By: Wilco (Wilco.Dijkstra.delete@this.ntlworld.com), March 6, 2015 4:04 am
Room: Moderated Discussions
David Kanter (dkanter.delete@this.realworldtech.com) on March 6, 2015 12:42 am wrote:
> > A57 is definitely more than 20% faster at the same clock, as you can easily conclude
> > from various benchmarks (eg. Geekbench shows 40% single threaded gain overall
> > in AArch32 mode between Galaxy S5 and Galaxy S6 at the same clock).
>
> That's a bogus comparison since Geekbench contains lots of AES
> stuff, and ARM64 contains special instructions for AES.
Across the whole suite it makes maybe ~5% difference. So my point that A57 shows far more than 20% gain on GB stands.
> What's the speedup on GCC?
I don't have numbers at hand but A57 does pretty well on SPEC and gains a lot from the extra registers in 64-bit mode.
Btw while on the topic of benchmarks, you might want to edit your recent article about CoreMark-Pro - the one subtest you claimed is most representative spends almost all of its time (~80%) in one tiny string function. So it's trivial to claim 5x gains with library optimizations...
Wilco
> > A57 is definitely more than 20% faster at the same clock, as you can easily conclude
> > from various benchmarks (eg. Geekbench shows 40% single threaded gain overall
> > in AArch32 mode between Galaxy S5 and Galaxy S6 at the same clock).
>
> That's a bogus comparison since Geekbench contains lots of AES
> stuff, and ARM64 contains special instructions for AES.
Across the whole suite it makes maybe ~5% difference. So my point that A57 shows far more than 20% gain on GB stands.
> What's the speedup on GCC?
I don't have numbers at hand but A57 does pretty well on SPEC and gains a lot from the extra registers in 64-bit mode.
Btw while on the topic of benchmarks, you might want to edit your recent article about CoreMark-Pro - the one subtest you claimed is most representative spends almost all of its time (~80%) in one tiny string function. So it's trivial to claim 5x gains with library optimizations...
Wilco
