- Mar 3, 2017
- 1,623
- 5,894
- 136
Thats why I see it to be 'possible'. Alienware doesn't participate production but could know something related to marketing ahead of any leakers.AMD pre-prod parts have no names, just an OPN.
Didn't AMD already show a slide where they confirmed Zen 5 Ryzen 8000 CPUs or am I mistaken?It is possible that former "8950X" "8950HS" leaks could be fake
this is a new advertising material for old users of Alienware.
View attachment 88602
Of course they are fake. Actual SKUs aren’t formed yet. Desktop will probably be 9000 while mobile will be 8000 (unless AMD moves DIY to new naming scheme). I actually wanted to point this out a couple days ago, but got distracted.It is possible that former "8950X" "8950HS" leaks could be fake
this is a new advertising material for old users of Alienware.
View attachment 88602
Not that I am aware of, maybe I missed it? 8xxx for mobile, 9xxx for desktop.Didn't AMD already show a slide where they confirmed Zen 5 Ryzen 8000 CPUs or am I mistaken?
I think Intel is using TSMC for their Gaudi2 card. I don't know about packaging, but it is possible Intel is outsourcing that as well.They should unironically contract IFS for dual sourcing MI300 packaging. I don’t think there’s much of an IP risk through packaging alone so it should be a serious consideration.
Nvidia is already dual sourcing packaging to keep up with demand and I’m fairly certain it’s through IFS.
BTW, who is making those interposers for Mi300. Is it TSMC?EMIB pitch insufficient and I don't think Intel can reticle stitch passive slabs.
Didn't AMD already show a slide where they confirmed Zen 5 Ryzen 8000 CPUs or am I mistaken?
Of course they are fake. Actual SKUs aren’t formed yet. Desktop will probably be 9000 while mobile will be 8000 (unless AMD moves DIY to new naming scheme). I actually wanted to point this out a couple days ago, but got distracted.
Not that I am aware of, maybe I missed it? 8xxx for mobile, 9xxx for desktop.
You have to wonder about the trajectory of those sales.This attitude could be the difference between 2.0B in revenue vs 2.5B in revenue from MI300.
They would have to be silent about it. They’ve got prepay customers for Intel 3 & 18A and we still don’t know who it is. The last earnings call Pat Gelsinger specifically said they got multiple packaging deals for AI chips in Q3.This. Lmao there is no way Intel grabbed a packaging contract of that scale and just went silent about it.
I was talking about MI300 and the comments during Q3 earnings call. Frankly I don’t know much about Gaudi’s performance or sales expectations.You have to wonder about the trajectory of those sales.
Gaudi became available in May 2023, and will not reach revenue in 100s of millions until 2024. And, BTW, that 2 billion projection from Intel - it did not have a time period on it.
Since the projections are vague, they can be retroactively redefined. Another example: 1 million SPR shipped. Most people assumed Intel was projecting selling 1 million SPR by mikd 2023, but Intel missed that target, and then it became that Intel would ship 1 million SPR to its warehouse.
The 1m sales goal was reached sometimes in Q3.
That's years and years away.They’ve got prepay customers for Intel 3 & 18A and we still don’t know who it is.
GEMM brrr vendors are dime a dozen and only two are relevant.they got multiple packaging deals for AI chips in Q3.
They would have to be silent about it. They’ve got prepay customers for Intel 3 & 18A and we still don’t know who it is. The last earnings call Pat Gelsinger specifically said they got multiple packaging deals for AI chips in Q3.
I was talking about MI300 and the comments during Q3 earnings call. Frankly I don’t know much about Gaudi’s performance or sales expectations.
If past naming conventions are anything to go by (AMD hasn't even followed their own convention) - Zen 4 Desktop APU's will be 8000 series and Zen5 Desktop will be 9000 series (unsure about Zen 5 APU's since we're a the end of the avaliable numbers).Of course they are fake. Actual SKUs aren’t formed yet. Desktop will probably be 9000 while mobile will be 8000 (unless AMD moves DIY to new naming scheme). I actually wanted to point this out a couple days ago, but got distracted.
Not that I am aware of, maybe I missed it? 8xxx for mobile, 9xxx for desktop.
If past naming conventions are anything to go by (AMD hasn't even followed their own convention) - Zen 4 Desktop APU's will be 8000 series and Zen5 Desktop will be 9000 series (unsure about Zen 5 APU's since we're a the end of the avaliable numbers).
For those who are curious, AMD released this slide in 2017 and stopped following it almost as soon as it was published.
View attachment 88631
I remember reading a couple of weeks ago that CapEx differs immensely WRT advanced packaging. TSMC alone invests more into advanced packaging than all the OSATs together, which could render them obsolete mid-term.You're all focusing too much on TSM's inhouse 2.5d capacity.
There are myriad other options and all are eager to get a slice of that high margin pie.
Capex is stuff-years-away and we're talking 2.5D capacity here and now.I remember reading a couple of weeks ago that CapEx differs immensely WRT advanced packaging
Man that's great, still a few years from coming online.TSMC alone invests more into advanced packaging than all the OSATs together, which could render them obsolete mid-term.
Where were you the last week? This was already discussed here.View attachment 88648
Lol, I was in stitches after reading that.
2000 CB2024 MT score for the 12-core Strix Point APU?
THAT is absurd. That is like 2.5x the MT of 7840HS.
I did not know what kind of fantasies you can have after being high on some good stuff. Well, now I know after reading that.
/s
Two important questions: Cost and throughput? As soon as cost for advanced packaging is significant and throughput is limiting OSATs won't have an issue. Even today packaging is more often deliberately kept simple for cost and yield reasons, I don't think advanced packaging will see such a widespread adoption that it is able to obsolete "lesser" packaging.TSMC alone invests more into advanced packaging than all the OSATs together, which could render them obsolete mid-term.
Haha.
I was talking about advanced packaging specifically. While classical packaging will not disappear, it is a low-cost, low-margin business.Two important questions: Cost and throughput? As soon as cost for advanced packaging is significant and throughput is limiting OSATs won't have an issue. Even today packaging is more often deliberately kept simple for cost and yield reasons, I don't think advanced packaging will see such a widespread adoption that it is able to obsolete "lesser" packaging.
And this is a serious problem. You won't know what you have missed.Haha.
I wasn't following this thread last week.
OSAT stands for..?