Continuing To Stress Ryzen

[vein]

As I already posted in the other thread, I used to see this bug with the original BIOS while doing kernel compilation and video encoding. However, since the last BIOS update (that contains AGESA 1.0.0.6a), I do not see it anymore. This is on Prime X370-Pro from ASUS w/ R7 1800X. Just for the hell of it I ran the test today for hours and I cannot make it fail. Perhaps this is buggy BIOS on some motherboards or perhaps something with voltages?

Hmm.. Interesting... My motherboard does not have a bios update with AGESA 1.0.0.6a out yet so i cannot try this... Can someone else try it out?

(If it has already been done, then disrecard this message... It is just hard to see in this massive thread)

[leipero]

wrong.

Have you ever thought about the following: "Hey Mr. Nadella, there are a few oddities in our CPU that could lead to undesired behaviour, but you can circumvent them by applying following workarounds in the Windows kernel: [funny list of assembler commands] BUT, Mr. Nadella, don't ever tell anyone, okay?!"

Few things before I respond to you, you created account recently with interesting name..., all of your posts are Ryzen related as far as I can see. So let me make some theory, you are paid by Intel to be some sort of marketing troll, since TR does go out soon, Intel is quite desparate and goes back to their own dirty tactics instead of fair competition. Isn't that fair assumption? It makes perfect sense, you made account just recently and all your posts are Ryzen related.

See what I did there? Ofc. it makes perfect sense for AMD to create conspiracy with Microsoft and protect their faulty product by paying Microsoft (who have more money than AMD...) and it's CEO who would be perfectly fine to directly involve himself in such fishy deal because if that comes out in public it would not ruin both his reputation and reputation of his product. And ofc. AMD isn't aware of free or open source software available (like Linux, BSD or whatever else exist out there...), and since it is very hard to do actual research of most popular server operating systems, AMD with it's tiny budget couldn't do such lenghty research, so being oblivious to the fact that there's other operating systems out there, AMD did begginers mistake and paid Microsoft to keep hush for their clearly hardware bug....

How could I be that stupid not to figure it out earlier, thank you my brother .

[sdack]

and because they (AMD) haven't - and if they will do, they should have done that since day zero - I assume that this new CPU is acting like the CPUs before. And because it doesn't, I'm "bitching" (to quote my buddy) around. It bugs me immensly when an 8 hour-"libreoffice" compilation fails in the last steps because there is a "bus error" of some kind and in the next run it builds without hiccups...

Why don't you take the microcode of an Intel Xeon 2678v3 and load it up into your Ryzen CPU? And don't ask why, just do it.

You can run the latest Nvidia driver on any AMD graphics card, as long it supports OpenGL 4.0. It's all instructions underneath and 1 + 1 equals 2 on every CPU.

...

[RyzenNewbie]

Few things before I respond to you, you created account recently with interesting name..., all of your posts are Ryzen related as far as I can see.

yep, as I have troubles with Ryzen, that name looked suitable to me. Could have been any other name, but I thought it would be the most neutral in that matter. "RyzenSucks" would have not been so fitting.

So let me make some theory, you are paid by Intel to be some sort of marketing troll, since TR does go out soon, Intel is quite desparate and goes back to their own dirty tactics instead of fair competition. Isn't that fair assumption? It makes perfect sense, you made account just recently and all your posts are Ryzen related.

I love your theory because then I'd be probably richer than I am now - but *yawn* let me quote myself - again:

"(...) And yes, I'm an Intel zealot because I'm the only guy here in our company that uses AMD processors (Athlon X2, Phenom II X6 1100T, Ryzen 1700) in his host and always fruitlessly promotes to colleagues/staff using AMD processors in workstations and servers."

source: https://community.amd.com/message/28...omment-2813834

And, for the case you haven't got it - that's somewhat sarcastic. And at the moment, I have to justify myself to my superiour why I wanted that new Ryzen system for around 1k$ that makes me so much trouble.

See what I did there? Ofc. it makes perfect sense for AMD to create conspiracy with Microsoft and protect their faulty product by paying Microsoft (who have more money than AMD...) and it's CEO who would be perfectly fine to directly involve himself in such fishy deal because if that comes out in public it would not ruin both his reputation and reputation of his product.

why not? It seems that the CPU is primarly advertised for being a gaming CPU (under Windows) but not labeled as one. Why shouldn't there be any cool workarounds in Windows that are yet to be included in Linux and other OSes?

And ofc. AMD isn't aware of free or open source software available (like Linux, BSD or whatever else exist out there...), and since it is very hard to do actual research of most popular server operating systems, AMD with it's tiny budget couldn't do such lenghty research, so being oblivious to the fact that there's other operating systems out there, AMD did begginers mistake and paid Microsoft to keep hush for their clearly hardware bug....

I think that AMD indeed is aware of other open-source operating systems; but tends to - let's say - takes care for them last. Nothing wrong with that, but at least a little statement would be nice - may it be positive or negative.

How could I be that stupid not to figure it out earlier, thank you my brother .

you're welcome, brother. Friends for lifetime?

[RyzenNewbie]

Why don't you take the microcode of an Intel Xeon 2678v3 and load it up into your Ryzen CPU? And don't ask why, just do it.

You can run the latest Nvidia driver on any AMD graphics card, as long it supports OpenGL 4.0. It's all instructions underneath and 1 + 1 equals 2 on every CPU.

I really don't care what a CPU or GPU sees inside itself - but I do care what I, as a user, see outside the CPU/GPU - or should I say, its behvaiour. If you find a way to push foreign microcode to any CPUs/GPUs internally in order to make them work externally, I would be very happy and call you pro.

any smilies postfixed to any statements do not make them less meaningful...

[sdack]

I really don't care what a CPU or GPU sees inside itself - but I do care what I, as a user, see outside the CPU/GPU - or should I say, its behvaiour. If you find a way to push foreign microcode to any CPUs/GPUs internally in order to make them work externally, I would be very happy and call you pro.

any smilies postfixed to any statements do not make them less meaningful...

I have some magic beans here I'd like to sell to you.

[nightmarex]

And that this damage happens is totally AMD's fault. I told them in the community forum thread that AMD urgently needs to lead this discussion. But AMD chose not to make public statements but instead to remain silent for months. Now reports are spiralling out of control and are confusing users. Someone wrote that they asked Kyle from HardOCP to run the Ryzen stress test on his Threadripper review samples. Is Kyle informed about the conftest problem? If not, shit's gonna hit the fan soon.

This is what I told Bridgman about. You can't ignore people having problems and that's the worst thing. A few simple reassurances would of kept this from being a lightning rod, but no. AMD_MATT jumped ship on these people.

Windows is much slower than Linux and *BSD and maybe it just doesn't hit these bugs.

Especially in MT I wondered that at some point if you look at... one second..

https://browser.geekbench.com/v4/cpu/3566292

That's my geekbench 4 run, you'll see systems clocked above that, however no windows systems get close to that MT yet somehow some are far above in ST so... Yeah maybe Linux just fires off many more thread hits that windows doesn't that exposes problems.

[carewolf]

I was part of the class-action lawsuit against Intel. Floating point was broken and you had to use software emulation under Winblows to prevent lockups. Took over a year but they finally replaced my CPU.

I hope this gets sorted out as I've already decided my next laptop will be a “Raven Ridge": http://wccftech.com/amd-raven-ridge-...phics-spotted/

It made a wrong rounding in a specific rare case. It wasn't even less accurate, just not following the standard.

Btw, compilers could work around it. If they haven't removed it recently gcc still has the option to compensate for the bug if you need fully standard complient floating point on an effected Pentium.

[RyzenNewbie]

(...)

don't worry, I haven't forgotten you - my answer currently is in moderation queue because I was so bold to include an URL...

[Brutalix]

Ok.. the tone in this thread seems to be a bit heated. Please keep the tone civil everyone. This does not become a linux forum.

Since my premature post yesterday, 2 of my computers, one 1700x and one 1600x both with ECC ram, 1 ubuntu and one debian testing, both have been running a 26 hour cycle test. One was running the Kill Ryzen script, and one the ryzen_segv_test-master script from the BSD error discussion. This last ran on the 1600x with: run.sh 12 250000. So far I have a 350 Mb log, and no seg. errors. Have not tested my 1700 that runs with non-ECC ram, but I will try that as well. So since I only got conftest seg error on all computers previously I can understand that AMD has trouble replicating this error.

Kind regards
Brut.