In a blunt video posted late Thursday night, outspoken former Intel principal engineer Francois Pidnoel provided his recommendation on “fix” Intel CPUs, criticized present management for not being engineers, stated AVX512 was a misadventure, and declared that it’s solely luck AMD hasn’t grabbed extra market share.
“First, Intel is really out of focus,” Piednoel stated within the practically hour-long video presentation. “The leaders of Intel today are not engineers, they are not people who understand what to design to the market.”
Piednoel stated Intel’s technical choices have largely been “nonsense” since 2016. Incidentally, Piednoel left Intel in 2017 after serving as a principal engineer and efficiency architect for 20 years, engaged on CPUs from the Pentium III to the Sixth-gen Core i7. The outspoken engineer usually made technical shows and demonstration pitches to the {hardware} press, passionately arguing why design choices made by Intel have been the fitting choices.
Pienoel admitted his data on Intel is basically “obsolete” and years outdated. That additionally lets him communicate freely as nothing he spoke of was from data obtained below an NDA, he stated. Instead, his evaluation was largely primarily based on public data that’s been swirling round Intel.
The whole video is price a look ahead to fanatics, however we’ve highlighted his most intriguing claims right here.
Gordon Mah Ung / IDGJamming AVX512 features right into a laptop computer CPU like this 10th-gen Ice Lake CPU is a mistake, a former Intel engineer stated.
AVX512 is a mistake
AVX512 is the premise of the DL Boost AI acceleration Intel makes use of in its Xeon server CPUs, and the know-how has discovered its means into shopper chips such because the 10th-gen Ice Lake laptop computer CPU. Pidnoel flat-out dismissed together with AVX512 in shopper chips as a mistake.
“You had Skylake and Skylake X for a reason,” Piednoel stated. “AVX512 is designed for a race of throughput that is lost to the GPU already. There’s two ways to get throughput. One is to get the throughput is by having larger vectors to your core, and the other way is to have more cores.”
Piednoel, who as soon as instructed me after Intel’s Pentium four misadventure that “we learned you can’t recompile the world,” appeared to indicate the software program recreation wasn’t profitable Intel any battle this time both.
This in all probability isn’t new to anybody who heard famed Linux creator Linus Torvalds attain deep to spew anger at Intel’s AVX512 method simply final month.
“I hope AVX512 dies a painful death, and that Intel starts fixing real problems instead of trying to create magic instructions to then create benchmarks that they can look good on,” the never-too-shy-to-cut-loose Torvalds stated.
Francois Piednoel Loss of focus
Another mistake Intel made was to defocus the corporate from its core enterprise of constructing quick CPUs, Piednoel stated.
Intel went on a diversified shopping for spree within the final half of the last decade that left the corporate unable to give attention to its CPU enterprise. This let bitter rival AMD meet up with it, and the one factor saving Intel from shedding a extra huge market share is AMD’s quantity constraints in making its in style CPUs.
“Intel is very lucky AMD cannot get the volume, to be able to compete,” Piednoel. “If they were getting volume, the price difference would definitely cost Intel market share a lot more than what they are losing right now.”
Just yesterday AMD reached an all-time excessive of 20 p.c market share in laptops, in accordance with numbers from Mercury Research. We known as AMD’s new 7nm Ryzen 4000-series laptop computer chips “game-changing” in our evaluation this spring.
“Intel is lucky AMD has capacity constraints and because of this, they can’t grab market share fast enough,” he stated. “We kind of had the same thing when AMD had Athlon 64 and we were basically trying to catch up with Pentium 4 to Conroe.”
Indeed, AMD had made an enormous dent…
