Why does AMD keep pissing off the press and inflicting needless wounds on their reputation? SemiAccurate is sick of this sh*t and this article, ostensibly on the Ryzen 7000 teaser, is going to make for an interesting dinner tonight.
What are we talking about? AMD just pissed away 1/4 of the time they set aside for this launch for us to write an article on the video that will go live about the time this article does. Instead of being able to ask questions, think about the interesting things to write, and make a cogent story, we are forced to choose between doing a good job and not getting done in time or a crappy job and being late. Late means doing it simply isn’t worth it so you are stuck with a lose/lose situation.
This pisses us off. Again. AMD keeps doing this for no sane reason, it hurts them, hurts the quality of the information they get out, and there is no up side. They will claim that it is for security reasons but that isn’t true for things we won’t get in to in this article. Instead AMD goes the collective punishment route, the floggings shall continue until morale improves. Again, why?
Oh yeah, did we mention in the 4.5 hours between the event ending and the embargo lifting, two hours of that is when we are supposed to be in another event? So that 4.5 hours, call it 5, is actually 3. To make matters more hilarious, they forbade pictures during the taping, (Hi Pat!) and promised to get us slides soon. That narrowed the time to write window to 40 minutes or not enough time to properly edit the pictures. It is a pathetic showing once again, self-inflicted wounds abound here in the outskirts of Austin, and they are completely unnecessary. SemiAccurate wants to do a good job but AMD seems hell bent on precluding that option. Today’s briefing is proof that they have mastered the art. Well done?
So now we will admit we lied, there were actually 41 minutes from when we got the slides which were absolutely necessary to start writing because we couldn’t transcribe the info fast enough during event. Don’t believe us? Watch the stream and try to take complete notes on the technical slides, it isn’t easy. Remember, no cameras. So that said there is now 24 minutes left to write the story. We aren’t going to bother, just put up a few of the slides and get snarky.
One more thing to point out, we are not going to parrot back AMD’s performance claims. Given their past behavior, we just don’t trust their numbers. We don’t think their drivers are fully functional although we do think they are likely getting better, but until we can get our hands on one and try, we will rely on past behavior and not take their word for it. On with the show, 22 minutes left….
The money slide, literally
This is about the limit of the useful information presented today. Four SKUs, none of which SemiAccurate considers to be safe to deploy due to remote back doors invisible to the user. We bought Intel and until we have proof otherwise, we recommend you do t0o even if it is slower and more expensive. This is not a joke. We would love to dive into the 170W TDP but, time. Suffice it to say you can get performance from jacking up the energy use and AMD did just that.
Important numbers
Here AMD compares their 5nm process cores to Intel’s 10++nm cores and AMD comes out way ahead, about half the area and significantly better claimed energy use. One thing to note is that AMD’s parts are on a full node ahead of Intel. When Intel gets their real 7nm, not the fake ‘7’ that is really 10++nm, parts out they will be about the same area as AMD and will pull significantly less power. AMD is ahead here but Intel’s response will likely be right there on area when it happens in a few months.
Sockets are new but compatible
Here is the new AM5 socket, look familiar? Any more questions about why drip feeds of data piss us off? Nothing new here really.
Chipsets in the Extreme
There are two new chipsets, again not really new but a little clarified. Lets wait for the details but the real differentiator is the PCIe5 lanes. AMD ominously mentioned that it was something about PCIe5 device types but we just think it is lane count. That said when there isn’t enough time to ask basic questions, accuracy suffers and messages are confused. Thanks AMD!
More teasing….
Last and definitely least is the teaser of the next AMD Vega3 based GPU. They showed it playing a game (theoretically) live and it was fast but that is what you expect from a new generation of GPU. Now you know all that we do on the topic.
Overall AMD had a good showing for Ryzen 7000. If we felt the CPU line was safe to deploy from a security standpoint, and we definitely don’t, we would potentially have good things to say about them. Unfortunately AMD precluded that option with their scheduling and now we get to talk more about PR behavior than the product. They set it up and they got what they asked for. Sorry we couldn’t tell you more about the product but we didn’t have a choice. Once again, why AMD?S|A
Author’s Note: Six minutes left to edit the pics, post the story, and off to the next thing.
Author’s Note 2: Proofed it with -4 minutes left. Posting will definitely eat into the other events. SIGH.
Charlie Demerjian
Latest posts by Charlie Demerjian (see all)
- Microsoft Hobbles Intel Once Again - Sep 20, 2024
- What is really going on with Intel’s 18a process? - Sep 9, 2024
- Industry pioneer Mike Magee has passed away - Aug 12, 2024
- What is Qualcomm launching at IFA this year? - Aug 9, 2024
- SemiAccurate is back up - Aug 7, 2024