The Intel Execution in [2025]

[tuna]

Their management got paid plenty.

 

[MfA]

This is exactly the problem I was saying about Intel: they really wanted to push x86 everywhere, without considering whether it's a good fit or not. Mobile is growing? Make a x86 CPU for it (Atom). Super computing? Make a x86 CPU array. They even wanted to push x86 into embedded computing (with limited success).

Poor timing and design decisions.

With Mediatek moving to designs with all big cores it's getting to the point where x86 isn't much of a problem any more for mobile either. As for Xeon Phi, CU/SM's in GPUs aren't small enough that using x86 for the scalar processing is a significant overhead ... it was how Intel decided to design the rest of the processor which made it impossible to compete with GPUs. Narrower SIMD, less SMT, strict coherence etc ... that killed the computational density, not x86.

 

[pcchen]

Poor timing and design decisions.

With Mediatek moving to designs with all big cores it's getting to the point where x86 isn't much of a problem any more for mobile either. As for Xeon Phi, CU/SM's in GPUs aren't small enough that using x86 for the scalar processing is a significant overhead ... it was how Intel decided to design the rest of the processor which made it impossible to compete with GPUs. Narrower SIMD, less SMT, strict coherence etc ... that killed the computational density, not x86.

I think this is partly a x86 problem, because some of these decisions probably was at least partly due to some limitations of x86. For example, x86 only had 8 GPRs, so SSE also has 8 registers, but since SSE is a completely new set of instructions, they don't really need to have only 8 registers, but yet Intel still made it with only 8 registers. They probably think it's better this way as some decoding hardwares can be reused or something.

It's only later with VEX encoding they are able to support 3-operand instructions and with EVEX to extended the number of registers to 32. They could do this earlier, but I think the baggages of x86 still shows in mnay places.

 

[Cyan]

Analyst: Intel's new CEO to refocus on chip design, get major customers for its fab business UBS analyst.

Timothy Arcuri says Intel's new CEO Lip-Bu Tan will refocus on chip design in short term, will secure big customers for its foundry business.

 

[Potato Head]

Does Timothy Arcuri know there’s no such thing as short term chip design?

 

[DegustatoR]

Based on the latest report by reliable insider and leaker, @Jaykihn0, it looks like the company has significantly altered its discrete GPU plans. It is said that the high-end Battlemage BMG-G31 GPU is more or less dead and that has been the case since Q3 of 2024. We did see some dies ship by the tail end of 2024, but it looks like those may just be for testing or evaluation purposes with plans for a retail launch being scrapped much earlier.

Intel Reportedly Cancels Its High-End Xe2 Arc Battlemage "BMG-G31" GPUs

Intel has reportedly canceled its high-end Xe2-based Arc Battlemage "BMG-G31" GPUs which were going to be used in gaming graphics cards.

wccftech.com

 

[trinibwoy]

Not surprising. They’re too far behind on PPA with Battlemage.