, dead as a doornail. So one of my 7950x's is out until I get it replaced.