News for hackers. Stuff that matters.

A comment on DRAM Prices in ‘Freefall’. Parent comment

Re:DRAM Prices in ‘Freefall’ by 0815test on Thursday March 07, 2019 @08:10PM (#19334275)
> In the ideal world, a program with loop unrolling and vectorization should always be faster, but in practice, the bigger binary size induces cache misses and may produce a slower problem.

The solution is to use abstract, arbitrary vectors in the ISA instead of packed SIMD and unrolled loops. RISC-V gets this right with their V extension.

[ Parent | on HN ]
Re:DRAM Prices in ‘Freefall’ by greglindahl on Thursday March 07, 2019 @08:23PM (#19334339)
There are many different kinds of loop unrolling, and RISC-V's V extension is not the only one. Itanium had the same issue: the loop unrolling assist in the instruction set was not viewed by many compiler people as being sufficient.
[ Parent | on HN ]
Re:DRAM Prices in ‘Freefall’ by colejohnson66 on Friday March 08, 2019 @03:15PM (#19341746)
ForwardCom has abitrary length vector support. http://www.forwardcom.info/ https://news.ycombinator.com/item?id=19340267
[ Parent | on HN ]