From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id 027EAA034C for ; Tue, 22 Feb 2022 20:35:27 +0100 (CET) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id 58CDB41144; Tue, 22 Feb 2022 20:35:25 +0100 (CET) Received: from mail-vs1-f50.google.com (mail-vs1-f50.google.com [209.85.217.50]) by mails.dpdk.org (Postfix) with ESMTP id B9FA041143 for ; Tue, 22 Feb 2022 20:35:23 +0100 (CET) Received: by mail-vs1-f50.google.com with SMTP id q9so749797vsg.2 for ; Tue, 22 Feb 2022 11:35:23 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=8/Jzp3GUpOpHNGZiy6lRdl+KNc3PFG5MHOmkP3AHtOg=; b=bbB5wkAba8X0br5QdJEvyEZolAgr31maia1BNOReK8aLgsPoeOroFCgCfBwB51Mwey yrZ9eRwpEAkSkOs3jHtHbQIg37NpqpOcPDsBAuS1v29+LWbndzNC/zcOfmwAHzpK5WxF llJTTodJKfLAObdbTh1T9u4Nl1MnGOHke1xSeU0aZ+hX14RfABHpLqmrRf11Yl/eJV+N bsAUnUMl6SklhP9sO2LqjW/bluanVw4kaBRyukJiBj4KvnWemgQ2HXWgX2WO65Jo0ipg qYQgNiFt/NzQlTDfYKzWYZdg0A5Ewnl4GEBXMzYaEb1hRI/Eqo+F7j8DWNqQhCOY1aS0 sxQQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=8/Jzp3GUpOpHNGZiy6lRdl+KNc3PFG5MHOmkP3AHtOg=; b=5irXcyN7jTiA62zwq5kUXboqNoKCyKz8tjydzm83/uYf9Af5CzPm/vFm/qijCf6kD6 K22Xbh9gWknUL8VrJhV4bzaM3L0vYunudD8kzyV3j8ync1/HGTb13T3IB82UaXDkC7ox ShWGifFbF9DAP6MM/HG7OsBLCC1y09haDHAgUGf/keJdgD6fvPHHveBiyvZoYWJevq5a 7ZIi9iwJS/N/yRkK94c6wZWybNpH29hZERv+7UwfGNT6U7airAK5s48FiRWK0JH67rpW gybZdz3Y0WISNMHi1SId9hiHUkTH1FOXc6whXcWobdcIIUzOsiptw8AK3Rhg/HhKxyDt 2xLA== X-Gm-Message-State: AOAM531hz74eCjlebh4ankebpt7da72fuX9MGkPc2pUVuaO0FokOJojy eEK1Gch8JUBwHt2GL0jERtSh5jWiCu3S8LeCu05UTDM2q8w= X-Google-Smtp-Source: ABdhPJxKeT0WnItiAWjcnSQ8SaerRPa5PniCVvdrN49wTr30c0A3wpC+8SezfoKRKpzqQmKgHhPDg5a5VjEIBRFhOVw= X-Received: by 2002:a05:6102:3f49:b0:31a:fe9a:81bb with SMTP id l9-20020a0561023f4900b0031afe9a81bbmr10230572vsv.47.1645558523059; Tue, 22 Feb 2022 11:35:23 -0800 (PST) MIME-Version: 1.0 References: <20220221090852.05f7819e@hermes.local> <20220221174902.115c1eed@hermes.local> In-Reply-To: <20220221174902.115c1eed@hermes.local> From: Staffan Wiklund Date: Tue, 22 Feb 2022 20:35:13 +0100 Message-ID: Subject: Re: Are Intel CPUs better than AMD CPUs for DPDK applications? To: Stephen Hemminger Cc: users Content-Type: multipart/alternative; boundary="00000000000039626d05d8a07338" X-BeenThere: users@dpdk.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: DPDK usage discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: users-bounces@dpdk.org --00000000000039626d05d8a07338 Content-Type: text/plain; charset="UTF-8" Thanks Stephen again for your help. I will check out what you mentioned about CPU features. By the way, regarding my earlier question about using DPDK memif in a Docker container. I made a patch in the memif driver of DPDK so it uses a TCP socket instead of a UNIX domain socket. This made it possible to communicate between a Docker container and an application outside the container using DPDK memif. This patch only works without the "zero-copy" option for DPDK memif. But the throughput of DPDK memif without zero-copy is rather good anyway. Regards Staffan Den tis 22 feb. 2022 kl 02:49 skrev Stephen Hemminger < stephen@networkplumber.org>: > On Mon, 21 Feb 2022 21:28:08 +0100 > Staffan Wiklund wrote: > > > Stephen, thanks for your answer. > > I realize the statement is very vague. > > I was thinking of if there was something common in the design of Intel > and > > AMD CPUs respectively that has an impact on their use by DPDK > applications. > > Do you know if there is such a common design difference between Intel and > > AMD CPUs or is it just a matter of using an Intel or AMD CPU with the > > requested performance? > > > > Regards > > Staffan > > I am not a CPU expert. But compare memory bandwidth, clock rate, PCI > Express version > and support for AVX support (for some features). There a few places in DPDK > that can use AVX512 but it is limited > https://doc.dpdk.org/guides/howto/avx512.html > > Also there is a tradeoff with more cores, NUMA, etc as well as cost. > > Don't believe simple tribal knowledge, you need to look under the covers. > --00000000000039626d05d8a07338 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
Thanks Stephen again for your help.
I will check out w= hat you mentioned about CPU features.

By the way, = regarding my earlier question about using DPDK memif in a Docker container.=
I made a patch in the memif driver of DPDK so it uses a TCP sock= et instead of a UNIX domain socket.
This made it possible to comm= unicate between a Docker container and an application outside the container= using DPDK memif.
This patch only works without the "zero-c= opy" option for DPDK memif.
But the throughput of DPDK memif= without zero-copy is rather good anyway.

Regards<= /div>
Staffan


=
Den tis 22 feb. 2022 kl 02:49 skrev S= tephen Hemminger <stephen@= networkplumber.org>:
On Mon, 21 Feb 2022 21:28:08 +0100
Staffan Wiklund <staffan491@gmail.com> wrote:

> Stephen, thanks for your answer.
> I realize the statement is very vague.
> I was thinking of if there was something common in the design of Intel= and
> AMD CPUs respectively that has an impact on their use by DPDK applicat= ions.
> Do you know if there is such a common design difference between Intel = and
> AMD CPUs or is it just a matter of using an Intel or AMD CPU with the<= br> > requested performance?
>
> Regards
> Staffan

I am not a CPU expert. But compare memory bandwidth, clock rate, PCI Expres= s version
and support for AVX support (for some features). There a few places in DPDK=
that can use AVX512 but it is limited
https://doc.dpdk.org/guides/howto/avx512.html

Also there is a tradeoff with more cores, NUMA, etc as well as cost.

Don't believe simple tribal knowledge, you need to look under the cover= s.
--00000000000039626d05d8a07338--