Re:BOOM v2: an open-source out-of-order RISC-V core
by dnautics on Tuesday September 26, 2017 @07:33PM
(#15343594)
The field is littered with the tombstones of people that have tried to make compilers for insert any foreign architecture. Obviously in-order CPUs are a challenge on top of that. I do think there is a real 'go-to-market' problem. If you think about it this is kind of solved by Nvidia (with CUDA). GPUs are in-order computational arrays with custom compiler.
[ Parent
| on HN ]
Re:BOOM v2: an open-source out-of-order RISC-V core
by wbl on Tuesday September 26, 2017 @09:01PM
(#15344063)
GPUs are vector processors under another name, that hide the latency of memory differently. They don't need to use the sort of compiler tricks you would with CPU code.
Re:BOOM v2: an open-source out-of-order RISC-V core
by deepnotderp on Wednesday September 27, 2017 @01:33AM
(#15345151)
Yeah, but GPUs target only a small subset of the workloads and they are very hard to program even then. I prefer processors like the Rex Neo or Adapteva Epiphany in that respect actually.
Re:BOOM v2: an open-source out-of-order RISC-V core
by dnautics on Wednesday September 27, 2017 @02:42AM
(#15345408)
have you tried programming rex neo?
Re:BOOM v2: an open-source out-of-order RISC-V core
by deepnotderp on Wednesday September 27, 2017 @02:53AM
(#15345452)
Nope, I know you have experience with it though, so perhaps you can help there ;)
I've toyed around with the older Epiphanies though, and they were definitely easier than GPUs in many respects.
Re:BOOM v2: an open-source out-of-order RISC-V core
by AlphaSite on Tuesday September 26, 2017 @09:54PM
(#15344280)
Also, as I understand it, GPUs are now out of order as well, aren’t they?
Re:BOOM v2: an open-source out-of-order RISC-V core
by TomV1971 on Tuesday September 26, 2017 @11:32PM
(#15344714)
No, they are not.
OoO is much more expensive to implement than simply scheduling a different warp to hide latency.
Re:BOOM v2: an open-source out-of-order RISC-V core
by deepnotderp on Wednesday September 27, 2017 @01:27AM
(#15345139)
Absolutely not.
Maybe you're confusing the GPU's threading techniques for OoO?