top | item 46706916

(no title)

owlbite | 1 month ago

I think calling VLIW "an adandoned design" is somewhat of an exaggeration, such architectures are pretty common for embedded audio processing.

discuss

order

matt_d|1 month ago

Worth adding on that note:

From JAX to VLIW: Tracing a Computation Through the TPU Compiler Stack, https://patricktoulme.substack.com/p/from-jax-to-vliw-tracin...

Google’s Training Chips Revealed: TPUv2 and TPUv3, HotChips 2020, https://hc32.hotchips.org/assets/program/conference/day2/Hot...

Ten Lessons From Three Generations Shaped Google’s TPUv4i, ISCA 2021, https://gwern.net/doc/ai/scaling/hardware/2021-jouppi.pdf

mike_hearn|1 month ago

Thanks, that JAX writeup was interesting.

mike_hearn|1 month ago

Sure. I did mention DSPs. But how many people write code for DSPs?

HarHarVeryFunny|1 month ago

x86-64 SSE and AVX are also SIMD

vel0city|1 month ago

SIMD and VLIW are somewhat similar but very different in the end.