4 ms·

Many have tried to derive an optimal instruction set, and RISC-V is such an attempt. But for now trying is failing, and one can only chose the way one will fail

Many have tried to derive an optimal instruction set, and RISC-V is such an attempt. But for now trying is failing, and one can only chose the way one will fail. With the possible exception of only the Mill architecture, but that architecture requires a grand scale. One common bottleneck, orienting instruction set design, is the need to pull in constants (indeed the instructions themselves are). One could theoretically share them between functions, and think of smart ways to reduce the amount of code. But in practice this is futile, as the modern CPU is a race car on a highway, if not a train on rails, so it can't be bothered to stop and deal with that kind of smartness. There's just to much inertia behind you... once you approach the speed of light. The road itself should have all information needed to let the CPU drive the correct track. And no special cases. In addition, almost all CPU's are designed to execute a C type machine abstraction, because that's the paradigm. Though I root for the Mill, my guess is we'll only see a different instruction set once the paradigm shifts towards another execution model, likely one that is less stack bound.

childintime 4y ago
Continuing.. For all the performance a modern CPU has, and all the visual magic it can do, for most of us our PC's are still just as fast as 20 years ago. That is something that needs mending first. 100x faster, yet where did all those cycles go? Seems pathological to me. Tech has been operating like politics: ever more promises. The vector that realizes itself from those promises is an increase of dependency, an increase of control and surveillance, and an erosion of self-reliance. They become stronger, not us. The paradigm forces you to think that the heroes of the paradigm, like big tech, are the solution, while they equally increase the problem. When a "Jedi instruction set" comes it is likely OoO like a christmas child.
- Nomentatus 4y ago
  I wish someone more competent than myself had taken you up on this. Now that we have big die sizes, is it possible to use some of that real estate to create a fast core with a different instruction set (with much less to worry about) that we can slip into with maybe just one app at a time and let 'er just zoom? If we had to reserve some ram exclusively to that core even when not in use; I'd still want it for the times I have a very demanding task I'm waiting on. This doesn't address all the bloat of course; but I'd like it. OoO ? Overlooked? Overwritten? Overlooked or Overwritten?
- Someone 4y ago
  > for most of us our PC's are still just as fast as 20 years ago. That is something that needs mending first. 100x faster, yet where did all those cycles go? Seems pathological to me. Nobody would buy a PC that’s 100 times as fast but doesn’t have new abilities. 20 years ago, Blu-ray didn’t exist yet, and it featured unprecedented full HD video. Today, 4K more or less is the norm. The original iPhone (about 15 years old) had a 480 × 320 pixel screen. The latest ones have 1000+ × 2500+, at ¿quadruple? the screen refresh rate and ¿double? the number of bits per pixel. Progress is even more enormous with cameras. Resolution had gone up relatively moderately (for iPhone from 2 megapixel ⇒ 48 megapixel), but the software behind it couldn’t realistically run on 20 year old hardware.

childintime 4y ago

Continuing.. For all the performance a modern CPU has, and all the visual magic it can do, for most of us our PC's are still just as fast as 20 years ago. That is something that needs mending first. 100x faster, yet where did all those cycles go? Seems pathological to me. Tech has been operating like politics: ever more promises. The vector that realizes itself from those promises is an increase of dependency, an increase of control and surveillance, and an erosion of self-reliance. They become stronger, not us. The paradigm forces you to think that the heroes of the paradigm, like big tech, are the solution, while they equally increase the problem. When a "Jedi instruction set" comes it is likely OoO like a christmas child.

Nomentatus 4y ago
I wish someone more competent than myself had taken you up on this. Now that we have big die sizes, is it possible to use some of that real estate to create a fast core with a different instruction set (with much less to worry about) that we can slip into with maybe just one app at a time and let 'er just zoom? If we had to reserve some ram exclusively to that core even when not in use; I'd still want it for the times I have a very demanding task I'm waiting on. This doesn't address all the bloat of course; but I'd like it. OoO ? Overlooked? Overwritten? Overlooked or Overwritten?
Someone 4y ago
> for most of us our PC's are still just as fast as 20 years ago. That is something that needs mending first. 100x faster, yet where did all those cycles go? Seems pathological to me. Nobody would buy a PC that’s 100 times as fast but doesn’t have new abilities. 20 years ago, Blu-ray didn’t exist yet, and it featured unprecedented full HD video. Today, 4K more or less is the norm. The original iPhone (about 15 years old) had a 480 × 320 pixel screen. The latest ones have 1000+ × 2500+, at ¿quadruple? the screen refresh rate and ¿double? the number of bits per pixel. Progress is even more enormous with cameras. Resolution had gone up relatively moderately (for iPhone from 2 megapixel ⇒ 48 megapixel), but the software behind it couldn’t realistically run on 20 year old hardware.