ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Nim, Ocaml, and more

Re:ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Ni... by kev009 on Saturday December 20, 2014 @08:40AM (#8776606)

I'm not really sure what this data means because amd64 and ARMv7 are ISAs. For instance, you could make a very deep and superscalar ARMv7 chip that blows a typical amd64 out of the water if you sacrifice size and power. Is the intent simply to show that some language backends are not optimized? Otherwise, without something like "These two chips and clock-for-clock or watt-for-watt it looks like this" it seems meaningless.

[ Parent | on HN ]

Re:ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Ni... by logicchains on Saturday December 20, 2014 @08:49AM (#8776617)

It's not meant to compare AMD64 and ARMv7 architectures, it's meant to compare the performance of various language compilers/runtimes on two common AMD64 and ARMv7 chips.

[ Parent | on HN ]

Re:ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Ni... by kev009 on Saturday December 20, 2014 @09:20AM (#8776675)

Not sure why the downvote as that was my request for clarification.

Point being, if you want to benchmark language backends, it doesn't make a lot of sense to cross chips without making mulch-dimensional benchmark (i.e adding in clock or power vectors).

After rereading closely, the intent of the article is saying "50% slowdown for backend X on x86, 70% slowdown for backend X on ARM." The subtle difference is that you are keeping comparisons of slowdown in the family and only making analogy to the other arch by slowdown percentage.

It's still susceptible to ISA implementation (for instance an in order Atom might fare a lot worse for typical backends), but mildly interesting.

[ Parent | on HN ]

Re:ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Ni... by Symmetry on Saturday December 20, 2014 @11:07AM (#8776940)

I expect that the differential performance between the various languages has just as much to do with architectural features such as prefetchers as it does with the ISA itself.

[ Parent | on HN ]