[sharpneli]

One reason is that they are already here. It's just called the GPU which happens to be way more parallel than puny 128 cores. That's a major reason why just bunch of low power CPUs are not really a thing, as they get trashed by GPUs.

What we are seeing is a hierarchy of processing cores. It's already visible even in the desktop market. Intel has their efficiency cores. Apple has the power and efficiency cores. And the GPU on top. So we already have 3 different stages here, at the top the least amount of cores but the best single threaded performance going all the way down to, let me check how many threads, almost 10000 threads on a flagship consumer GPU running simultaneously right now.

[moonchrome]

Except GPU cores are nothing like CPU cores, GPGPU was around 10 years ago, and whatever scala language/framework features he was shilling are probably useless for that kind of programming - so that's a failed prediction.

[sharpneli]

On that I fully agree. And the reason why scala approach doesn't work is because it's horrendously inefficient. What actually does work at massive scale is the GPU style workload.

Intel tried with their Larrabee on what happens if they just toss in ton of traditional low performance CPU cores. It failed to perform. It's really hard to beat the modern SIMT style GPUs when it comes to massively parallel computation.

[nine_k]

Indeed. MCUs are even further in this regard, combining 32-bit and 64-bit cores.