User Profile

[coder]

That's not true. I agree with L_A_G that the bulk of blockchain use is for illicit purposes, but there are plenty of problems in the world where you...

[coder]

Don't most of those projects just reuse Etherium or other established implementations, in which case the nodes could just use existing ASICs already optimized...

[coder]

At lower clock speeds, though! One developer found the impact on clock speed was so dramatic that using AVX-512 for crypto resulted in a net decrease...

[coder]

Aside from the penalty you incur in mixed 128-bit or 256-bit + 512-bit workloads, there's the unavoidable downside of bigger registers and larger context....

[coder]

Sadly, AVX-512 is broken by design.

If just one library call executes just one AVX-512 instruction, suddenly every SSE and AVX operation...

[coder]

You're far from the only one here to have contributed to open source, however you do appear to be in the minority that believes it entitles them to be...

[coder]

Same old tired trolling.

Does anyone force you to use Linux? You have freedom to go and use proprietary shit, so why not do it and leave...

[coder]

Again, if you look at the TILERELEASE instruction, that talks about putting AMX in some sort of "INIT" state, which presumably could disable...

[coder]

It's obviously not for that use case.


A generalized matrix multiply requires repeated access to each row and column, which would be...

[coder]

I didn't say it didn't do dot products, just that the raw computation isn't the key point of it.


Unless what you actually need is...

[coder]

Not to cast myself as an AMX proponent, but it's more constructive to think of it like crypto-acceleration extensions. In both cases, they're very specialized...

[coder]

If you think it's about computing dot-products, you're missing the point. This is really about optimizing data-movement, and quite obviously for deep...

[coder]

More on the state of the Vulkan backend, here: https://github.com/Tencent/ncnn/blob...ncnn-vulkan.md

It...

[coder]

Yeah, take a closer look at both articles. The first was super heavy-weight, with results taking on the order of tens of seconds. This article features...

[coder]

I'm not sure how to tell. They have int8 versions of some, which should run well on Radeon VII and Pascal + Turing Nvidia GPUs, though it seems those...

[coder]

Would be nice to see benchmarks with batching - especially on theses smaller networks. That would be somewhat more realistic, as well as overcome some...