5 ms·

Hyperscale and large corporate buyers aren't dumb enough to fall for this. Gimme everything enabled or I'll buy a neoverse or AMD thanks. And I'd reckon at ret

Hyperscale and large corporate buyers aren't dumb enough to fall for this. Gimme everything enabled or I'll buy a neoverse or AMD thanks. And I'd reckon at retail that sub-brand and product confusion leads to most customers buying the least costly and lowest spec version and tossing the machine in the bin if it's unsuitable. Any undifferentiated "intel core i" processor is just the same as any other undifferentiated "intel core i" processor to the layman. Most laptops at retail aren't even labeled with the cpu model number. If the cpu is slow it's just because Intel products are crap.

userbinator 5y ago
I believe this practice actually started out with the enterprise stuff, i.e. the "large corporate buyers", and if anything, they are more likely to not care or embrace it.
monocasa 5y ago
Hyperscalers have already been bought in for years. There's custom Amazon, Google, etc. silicon in Xeons that either gets fused off in the general case or needs a secret MSR knock to turn on, or what have you.
- kmeisthax 5y ago
  Is there any public documentation as to what these features would be, and has anyone discovered ways to turn them on? I remember hearing a few years back about Intel rolling a special weird Xeon SKU with VNNI/bfloat16 tacked onto a different core architecture that got cancelled, specifically because Facebook backed out of the project. However, I think that was intended to be a publicly-available SKU that just didn't make monetary sense to produce in volume. There's also the occasional off-roadmap SKUs that eventually wind up on eBay or wherever... but my impression was that these were just ways for Intel to obfuscate whatever special volume deals they were giving certain hyperscalars.
  - vsrinivas 5y ago
    Intel Cooper Lake - it's a 14 nm server part that fit a particular niche (4S support, VNNI/bfloat16). It was partly due to FB, but also partly to provide a platform that was larger than ICX server could / can.
paulmd 5y ago
enterprise customers have been the leading group demanding software-defined "capacity on demand" features for literally decades

userbinator 5y ago

I believe this practice actually started out with the enterprise stuff, i.e. the "large corporate buyers", and if anything, they are more likely to not care or embrace it.

monocasa 5y ago

Hyperscalers have already been bought in for years. There's custom Amazon, Google, etc. silicon in Xeons that either gets fused off in the general case or needs a secret MSR knock to turn on, or what have you.

kmeisthax 5y ago
Is there any public documentation as to what these features would be, and has anyone discovered ways to turn them on? I remember hearing a few years back about Intel rolling a special weird Xeon SKU with VNNI/bfloat16 tacked onto a different core architecture that got cancelled, specifically because Facebook backed out of the project. However, I think that was intended to be a publicly-available SKU that just didn't make monetary sense to produce in volume. There's also the occasional off-roadmap SKUs that eventually wind up on eBay or wherever... but my impression was that these were just ways for Intel to obfuscate whatever special volume deals they were giving certain hyperscalars.
- vsrinivas 5y ago
  Intel Cooper Lake - it's a 14 nm server part that fit a particular niche (4S support, VNNI/bfloat16). It was partly due to FB, but also partly to provide a platform that was larger than ICX server could / can.

paulmd 5y ago

enterprise customers have been the leading group demanding software-defined "capacity on demand" features for literally decades