This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] Add math benchmark latency test
- From: Siddhesh Poyarekar <siddhesh at gotplt dot org>
- To: Arjan van de Ven <arjan at linux dot intel dot com>, Wilco Dijkstra <Wilco dot Dijkstra at arm dot com>, "libc-alpha at sourceware dot org" <libc-alpha at sourceware dot org>
- Cc: nd <nd at arm dot com>
- Date: Wed, 16 Aug 2017 19:56:19 +0530
- Subject: Re: [PATCH] Add math benchmark latency test
- Authentication-results: sourceware.org; auth=none
- References: <0e008f2e-f41a-1bb8-803c-2f798e2c3541@gotplt.org> <d8612b78-7a02-40e8-d8bd-9b221eff4b7b@linux.intel.com>
On Wednesday 16 August 2017 07:53 PM, Arjan van de Ven wrote:
> well not really...
>
> I've been working on making expf() faster for x86 (see HJ's email
> earlier), and
> with a massive out of order/pipelined cpu, latency and throughput are
> very distinct things.
> expf() can run at a throughput of somewhere in the 10 to 11 cycles
> range, while the latency
> can be in the 45 to 55 cycles range.
> (not trying to do benchmarking here, just wanting to show an order of
> magnitude)
Ah sorry, I should have been clearer - inverse in terms of units, not
value, i.e. throughput is typically expressed in items/unit time while
latency is expressed in time/unit item.
Siddhesh