Ryzen 3000 boost clock controversy - der8auer publishes his survey results, not a good look for AMD

That’s an article from today :facepalm:
Just because today is the first day they mentioned it does not meant today was the first day they became aware of the issue.
If they already have a release date it means the code has been in testing for a while.
This ain't rocket surgery.
 
Just because today is the first day they mentioned it does not meant today was the first day they became aware of the issue.
If they already have a release date it means the code has been in testing for a while.
This ain't rocket surgery.
If they had to be dragged kicking and screaming into acknowledging it because the bad publicity reached a boiling point, that's a lot worse than if they'd gotten out in front of it weeks ago with a statement, instead of just stealth-nerfing AGESA and calling it a day.
 
Last edited:
If they had to be dragged kicking and screaming into acknowledging it because the bad publicity reached a boiling point, that's a lot worse than if they'd gotten out in front of it weeks ago.
It is what it is. 3/4's of the folks who bought one don't know or care about it. 20% are aware of the issue. 3% percent are upset but have no reason to be, 1% are extremely vocal about it but would never notice the extra 125 mhz in real world usage and the other 1% need to have their meds balanced.
imho
 
Well, this has been quite the entertaining thread so far.

tenor.gif
 
Normally all the aw folks would receive their spankings and free vacations for a few days. Must be holiday weekend lag.


Yeah I am bit surprised Kyle's drunk ban hammer didn't materialize over the weekend. Its funny that the most vocal refuse to acknowledge that the latest release actually improved performance vs what you got at launch, despite the clocks not reading as high as people want. I will take the guaranteed performance vs seeing a slightly higher number any day of the week but that's just me.
 
Sometimes I get the feeling the same five people are responsible for every argument on the internet. The subjects are completely different, the forum is completely different, but I'll be damned if it doesn't seem like the same argument with the same insults over and over.
 
I just want to know if all the people up in arms here over their missing 50mhz are also going to stand in line with the same pitchfork on their hard drives not matching what's advertised.

A non insignificant number of 3900x owners are only getting clocks of 4275 with a small group as low as 4025. Thats far more than 50mhz. Watch the video.

Its 125mhz for me.

This has been covered about 30 times in the thread.

Nobody has a pitchfork. It's been an issue since july but no pitchforks arrived until outsiders saw the derbauer video and decided to come accuse us of pitchforks and crying. The drama is coming vastly from non ryzen 3000 owners. Funny red herring if you ask me. Its like asking your girlfriend what's wrong 1000 times until you find what's wrong is you asking the same question repeatedly :p
 
Box says 4.6. Mine ran for a month at 4.3, until an updated bios fixed that and it will hit 4.575 in some rare cases.
Really? Mine says "4.6Ghz Max Boost, 3.8Ghz Base Clock". Call me cynical, but any time a company uses language like "max" or "up to" I ignore whatever performance metric follows since outside of their testing environment I'm unlikely to see similar results. While I'll agree AMD needs to shut the hell up about potential boost clocks in their advertising, I'm not too upset - I mean I did buy one to replace the 7700K I'm running now.
 
Really? Mine says "4.6Ghz Max Boost, 3.8Ghz Base Clock". Call me cynical, but any time a company uses language like "max" or "up to" I ignore whatever performance metric follows since outside of their testing environment I'm unlikely to see similar results. While I'll agree AMD needs to shut the hell up about potential boost clocks in their advertising, I'm not too upset - I mean I did buy one to replace the 7700K I'm running now.

You like to live dangerously ;)

I wonder if we'll see the max boost crank up when all-core overclocks are active or will it only be when all cores are at base clock and it kicks-in for whatever single thread perf app needs it ? (I don't think that would be games as they peg other cores too).
 
You like to live dangerously ;)

I wonder if we'll see the max boost crank up when all-core overclocks are active or will it only be when all cores are at base clock and it kicks-in for whatever single thread perf app needs it ? (I don't think that would be games as they peg other cores too).

I imagine somewhere in between.

I have trouble a bios update remedying this all the way. When max manual OC’s are a few hundred Mhz under the boost on the box...

I am also curious how fast it’ll happen. I’d like to think the amount of testing is insane. Imagine if they mess it up and millions of CPUs go unstable.
 
I imagine somewhere in between.

I have trouble a bios update remedying this all the way. When max manual OC’s are a few hundred Mhz under the boost on the box...

I am also curious how fast it’ll happen. I’d like to think the amount of testing is insane. Imagine if they mess it up and millions of CPUs go unstable.
No different than Intel CPU's needed the hundreds of BIOS updates/software fixes for their vulnerabilities(features). Still more to come.
They both have the same problems and glad to see when they fix them. When they do not, that is when you can cry.
 
I read hundreds of customer experiences on forums, note the results of dozens of reviewers and read the resulting statements from the manufacturer itself finally acknowledging the issue.

I never said that the issue did not exist for "Some people"...I said that while people are up in arm's over missing a small of boost, they are completely overlooking the fact that the LADT AGESA ND CHIPSET IMPROVED PERFORMANCE VS THE LAUNCH PERFORMANCE.

I am one of those that is missing ~50Mhz of boost , as my CPU was boosting to 4.415Ghz on the launch AGESA code. I now see 4366Mhz, but again, PERFORMANCE has INCREASED despite the clockspeed "loss"..

So yeah, I understand the difference between an apple and an orange. I just don't care since my number of apples has increased since launch week. And the vest thing about it is that they were free!:rolleyes:
 
I never said that the issue did not exist for "Some people"...I said that while people are up in arm's over missing a small of boost, they are completely overlooking the fact that the LADT AGESA ND CHIPSET IMPROVED PERFORMANCE VS THE LAUNCH PERFORMANCE.

I am one of those that is missing ~50Mhz of boost , as my CPU was boosting to 4.415Ghz on the launch AGESA code. I now see 4366Mhz, but again, PERFORMANCE has INCREASED despite the clockspeed "loss"..

So yeah, I understand the difference between an apple and an orange. I just don't care since my number of apples has increased since launch week. And the vest thing about it is that they were free!:rolleyes:
So you would have no problem if 3950X came out, had great performance, beat every Intel CPU in games, and was marketed as a 6 GHz boosting CPU? How about they advertise it as a 32/64 core CPU? The performance would still be great and the best in the Ryzen 3000 series. Why does it matter how many cores there are in reality?

They shouldn't even bother with true spec anymore, they should just put a CB score on the box and write whatever they please elsewhere. That is the logical conclusion of your position.
 
So you would have no problem if 3950X came out, had great performance, beat every Intel CPU in games, and was marketed as a 6 GHz boosting CPU? How about they advertise it as a 32/64 core CPU? The performance would still be great and the best in the Ryzen 3000 series. Why does it matter how many cores there are in reality?

Well there was a time where they would release it and call it...

Ryzen 2 16 6000+ .... as according to their tests it would take intel 6ghz to match it.

Careful what you all wish for... PR rating marketing was stupid.
 
AMD themselves officially acknowledged this as an issue, so I'm not sure where the apologists are coming from.

An issue sure... but it's not a major pitchfork 20 page thread issue either. In general most people seem to be 50-75 mhz off the listed max single core boost. That is really not a big deal. I hope their next bios update pushes most people so they are exceeding the rated max boost by the same. But lets all get real... actual performance will be un effected. If it took you 2 min to render something right now it will take you 2 min later even if AMDs bios gives you a 200mhz single core boost bump. If you got 110 FPS in your favorite game before you will get no better after.

If they where missing the single core boost speed by 10% or something that would be news... missing it by 1% who the fuck seriously cares.
 
An issue sure... but it's not a major pitchfork 20 page thread issue either. In general most people seem to be 50-75 mhz off the listed max single core boost. That is really not a big deal. I hope their next bios update pushes most people so they are exceeding the rated max boost by the same. But lets all get real... actual performance will be un effected. If it took you 2 min to render something right now it will take you 2 min later even if AMDs bios gives you a 200mhz single core boost bump. If you got 110 FPS in your favorite game before you will get no better after.

If they where missing the single core boost speed by 10% or something that would be news... missing it by 1% who the fuck seriously cares.
Almost no one took it as a huge issue, but they took issue with people claiming it was no issue at all. And if it weren't for threads like these, AMD wouldn't have been working on improving the CPU further, so I don't see why they're fighting against it.
 
Almost no one took it as a huge issue, but they took issue with people claiming it was no issue at all. And if it weren't for threads like these, AMD wouldn't have been working on improving the CPU further, so I don't see why they're fighting against it.

BS, like your biased OP did not lead posters into a narrative. Did anyone debauer notify AMD? Making an exploitative video gets him more hits ya know, more money. He's not in the business to not make money.
 
Last edited:
Almost no one took it as a huge issue, but they took issue with people claiming it was no issue at all.

Fair enough. I'm not suggesting it shouldn't be brought to AMDs attention... so hopefully bios updates can get us to where it should be. I assume AMD really had their chips there and where not faking anything... its possible production level bioses regressed something. (I would even put money on the MOBO venders being overly cautious I mean 50mhz at 4350 mhz could easily be the mobo companies purposely feeding the chips a bit less for max stability) I'll give AMD the benefit of the doubt as its 2 months after launch and not 7 or 8. I understand most folks just want to get the max out of their chips. Hey I would love to see my 3600x hit 4.4 right now it is just shy. Some folks have made it sound like AMD has went pure evil here shipping bad chips with far less performance then marketed. Which isn't exactly true.

Anyway looking forward to the next bios release whenever Gigabyte feels like giving it to me after AMD releases the code. :)

In no way at all did AMD mess up any launch however, as you state in your op. I have yet to talk to anyone that picked up a Ryzen 2 that hasn't reported massive bumps in performance over their last generation Ryzens... or even 2-3 year old Intel parts. Ryzen 2 is the real deal performance wise.
 
Last edited:
BS, like your biased OP did not lead posters into a narrative. Did anyone debauer notify AMD? Making an exploitative video gets him more hits ya know, more money. He's not in the business to not make money.
A narrative that a large number of CPUs aren't reaching their spec? I hope it did that as it's a fact.
Yes, he did, as have others. They even told him how to test optimally for the boost behaviour.

In no way at all did AMD mess up any launch however, as you state in your op. I have yet to talk to anyone that picked up a Ryzen 2 that hasn't reported massive bumps in performance over their last generation Ryzens... or even 2-3 year old Intel parts. Ryzen 2 is the real deal performance wise.
This is the second Ryzen launch AMD had prominent issues that had to be fixed with several quick AGESA updates, not counting the various issues on the GPU front.
I haven't seen anyone disagree with the latter.
 
Last edited:
This is the second Ryzen launch AMD had prominent issues that had to be fixed with several quick AGESA updates, not counting the various issues on the GPU front.
I haven't seen anyone disagree with the latter.

I'll take launches that take a few months to be running 100%.... over requiring me to buy a new MOBO and CPU fan bracket for every rehash. Ya getting support for new chips working 100% on chipsets designed before those chips existed takes a bit more software work. I'm ok with that. It's hardly like the other guys have never had a bios issue.
 
I'll take launches that take a few months to be running 100%.... over requiring me to buy a new MOBO and CPU fan bracket for every rehash. Ya getting support for new chips working 100% on chipsets designed before those chips existed takes a bit more software work. I'm ok with that. It's hardly like the other guys have never had a bios issue.
I completely agree with you, but that is whataboutism. Both things can be true at the same time.
 
An issue sure... but it's not a major pitchfork 20 page thread issue either. In general most people seem to be 50-75 mhz off the listed max single core boost. That is really not a big deal. I hope their next bios update pushes most people so they are exceeding the rated max boost by the same. But lets all get real... actual performance will be un effected. If it took you 2 min to render something right now it will take you 2 min later even if AMDs bios gives you a 200mhz single core boost bump. If you got 110 FPS in your favorite game before you will get no better after.

If they where missing the single core boost speed by 10% or something that would be news... missing it by 1% who the fuck seriously cares.

The only reason the thread got as long as it did was due to people acting personally insulted by people discussing the issue in a way that wasn't 100% in line with their own view. Outside of people making up stories about people "crying" about this issue most of the conversation in the thread was calm, on topic, and not filled with too much hyperbole. It's a minor issue and the vast majority of people were treating it as such and just sharing their thoughts on it. It only turned into something big because people wanted to act like drama queens.
 
I completely agree with you, but that is whataboutism. Both things can be true at the same time.

It's more a case of choose your poison.

One company has decided that any chip with a different ID gets a new board... even if the chip is nothing but a rebrand. Yes that makes it much easier to release a bios on day 0 that performs exactly as expected cause vendors will be dealing with one specific design. I guess that is an advantage for early adopters no doubt. As long as those early adopters are ok with only upgrading every few years.... or have unlimited pockets.

The other company has decided to stay loyal to one socket for as long as possible. Sure that means bios updates for 3 year old designs are possible. It also means that on day 0 they are going to have to support a lot of boards some of which very wildly from the reference AMD designs in terms of power delivery ect. Still a few months after launch you can in general expect damn near perfect support. My 3600x in a b450 board may miss its rated max boost by 25-50mhz but its rock solid stable and if I was not watching for the boosts I would have no idea it was hitting 4.375 instead of 4.4.

It's a different design philosophy. And as horrible as it is that AMDs release bioses are in general not perfect.... and it's the bios released 3-6 months after each chip that ends up being the one people forget about for 3 years. The alternative is Intels design... great for early adopters with unlimited budgets. While punishing early adopters without deep pockets.

Whatever you prefer. I'll take AMDs approach myself... drop a new CPU in almost every year if you want and sell the old one off for damn cheap upgrades. If you started 3 years ago with a 16/17/1800 you could easily have dropped a 26/2700 in last year and sold your old chip off for a 50%(or better) cost upgrade, then you could do the same today dropping in a 36/37/38/3900 and again selling the old chip. Yes the catch is your 370 board is going to need bios updates each gen, and at some point the ride will end.

A low single digit % off the single core boost clock for a few months... is hardly anything to get too bend over. That AMD has pulled off as as smooth a launch as they have with insane numbers of chips moved and most of them going into older non x570 boards. Is perhaps proof that their chipset software team is in fact doing a pretty damn bang up job. We are not talking about crashing locking up dying machines here... we are talking about 0.5-3% off single core boost MAX ratings. That will likely be tuned. IMO I prefer that they air on the side of caution at launch and undervolt/shave a bit of the multipliers to ensure stability. Then when everything is ironed out drop the performance update. (I mean even on the Intel side is that not the way basically every launch goes... launch it stable, tune it for max go after)
 
Der8aur: hey there looks to be a discrepancy with AMD's advertised clockspeeds...

AMD fanboys: NYUH UH HOW DARE YOU FUCKING SHILL WHAT A INSIGNIFICANT HACK THIS GUY IS WHAT AN IDIOT

AMD: hey there looks to be a discrepancy with our advertised clockspeeds...
 
Fair enough. I'm not suggesting it shouldn't be brought to AMDs attention... so hopefully bios updates can get us to where it should be. I assume AMD really had their chips there and where not faking anything... its possible production level bioses regressed something. (I would even put money on the MOBO venders being overly cautious I mean 50mhz at 4350 mhz could easily be the mobo companies purposely feeding the chips a bit less for max stability) I'll give AMD the benefit of the doubt as its 2 months after launch and not 7 or 8. I understand most folks just want to get the max out of their chips. Hey I would love to see my 3600x hit 4.4 right now it is just shy.

Agreed. (thanks for the logical point of view)

Some folks have made it sound like AMD has went pure evil here shipping bad chips with far less performance then marketed. Which isn't exactly true.

Where have you seen this? Granted I haven't read all 10 pages, after the first page the pattern was clear... issue pointed out, fanbois on the defensive blaming everything else but AMD...

In no way at all did AMD mess up any launch however, as you state in your op...

Have to disagree on that one, facts speak for themselves.
 
Last edited:
I never said that there wasn't an issue. I simply said that people, in their zeal to have something to complain about, have been screaming about how bad it is and that they aren't getting the maximum performance that they should based solely on the fact that the MAX listed boost clock is not being reached enough for their liking...

The issue I have with it is that in their rabbit fury their brains stopped working and they overlooked the fact, yes MAX boost clocks are down a bit, but performance IS UP. I have a very good 3700x. It was boosting over 4.4Ghz on the wraith Prism in case with just a single 120mm intake fan running, while my 5700XT was on the stock blower helping to create heat. The performance I got in that setup is being exceeded despite the max boost clock only showing 4366Mhz now. Would I like to see my boost clock number go back over 4.4ghz? Sure why not, as long as performance is not decreased from what it is at now.


I do not have an issue with this, since the boost clock is listed as a max spec, and performance INCREASED. The fact that AMD acknowledged the issue is being held up as some crazy proof of their attempt to lie to customers when in fact they had to respond. It doesn't matter that the number of people complaining most like represent 5% (and I am being generous here) of their total Ryzen sales in the DIY channel. The rest of the complaining is coming from the trolls/IDF force. People that have never owned a Ryzen SKU (even a fair number here) are screaming and stirring the pot up, which then will turn uninformed buyers away since they think they are going to be slighted if they opt for a Ryzen 3K SKU at this time.

The other side of this coin is the Mobo companies, and I feel that they should be roasted over the coals. Asus represents a HUGE chunk of the DIY market, especially in higher end boards, and their bios support has been absolute garbage. They have released BIOS images that have bugs that have been there for over a year, not to mention the high voltage issues and such. I personally will not ever buy another Asus board after the experience I have had with them over my last two boards.


PS: I am personally kind of sick and tired of way social media has changed the way people think and act anymore. Everyone has this sense of entitlement and are quick to blame any issue they have on anyone and everyone else when a large majority of the time they are the ones at fault.

The fact that companies now have to worry about a tiny number of people (and I mean ANY company, not just AMD) that have unrealistic expectations and no desire to ensure that they are not the ones at fault hurts all of us. It creates a higher operating cost for the company, who then decides to either shoulder the burden of this and pass the cost on to the consumer or to just stop supporting things very quickly after launch. I hate having to pay more money because of stupidity, but that seems to be the way the world is going.

/get off my grass rant over
 
Last edited:
I don't know. Without reading 10 pages, I think it's a lot to do about nothing. I don't know that you'd ever be able to isolate the core that boosts the highest (up to the available max boost), and even if you did, something else is going on in the background that causes other cores to kick in. My 3600 will boost to 4.195Ghz (never actually to that "magic" 4.2Ghz boost) but 3 of the 6 cores have some load on them even when idle due to background processes. All cores run at 4.1Ghz pretty regularly. I don't really feel like the testing methodology used is ideal personally. Also, my board doesn't run at exactly 100mhz bus speed either.

That being said, if AMD isolated a firmware issue and is pushing out a fix, great... I don't feel like I'm missing out in the meantime though. I'm not holding onto every scrap of documentation for a $35 class action payout 10 years from now over this one though.
 
A quote from an article that links a Reddit post as source that has a statement from an Asus employee. Oh, Intel, never change.

And AMD, why on Earth do you give them any ammunition even when you're beating them...
 
Back
Top