5 ms·

> Your "painstakingly optimized with SIMD" doesn't apply much to those 4 year old programs. What's this then? https://github.com/greensoftwarelab/Energy-Langua

> Your "painstakingly optimized with SIMD" doesn't apply much to those 4 year old programs. What's this then? https://github.com/greensoftwarelab/Energy-Languages/blob/master/C/n-body/nbody.gcc-4.c#L100 https://github.com/greensoftwarelab/Energy-Languages/blob/ma... There are some explicit SIMD instructions, and some programs written to allow autovectorization. > Never throw the baby out with the bath water. /eyeroll There's no baby. There's no significant investment in applying equal levels of optimizations to the programs across languages. TypeScript vs JavaScript is a telling example but hardly the only one.

igouy 5y ago
> There are some explicit SIMD instructions, and some programs written to allow autovectorization. Yes there were some way back then. Many more have been contributed since then. > … equal levels of optimizations to the programs across languages. How do you propose to measure that exactly? Or will we like Goldilocks know it when we see it? Like these — https://benchmarksgame-team.pages.debian.net/benchmarksgame/performance/simple.html https://benchmarksgame-team.pages.debian.net/benchmarksgame/...
- scottlamb 5y ago
  Oh, it just clicked for me that you are the Game' maintainer. > How do you propose to measure that exactly? I don't have a proposal, other than to avoid overselling. I like that the Game's website has the text "If the fastest programs are manually vectorized SIMD, does the host language matter? You might be more interested in less optimised programs." The "Energy Efficiency across Programming Languages" paper though is overselling its rigor.
  - igouy 5y ago
    Yeah, sorry about that, peoples comments usually seem more open before they realize. It's difficult to find a balance that works — otoh for some hints give a context, otoh for others hints give stones to throw ;-)

igouy 5y ago

> There are some explicit SIMD instructions, and some programs written to allow autovectorization. Yes there were some way back then. Many more have been contributed since then. > … equal levels of optimizations to the programs across languages. How do you propose to measure that exactly? Or will we like Goldilocks know it when we see it? Like these — https://benchmarksgame-team.pages.debian.net/benchmarksgame/performance/simple.html https://benchmarksgame-team.pages.debian.net/benchmarksgame/...

scottlamb 5y ago
Oh, it just clicked for me that you are the Game' maintainer. > How do you propose to measure that exactly? I don't have a proposal, other than to avoid overselling. I like that the Game's website has the text "If the fastest programs are manually vectorized SIMD, does the host language matter? You might be more interested in less optimised programs." The "Energy Efficiency across Programming Languages" paper though is overselling its rigor.
- igouy 5y ago
  Yeah, sorry about that, peoples comments usually seem more open before they realize. It's difficult to find a balance that works — otoh for some hints give a context, otoh for others hints give stones to throw ;-)