If she didnt mention RDNA4 there s few chances that it will be exposed, hope that she ll eventually give some hints for what is to be expected, i mean, DT Ryzen will have to rely on GPUs for AI, that would be uncautious to instillate the idea that it could be only a GPU from Nvidia.
That s still 2 ops/cycle/ALU since ages, at least for 32b and 64b FP, granted that the format of the operands have been extended from 32b/64b to 16b, or even INT8, and the exe units surely splitted to double the throughput when the required precision is halved.
They share the underlying designs, they use the same ALUs FI, it s what is around that is adapted to the different tasks, FI rendering back end/GDDR controler for client and HBM controler for GPGPU that do not need the rendering part.
Possible, but that was surely on the drawing boards well before the AI market surged exponentialy, so that cost not much to release what was already well advanced last year.
The success of the 7800XT is a testimony that there s more sales around 500-600$, not that something that is faster is...
AMD just confirmed these days 4bn $ shipment of MI300 for 2024, given the current AI frenzy they can easily double this number in 2025 if they have some technical advantage, that would surpass their servers or CPU client sales within 2 years.
Guess that s where they ll put their priorities...
With 50% more cores than Hawk Point you get 2.5x the perf/watt at same throughput, and that would be with 12 Zen 4 cores, so given that it s cores with higher IPC i let you imagine how much the competition is lagging since it couldnt even match the previous AMD 8C APU in the 15 - 45W range.
They can release AM5 versions of Strix Point limited to OEMs in a first step, and it can be used as it is in SFFs with soldered BGAs, 8-12C is even more than necessary, after all Phoenix ended on AM5 in the consumer retail market.
5950X is 14% faster and 5800X is 10% faster than the 10900K, that s at 720p.
https://www.computerbase.de/2023-10/intel-core-i9-14900k-i7-14700k-i5-14600k-test/3/#abschnitt_taktraten_in_spielen
9:30 Taiwan, wich here in Western Europe is 3:30 AM, at the same time too late and too early, if they wanted to bother us they couldnt have done better.
That s 11% higher perf at same power, and 22% lower power at same frequency, the latter amount to 28% higher perf/watt at isofrequency because 1/0.78 = 1.28.
It s either 11% higher perf or 22% higher efficency, the two possibilities are exclusive, you cant have both at the same time, although there s something wrong in the efficency number, it s rather 22% lower power, wich amount actually to 28% better efficency.
They will surely have sligtly better MT perf, same as the 5600GT vs 5600G,
so not that pointless.
https://www.computerbase.de/2024-01/der-sockel-am4-lebt-5700x3d-und-5800x3d-trennen-200-usd-5000gt-ab-125-usd/
Where did i say that it can be generalized.?
But you ll notice that even the corean market is far from Mercury s and JPR numbers that put AMD in the vicinity of 20% CPU marketshare, they just greatly underestimate the retail market numbers, just look at Microcenter annual revenue, there s a big...
Mercury account only OEMs sales, and only the biggest ones, the Minisforum and other Geekom are not accounted, let alone the retail CPU market.
Notice that they do the same for GPUs, they account only the GPUs sold in commercial PCs, be it laptops or DTs, when you buy a dGPU or CPU at...
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.