We'll rarely find a better occasion to prove our point.Gamers are a loud bunch. But you are right, is anyone really losing out on a gaming experience by having only a 5600X/5800X3D? I very much doubt it. I can't understand why people call it a fail because it didn't improve gaming much. They need to chill out. I think the techtubers are doing it for views.
X3D delay seems to speak to real internal execution or development problems with Zen 5. At first it was rumored to be coming sooner than it did last generation (and apparently for good reason), now it seems like it will be the same difference in timeframe or possibly even longer than last gen. Perhaps MLID was right about the issues with Zen 5 development all along?
Makes you wonder, whats the cause of the delay? Same 7/6nm process for X3D die, presumably, and same 4nm process for CCD. New CCD stepping required for X3D possibly, or something else? We've already seen that Turin classic seems to be a new CCD stepping from launch desktop, maybe the delay is to produce more volume for EPYC?
Then we've still yet to see 3nm Zen 5C in action. Who the hell knows what joy it will bring? Overall, seems Zen 5 development and execution has really kicked AMDs ass...
I understand it that such instructions will both write to PRF and to memory, so the number of store pipelines still matters.
Yeah, can't delay something that's never had an official release date in the first place.Rumors being wrong does not equal a delay.
This reminds me a thing I found during the design of a CPU I was involved in: there was a system register that when written to was flushing many structures because some of its bits controlled various parts of the CPU (MMU, caches etc). As this register was not supposed to be written often, no special care was taken when writing values in it. That's when I realised Linux was writing to this register on every syscall, rewriting the same value as before (don't ask why, I don't know). Needless to say, the designer I talked to then made sure he compared old and new values of the register to prevent useless flushes. These details can have very measurable impacts on end-user applications.
Blame the current launch debacle.So much for the September rumor. it seems 9000X3D is launched at CES 2025:
Software readiness. Have you SEEN what you have to do to Windows on 2CCD parts for example?X3D delay seems to speak to real internal execution or development problems with Zen 5. At first it was rumored to be coming sooner than it did last generation (and apparently for good reason), now it seems like it will be the same difference in timeframe or possibly even longer than last gen. Perhaps MLID was right about the issues with Zen 5 development all along?
Makes you wonder, whats the cause of the delay? Same 7/6nm process for X3D die, presumably, and same 4nm process for CCD. New CCD stepping required for X3D possibly, or something else? We've already seen that Turin classic seems to be a new CCD stepping from launch desktop, maybe the delay is to produce more volume for EPYC?
Then we've still yet to see 3nm Zen 5C in action. Who the hell knows what joy it will bring? Overall, seems Zen 5 development and execution has really kicked AMDs ass...
I spoke with someone that works for a partner, and the September date was communicated with them, however the poster below is right. An unannounced, unreleased product technically can’t be delayed.Rumors being wrong does not equal a delay.
Local cult expects Apocalypse any day now.WTFTech seems to think one Zen5 X3D part will launch this year. We will see.
Release your eekBench, man!Blame the current launch debacle.
Do you know acquaintances that happen to work for GPU partners 🤔I spoke with someone that works for a partner
Not anymore. I quit working some time ago (medical issues) and the few people I did know have been gradually changing jobs, so soon I will know nobody and I will be sad.Do you know acquaintances that happen to work for GPU partners 🤔
Powercolor rep thinks RDNA 4 is launching only next year but MLID disputes it. Is AMD doing a jebait again ?
Internal roadmaps don't exist I guessYeah, can't delay something that's never had an official release date in the first place.
This works for me.So much for the September rumor. it seems 9000X3D is launched at CES 2025:
Yeah, can't delay something that's never had an official release date in the first place.
Do share.Internal roadmaps don't exist I guess
The one that was supposed to be released in April!It is delayed in the same way Zen 5 failed to deliver the promised 40% IPC gain!
I'm talking in general terms, the idea that something can't be delayed because it wasn't put in a public roadmap is incredibly stupid.Do share.
“At the end of the day, there were a series of decisions that AMD made, that differed from how reviewers were testing, that led to different conclusions,” McAfee said. “This is not saying that reviewers did anything wrong, or that there’s anything in the process of how these reviews were done that was incorrect. We were as puzzled by some of the results that reviewers were seeing as the Ryzen fans are out there in the world, and that’s really what led us to this sort of multi-layered conclusion that there’s a number of factors going on here that created that separation.
This is no shade whatsoever on reviewers and how they tested, this is completely on things that we didn’t really understand as a part of the way that we tested and configured our products, and how that differed from both reviewer game suites moving in a slightly different direction, as well as how they tested the product versus how we did. It’s not a criticism of reviewers or how they test, simply significant differences in setup and configuration that got us from point A to point B and I think we understand that now.”
Invest more on QAMoral of the story: don't continue to employ marketing and public relations people that are terrible at their jobs.
From a consumer standpoint, the only thing that warrants being named a delay is when a company fails to meet deadline that they publicly committed to.I'm talking in general terms, the idea that something can't be delayed because it wasn't put in a public roadmap is incredibly stupid.
So if Nova Lake releases in 2035 I guess it's not delayed.From a consumer standpoint, the only thing that warrants being named a delay is when a company fails to meet deadline that they publicly committed to.
I didn't say that.What's even more stupid, or shall I say tone-deaf is suggesting that this thread of all places should continue eating up any rumour popping up as gospel.