Thursday, October 31st 2019
AMD CEO Lisa Su Talks About 3rd Gen Ryzen Boost Issue in Q3 Earnings Call
AMD CEO Dr Lisa Su in response to a question, spoke about 3rd generation Ryzen processor boost issue. Dr. Su was responding to a question by Mitch Steves of RBC Capital on whether she had comments on "the software side" of 3rd gen Ryzen, and articles in the press still popping up about them despite AMD's fix. This was interpreted by the AMD CEO as a question specific to the Precision Boost controversy surrounding 3rd gen Ryzen chips, in which processors would seldom/never hit the advertised maximum boost frequency. AMD tried to address this by issuing updates to its processor microcode under AGESA Combo 1.0.0.3 ABBA, distributed through motherboard BIOS updates. The new microcode is supposed to increase the maximum turbo clock-speeds for "the vast majority" of users.
In her response, Dr. Su began by stating that the company is pleased with the sales of these processors. She then mentioned that AMD is working with its motherboard partners and ODM partners to "improve the optimization of the maximum boost frequency." She notes that the issue has been "largely addressed over the last couple of weeks" (referring to 1.0.0.3 ABBA). She goes on to state that AMD sees its response to the boost issues as more of an "optimization," rather than a "major update," possibly trying to allay investor fears that AMD is firefighting a costly problem with its products. "We're going to continue to improve the platform," she adds, possibly referencing the upcoming AGESA 1.0.0.4 Patch B microcode that's beginning to ship out by motherboard vendors. The earnings call can be accessed here. The specific question can be found at 47:00.
In her response, Dr. Su began by stating that the company is pleased with the sales of these processors. She then mentioned that AMD is working with its motherboard partners and ODM partners to "improve the optimization of the maximum boost frequency." She notes that the issue has been "largely addressed over the last couple of weeks" (referring to 1.0.0.3 ABBA). She goes on to state that AMD sees its response to the boost issues as more of an "optimization," rather than a "major update," possibly trying to allay investor fears that AMD is firefighting a costly problem with its products. "We're going to continue to improve the platform," she adds, possibly referencing the upcoming AGESA 1.0.0.4 Patch B microcode that's beginning to ship out by motherboard vendors. The earnings call can be accessed here. The specific question can be found at 47:00.
32 Comments on AMD CEO Lisa Su Talks About 3rd Gen Ryzen Boost Issue in Q3 Earnings Call
seriously, what the fuck?
www.anandtech.com/show/14873/reaching-for-turbo-aligning-perception-with-amds-frequency-metrics-
I've sat through far too many of those kind of events. The only thing more annoying are certain Germans at tech events, as they always try to prove that they know more than the company does about their own products...
That said, I am in need of a new system, badly. ill wait it out for DDR5
On one hand most people dont give much fluff about 25-50MHz 1-2C boost, on the other this should've been resolved before actual mass release... especially when some parts did not meet clocks that were written on the box.
Like what a company's revenue per share is is what the stocks "supposed" to be, and when the next recession hits and stocks go way back down to where they're supposed to be alot of people are gonna be sol.
My point is, not everyone has the exact same build - it's just not possible. Slight variances in different hardware being used, different software installed, different BIOS in use and so on and so forth.
For anyone to expect things to be perfect through all the variations of hardware and software out there, you're crazy. Ideally, it would be nice if all the CPUs worked without fail.
Anyway, maybe at AMD they didn't run into issues with the handfuls of hardware setups they used for testing. But once things are released to the wilds and no longer in a controlled environment it's very possible for some issues to crop up.
This is something that should have been caught #1 - a lot earlier, #2 by AMD themselves. Looking around the web, it seems like more people than not had this issue.... how could that NOT be caught earlier???
As for AMD, unless you're part of their engineering team and know for sure, it's just an assumption on your part that AMD dropped the ball. We can draw conclusions all day, but we'll just never know. Perhaps there was something new they didn't account for and missed and results worked correctly from all their other testing. From things I've read it wasn't something that impacted a majority of owners. Some had issues, some didn't and some has CPUs that boosted above the suggested rage - and this is just from a small sampling of folks that have posted. You have that whole other group of people that never looked or reported issues or no issues. It's not fair to claim it was a majority of people.
Wasn't it the 2080Ti cards that some in the first batch were failing and others weren't? You can't claim that a majority of users on the 2080Ti were having issues with faulty cards without taking note from every single person that purchased it upon release. If you only take a small pool of people that are verbal about their failures and you don't get the big picture and you're left with skewed results and incorrect conclusions get drawn. Where there issues? Sure. Did they get fixed, it appears so.
Just like with AMD's Zen 2 - where their issues? Apparently. Some folks had boosting issues and AMD has been addressing them and fixing them.
I'm not defending AMD and saying they're in the clear, but perhaps this whole issue will allow AMD to re-evaluate their testing procedures to make sure it doesn't happen again.
I guess I'm just hoping folks will calm down and relax some. Then again, it is entertaining to read through the posts of some folks on these topics, get so worked up.
The problems with the 2080ti showed up months after release, no? Configuration wasnt relevant. You drop an amd chip in any mobo and it's a crapshoot if you received the specs on the box. Different situations to me.
I also want to note that I'm personally not fired up. It is what it is. I said it before in the discussion threads this isnt a huge deal (nvidia cards dieing happened to a much smaller percentage, but still worse). But make no mistake about it, this is AMDs snafu. If it wasnt, the same chips would work in the same boards...but some worked and some didnt. It wasnt the boards.
www.cnbc.com/video/2018/11/16/nvidia-stock-sent-crashing-after-missed-earnings.html
The problems with the 2080ti showed up months after release, no? Configuration wasnt relevant. You drop an amd chip in any mobo and it's a crapshoot if you received the specs on the box. Different situations to me.
I also want to note that I'm personally not fired up. It is what it is. I said it before in the discussion threads this isnt a huge deal (nvidia cards dieing happened to a much smaller percentage, but still worse). But make no mistake about it, this is AMDs snafu. If it wasnt, the same chips would work in the same boards...but some worked and some didnt. It wasnt the boards.
[/QUOTE]
I never said you were fired up, just a general comment on reading through topics in regards to such things.
You mean the poll info that was discussed here about der8auer's youtube video?
www.techpowerup.com/forums/threads/der8auer-only-small-percentage-of-3rd-gen-ryzen-cpus-hit-their-advertised-speeds.258840/#post-4108681
Folks posting in that thread couldn't agree on his data results as being useful or not - read through them all again if you can't recall.
Point is, you and no else outside of AMD still know if the issue was a result of poor work on AMD's side or if it was a circumstance beyond their sight and scope of testing they did.
It just seems salty that folks outright blame a company based on what little information is actually available.