From: Wilco Dijkstra <Wilco.Dijkstra@arm.com>
To: paul zimmermann <Paul.Zimmermann@inria.fr>
Cc: "libc-alpha@sourceware.org" <libc-alpha@sourceware.org>,
"adhemerval.zanella@linaro.org" <adhemerval.zanella@linaro.org>
Subject: Re: [PATCH 1/5] benchtests: Add exp10f benchmark
Date: Wed, 15 Apr 2020 08:51:36 +0000 [thread overview]
Message-ID: <DB8PR08MB503616C37D99F58F09066ADE83DB0@DB8PR08MB5036.eurprd08.prod.outlook.com> (raw)
In-Reply-To: <mwtv1ltbq2.fsf@tomate.loria.fr>
Hi Paul,
>> Also keeping the workload line means we get more accurate benchmark
>> results, including throughput and latency.
>
> however, it makes difficult to compare to other functions which don't have
> the workload line, for example sinf. Why don't all float functions have the
> same workload?
Switching off the workload feature doesn't make it possible to compare performance
between different math functions - the input files are different anyway.
What we want to know is how each math function performs on typical inputs.
The workloads contain traces from real applications and the workload feature
processes these traces in a realistic way to get accurate throughput and latency
results. So the datasets are unique for each math function.
The performance results of the workloads can be compared since it is a measure
how each function performs on typical inputs. Comparing performance of sinf(0)
with logf(0) on the other hand is useless.
Cheers,
Wilco
next prev parent reply other threads:[~2020-04-15 8:51 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-04-14 11:15 Wilco Dijkstra
2020-04-14 12:18 ` Adhemerval Zanella
2020-04-15 8:05 ` paul zimmermann
2020-04-15 8:51 ` Wilco Dijkstra [this message]
-- strict thread matches above, loose matches on Subject: below --
2020-06-18 16:17 Wilco Dijkstra
2020-04-09 19:59 Adhemerval Zanella
2020-04-16 20:41 ` Adhemerval Zanella
2020-04-29 17:11 ` Adhemerval Zanella
2020-05-20 20:18 ` Adhemerval Zanella
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=DB8PR08MB503616C37D99F58F09066ADE83DB0@DB8PR08MB5036.eurprd08.prod.outlook.com \
--to=wilco.dijkstra@arm.com \
--cc=Paul.Zimmermann@inria.fr \
--cc=adhemerval.zanella@linaro.org \
--cc=libc-alpha@sourceware.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).