“they don't seem to have simulated macro-op fusion at RTL level, much less in an FPGA or a MOSIS run.”
The Renewed Case for the Reduced Instruction Set Computer: Avoiding ISA Bloat with Macro-Op Fusion for RISC-V
论证 RISC-V 配合宏操作融合,可在性能和代码密度上追平 CISC
报告在 SPEC CINT2006 上比较 ARMv7、ARMv8、IA-32、x86-64 与开放的 RISC-V RV64G、RV64GC 的动态指令数和取指字节数。RV64G 平均比 x86-64 多执行 16% 的指令,但与 x86-64 退休的微操作数相差在 2% 以内;压缩变体 RV64GC 是最密的 ISA,取指字节比 x86-64 少 8%。利用宏操作融合可再把有效指令数平均减少 5.4%,因此无需再往 ISA 里添加更多指令。
HN 上的人反复引用它来论证 ISA 本身对性能影响很小、简单正交的 ISA 加上宏操作融合就能追平 x86 和 ARM。
HN 书评
推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及“I think RISC-V is showing that RISC ISAs can still compete in this space: The Renewed Case for the Reduced Instruction Set Computer.”
“Load/store multiple, for starters. Even just load-pair-increment-address is a pain as it requires 3 write ports.”
“I just so happen to have written a tech report on this very topic! The Renewed Case for the Reduced Instruction Set Computer”
“The ISA is going to have close to zero impact on a processor's performance.”
“Yeah, I drew my conclusions from your papers.”
看全部 11 条 →
“So with RISC-V GC + macro fusion of a few common idioms you get slightly less micro-ops than x86-64, armv7, or armv8.”
“With a few simple macro-op fusion pairs, it also comes right up to x86_64 in dynamic uOp count.”
“Also, a huge amount of instructions end up being used very little, there are a few "hotspot" instructions that make up a huge percent of actual instructions executed.”
“is a widely cited paper that argues that simple and orthogonal ISA's are preferable, as fusion can fix up those common cases”
“The only other source I know of was this 2016 paper that suggested specific macro fusions.”
- Avoiding ISA Bloat with Macro-Op Fusion for RISC-V ↗phkahler · 2016-09-26 · 3 分