Der8auer: Only Small Percentage of 3rd Gen Ryzen CPUs Hit Their Advertised Speeds

[TheLostSwede]

I'm not convinced there is a real firmware fix coming. Some minor tweaks perhaps but I suspect the announcement is just damage control to shut up the vocal minority who are making a big deal about this.

There's enough of a spread in the Der8auer survey results to show a clear bell-curve of results implying that this isn't a firmware limitation but simply the spread of results from the silicon lottery. The peak of the bell curve is typically 25-50MHz lower than AMD's figures and if the survey data is realistic then AMD either miscalculated slightly or rounded up the figures to the nearest 0.1GHz.

It's still comical that this topic has even come up, firstly because Intel's CPUs have arbitrary time-limits to their boost, after which they slow down again far more than Zen2 chips do, and secondly because the number of situations where only one core is active in a modern machine is zero. The only people who care about this "peak single-core boost frequency" aren't people who are actually using the chips to do stuff. The minute you give any multi-core CPU a real-world workload, the OS scheduler is going to use all available cores to run background tasks, meaning that 'single core' is never achieved.

Hell, the monitoring software uses a core to monitor the single-threaded synthetic load, thus using a second core. It's so dumb that the only people left arguing it seriously are just in it for the arguing, not actually giving a damn about the topic at all

So how do you explain that some of us have already had the problem resolved courtesy of an updated UEFI/AGESA? I was as I've explained time and time again in this thread, a hard upper clock limit of 4,400MHz until recently. Now my CPU boosts to 4,525MHz no problem. But hey, I'm just making that up, right? As it's easier to make crap up, like you...

Oh and it's also on AMD's official Twitter account now.

https://twitter.com/i/web/status/1168901636162539536

 

[EarthDog]

I would like to see how Intel i7-6950X compete with new R9-3900X.
Because we talk about 4 years old CPU with lower frequency it's logic to OC both to the maximum and then to compare them.
That mean i7-6950X 4.4GHz boost on all cores, 3.8-4.0GHz Cache frequency vs R9-3900X on how much is boost...
no one know that for sure, enthusiasts community still examine is it boost as AMD advertised.

That has absolutely nothing to do with this thread.

 

[Chrispy_]

So how do you explain that some of us have already had the problem resolved courtesy of an updated UEFI/AGESA? I was as I've explained time and time again in this thread, a hard upper clock limit of 4,400MHz until recently. Now my CPU boosts to 4,525MHz no problem. But hey, I'm just making that up, right? As it's easier to make crap up, like you...

Oh and it's also on AMD's official Twitter account now.

https://twitter.com/i/web/status/1168901636162539536

You had a nearest 100MHz round-numbered hard upper clock limit. That's clearly not your precision boost fluctuating in 25MHz steps as it's supposed to, that was your board vendor screwing up their UEFI implementation of boost altogether.

Hundreds of results in the video are in the 25/50/75MHz increments which means that precision boost IS working properly, but they're topping out at that level. Your CPU now tops out at 4525 which makes it one of these results, as expected.

Congratulations you got a chip that is slightly better than average.

 

[R-T-B]

well then mate apply for an editorial role at tpu, you picked a topic of debate that keeps people at the site to either shitpost or take things

I was born in a different age...

silly to go down this hole of a GPU analog.

I don't agree here. They (gpu makers) started it, and share a lot of similar issues with the idea, and how they advertise their specs (appropriately, or inappropriately) is relevant.

 

[EarthDog]

I don't agree here. They (gpu makers) started it, and share a lot of similar issues with the idea, and how they advertise their specs (appropriately, or inappropriately) is relevant.

I see what you are saying, but, they list minimum boosts, essentially. And they all hit that when not banging off power/thermal limits. Here with CPUs, both Intel and AMD list the maximum boost... what they are EXPECTED to run at in 'nominal' conditions. While similar, they work in a completely opposite manner compared to GPUs. So, to me, no point in going down that road in a CPU thread about boost.

...and staff said so anyway, lol.

 

[R-T-B]

I see what you are saying, but, they list minimum boosts, essentially. And they all hit that when not banging off power/thermal limits. Here with CPUs, both Intel and AMD list the maximum boost... what they are EXPECTED to run at in 'nominal' conditions. While similar, they work in a completely opposite manner compared to GPUs. So, to me, no point in going down that road in a CPU thread about boost.

Fair point. I mean that statement alone is perhaps relevant, but not much more to discuss there now is there?

 

[EarthDog]

but not much more to discuss there now is there?

Depends on who you ask... some neophytes are still making up excuses and droning on about it, however...

See.

 

[R-T-B]

You act surprised.

 

[E-curbi]

I believe everyone here would agree on this but this is not the topic here !

The topic is AMD advertised speeds very few CPUs can hit just to be able to put a 0,1 or 0,2 bigger number on the box . That's called false advertisement and AMD doesn't need this BS .

I apologize, guess I'm not so skilled in AMD Intel argument threads.

More of an overview guy - always trying to see the brighter side of mankind's progress with amazing tech. Limits my debating skills severely.

 

[Darmok N Jalad]

Yep! See above. I added in the excerpt to my post.

I wonder what those who thought otherwise will say now? Do you think we will hear from anyone after this?

I’m sure AMD liked the timing of this boiling over—right before a three day weekend. Gave people plenty of time to sharpen the pitchforks before corporate could issue a presser!

 

[lexluthermiester]

It was straight from AMD. If the AIBs were to blame, you're damn right AMD would have said so. They didn't.

Ok, that only shifts accountability from the AIB's to AMD, sort of. Doesn't change the solution. Manually drop the voltage, problem solved. Don't believe me? Try it.

But here's a thought, one would think that the AIB's are the ones looking over the software supplied by AMD. They're presumably smart enough to know a problem when they see one, so why didn't the AIB's blow the whistle and fix it themselves? Even with AMD making that statement, the AIB's still have some level of accountability here. The fix is trivial. Laughably easy even. So why did they not do so themselves?

 

[EarthDog]

Ok, that only shifts accountability from the AIB's to AMD. Doesn't change the solution. Manually drop the voltage, problem solved. Don't believe me? Try it.

But here's a thought, one would think that the AIB's are the ones looking over the software supplied by AMD. They're prsumably smart enough to know a problem when they see one, so why didn't the AIB's blow the whistle and fix it themselves? Even with AMD making that statement, the AIB's still have some level of accountability here. The fix is trivial. Laughably easy even. So why did they not do so themselves?

"Ohhhhhhhh we're halfway there.........OhhhhhHHHHH living on a prayer! Take my hand, we'll make it I swear" - Bon Jovi

Anyway, joking aside - I've tried dropping voltage and like TLSwede, I ran into instability. That wasn't the answer for us at least. That also isn't remotely the point. OUT OF THE BOX WITH NO CHANGES, users should reach the listed clocks on the box. AMD agrees admitted as much and is doing something about it.

Let's not move the goal posts though, eh? I don't care if it was easy or difficult. The point is that AMD admitted there is a problem with coding THEY inject (AGESA) in how the CPU behaves. I recall TLSwede mentioning to you that the AGESA isn't editable by AIBs. They can make changes on top of it, but clearly, that is not the issue here or, as I said earlier, they would have said something to that effect instead of just taking one for the team. This is NOT an AIB issue, 'let it go' Elsa.

 

[lexluthermiester]

I recall TLSwede mentioning to you that the AGESA isn't editable by AIBs.

I didn't see it. After reading his imature nonsense I didn't care either. If you folks can't solve this issue, you have the problem. Sucks to be you. Have fun with that.

 

[newtekie1]

No, they didn't.

An FE 1080 for example was advertised to have a boost clock of 1733mhz but you can look at various reviews that under load it would drop well below that. There was no "maximum" just this one "boost clock". What it means, well be my guest, it's certainly not a maximum nor a minimum though. That's for sure.

View attachment 130861

In addition to that one can say AMD doesn't have full control over cooling, power delivery and whatnot but Nvidia did, they knowingly shipped cards with the sort of cooling that wouldn't support those boost clocks all the time. And don't get me wrong, AMD does the same for their GPUs. The point is no one is truthful with their boost clocks, there is always caveat, so either everyone is right or no one is.

No one cared though, because it's all about expectations not how truthful you are.

Interesting, you're correct.

 

[EarthDog]

I didn't see it. After reading his imature nonsense I didn't care either. If you folks can't solve this issue, you have the problem. Sucks to be you. Have fun with that.

Sometimes it's just easier to take your ball and go home I guess.

Interesting, you're correct.

It's hitting the temperature limit... of course it will throttle below the minimum boost. You can raise that limit or turn the fans higher if some titles manage to do so. Otherwise, it's as we said it was. That is a MINIMUM value that will always be hit UNLESS power/current/temp limits come into play. Either way, it has little to do with this thread as they work in opposite ways and the CPUs we are talking about are not hitting those limits.

 

[jaggerwild]

I hate to say I told you so, I told you so(but I did I did told you so!) They advertised 4.725GHz, no CPU hits that limit not 1...

 

[lexluthermiester]

I hate to say I told you so, I told you so(but I did I did told you so!) They advertised 4.725GHz, no CPU hits that limit not 1...

While that's a fair point, it isn't the topic here..

 

[cadaveca]

Let's not move the goal posts though, eh? I don't care if it was easy or difficult. The point is that AMD admitted there is a problem with coding THEY inject (AGESA) in how the CPU behaves. I recall TLSwede mentioning to you that the AGESA isn't editable by AIBs. They can make changes on top of it, but clearly, that is not the issue here or, as I said earlier, they would have said something to that effect instead of just taking one for the team. This is NOT an AIB issue, 'let it go' Elsa.

The AGESA update might take until the end of October. I'm going to crawl back in my hole now; take it easy till we talk again.

(I never expected this thread to blow up with 5 more pages in such a short time.)

 

[EarthDog]

The AGESA update might take until the end of October. I'm going to crawl back in my hole now; take it easy till we talk again.

(I never expected this thread to blow up with 5 more pages in such a short time.)

Rabid misinformed fanatics (on both sides) gets ya every time!

Get in your home!!!!!!

 

[cadaveca]

Rabid misinformed fanatics (on both sides) gets ya every time!

I wanna place blame at the usual places I do, actually. Those fanatics get their info from somewhere...


Now, I'm going to point out something that I said countless times when I was a reviewer... I don't see many people doing reviews using a clamp-on meter over the 8-pin, which allows you to directly see CPU power use, and also allows you to check things like power draw increases as core speeds increase, or when temps increase (yes, this still happens)… or other cores are used...

Many times when issues like this crop up, a simple look with some simple tools tells you the real picture as to what is going on, but we rarely see this in the enthusiast communities, and it pains me so... because to me, as an enthusiast overclocker, these measurements are so valuable that I don't know how anyone does it without them! With so many claims of not relying on software, so many do for this, and I don't know why.

Its been so long since we've seen true detailed analysis of hardware, and while I understand why...

 

[newtekie1]

It's hitting the temperature limit... of course it will throttle below the minimum boost. You can raise that limit or turn the fans higher if some titles manage to do so. Otherwise, it's as we said it was. That is a MINIMUM value that will always be hit UNLESS power/current/temp limits come into play. Either way, it has little to do with this thread as they work in opposite ways and the CPUs we are talking about are not hitting those limits.

That's very true, at least the nvidia cards actually hit the advertised boost clocks. They just got hot and lowered the clock speed. The AMD CPUs aren't even doing that, they just never hit the advertised boost clock.

So this very much is a different situation.

 

[Nordic]

I wanna place blame at the usual places I do, actually. Those fanatics get their info from somewhere...


Now, I'm going to point out something that I said countless times when I was a reviewer... I don't see many people doing reviews using a clamp-on meter over the 8-pin, which allows you to directly see CPU power use, and also allows you to check things like power draw increases as core speeds increase, or when temps increase (yes, this still happens)… or other cores are used...

Many times when issues like this crop up, a simple look with some simple tools tells you the real picture as to what is going on, but we rarely see this in the enthusiast communities, and it pains me so... because to me, as an enthusiast overclocker, these measurements are so valuable that I don't know how anyone does it without them! With so many claims of not relying on software, so many do for this, and I don't know why.

Its been so long since we've seen true detailed analysis of hardware, and while I understand why...

Come back then! Be the change you want to see.

Rabid misinformed fanatics (on both sides) gets ya every time!

My ignore list grows long...

 

[GreiverBlade]

Ok, now that's "mountain out of molehills"


Time to opt out (thread) and opt in (AMD next build, because Intel did far worse)


Also -7%, or something like that, ipc with a 2014 Intel CPU that can OC better than AMD's boost ( ooooh what about caring on OC and no boost enabled for AMD, sorry if I can't remember the name... Posting on mobile) before or after mitigation patch (well even with -7% pre mitigation... Price wise it's not green compared to a R5 3600/3600X, but self conviction can be hard to overcome)
And to say, when AMD had that kind of -% IPC gen for gen people's said Intel was stomping on them... I guess when the situation is reversed... Intel is still stomping on them for some people (if you didn't test and base out of personal experience.... "Zip it!" Pretty please)

Oh and nope Intel induced issues of my 6600k are not due to my VCore which never exceeded 1.35 it only happened after the microcode update they pushed via WUpdate.

Funny I re attempted my standards OC (meager 4.4) well... Indeed I saw a lot of blue ( and not the Intel's one)

 

[Dave65]

Do we really need another thread for this?

Yeah, we do!

 

[xdregox]

AMD made an error should have shipped stock coolers on high switch position instead of low . . .Boost gets dam close my 3800x hits 4.575 ghz. On low not a chance.we would then complain about noise hope they don't ship 3950x with that cooler or switch it to high just set silent in bios.