From: Daniel Stone <daniel@fooishbar.org>
To: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Arnd Bergmann <arnd@arndb.de>, Dave Airlie <airlied@gmail.com>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Maciej Kwapulinski <maciej.kwapulinski@linux.intel.com>,
Jonathan Corbet <corbet@lwn.net>,
Derek Kiernan <derek.kiernan@xilinx.com>,
Dragan Cvetic <dragan.cvetic@xilinx.com>,
Andy Shevchenko <andy.shevchenko@gmail.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
"open list:DOCUMENTATION" <linux-doc@vger.kernel.org>,
DRI Development <dri-devel@lists.freedesktop.org>
Subject: Re: [PATCH v3 00/14] Driver of Intel(R) Gaussian & Neural Accelerator
Date: Mon, 17 May 2021 20:32:43 +0100 [thread overview]
Message-ID: <CAPj87rMqYNdHMT5v9fiMuDzcB8462nJuthB9To70JOsORgxk=w@mail.gmail.com> (raw)
In-Reply-To: <503d101d-7273-757a-2809-e272db93c45d@suse.de>
Hi,
On Mon, 17 May 2021 at 20:12, Thomas Zimmermann <tzimmermann@suse.de> wrote:
> Am 17.05.21 um 09:40 schrieb Daniel Vetter:
> > We have, it's called drivers/gpu. Feel free to rename to drivers/xpu or
> > think G as in General, not Graphisc.
>
> I hope this was a joke.
>
> Just some thoughts:
>
> AFAICT AI first came as an application of GPUs, but has now
> evolved/specialized into something of its own. I can imagine sharing
> some code among the various subsystems, say GEM/TTM internals for memory
> management. Besides that there's probably little that can be shared in
> the userspace interfaces. A GPU is device that puts an image onto the
> screen and an AI accelerator isn't.
But it isn't. A GPU is a device that has a kernel-arbitrated MMU
hosting kernel-managed buffers, executes user-supplied compiled
programs with reference to those buffers and other jobs, and informs
the kernel about progress.
KMS lies under the same third-level directory, but even when GPU and
display are on the same die, they're totally different IP blocks
developed on different schedules which are just periodically glued
together.
> Treating both as the same, even if
> they share similar chip architectures, seems like a stretch. They might
> evolve in different directions and fit less and less under the same
> umbrella.
Why not? All we have in common in GPU land right now is MMU + buffer
references + job scheduling + synchronisation. None of this has common
top-level API, or even a common top-level model. It's not just ISA
differences, but we have very old-school devices where the kernel
needs to register fill on every job, living next to middle-age devices
where the kernel and userspace co-operate to fill a ring buffer,
living next to modern devices where userspace does some stuff and then
the hardware makes it happen with the bare minimum of kernel
awareness.
Honestly I think there's more difference between lima and amdgpu then
there is between amdgpu and current NN/ML devices.
Cheers,
Daniel
next prev parent reply other threads:[~2021-05-17 19:33 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-13 11:00 [PATCH v3 00/14] Driver of Intel(R) Gaussian & Neural Accelerator Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 01/14] intel_gna: add driver module Maciej Kwapulinski
2021-05-13 11:13 ` Greg Kroah-Hartman
2021-05-13 11:00 ` [PATCH v3 02/14] intel_gna: add component of hardware operation Maciej Kwapulinski
2021-05-13 11:15 ` Greg Kroah-Hartman
2021-05-13 11:00 ` [PATCH v3 03/14] intel_gna: read hardware info in the driver Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 04/14] intel_gna: add memory handling Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 05/14] intel_gna: initialize mmu Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 06/14] intel_gna: add hardware ids Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 07/14] intel_gna: add request component Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 08/14] intel_gna: implement scoring Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 09/14] intel_gna: add a work queue to process scoring requests Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 10/14] intel_gna: add interrupt handler Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 11/14] intel_gna: add ioctl handler Maciej Kwapulinski
2021-05-13 11:24 ` Greg Kroah-Hartman
2021-05-14 8:20 ` Maciej Kwapulinski
2021-05-14 8:32 ` Greg Kroah-Hartman
2021-05-24 10:43 ` Maciej Kwapulinski
2021-05-24 10:49 ` Greg Kroah-Hartman
2021-05-25 7:50 ` Maciej Kwapulinski
2021-05-13 14:16 ` Matthew Wilcox
[not found] ` <20210514101253.1037-1-hdanton@sina.com>
2021-05-14 15:06 ` Maciej Kwapulinski
2021-05-13 11:00 ` [PATCH v3 12/14] intel_gna: add a 'misc' device Maciej Kwapulinski
2021-05-13 11:18 ` Greg Kroah-Hartman
2021-05-13 17:06 ` Maciej Kwapulinski
2021-05-13 17:15 ` Greg Kroah-Hartman
2021-05-13 11:00 ` [PATCH v3 13/14] intel_gna: add file operations to " Maciej Kwapulinski
2021-05-13 11:19 ` Greg Kroah-Hartman
2021-05-13 11:00 ` [PATCH v3 14/14] intel_gna: add power management Maciej Kwapulinski
2021-05-14 8:34 ` [PATCH v3 00/14] Driver of Intel(R) Gaussian & Neural Accelerator Greg Kroah-Hartman
2021-05-14 9:00 ` Arnd Bergmann
2021-05-17 7:40 ` Daniel Vetter
2021-05-17 8:00 ` Greg Kroah-Hartman
2021-05-17 8:49 ` Daniel Vetter
2021-05-17 8:55 ` Greg Kroah-Hartman
2021-05-17 9:12 ` Daniel Vetter
2021-05-17 18:04 ` Dave Airlie
2021-05-17 19:12 ` Thomas Zimmermann
2021-05-17 19:23 ` Alex Deucher
2021-05-17 19:39 ` Daniel Vetter
2021-05-17 19:49 ` Thomas Zimmermann
2021-05-17 20:00 ` Daniel Vetter
2021-05-17 20:15 ` Thomas Zimmermann
2021-05-17 19:32 ` Daniel Stone [this message]
2021-05-17 20:10 ` Thomas Zimmermann
2021-05-17 21:24 ` Daniel Vetter
2021-05-17 21:36 ` Dave Airlie
2021-06-16 7:38 ` Maciej Kwapulinski
2022-06-20 9:49 ` maciej.kwapulinski
2022-06-20 9:56 ` Greg KH
2022-06-20 10:08 ` Maciej Kwapulinski
2022-06-20 10:26 ` Greg KH
2022-06-25 17:25 ` Daniel Vetter
2021-05-20 11:58 ` Linus Walleij
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='CAPj87rMqYNdHMT5v9fiMuDzcB8462nJuthB9To70JOsORgxk=w@mail.gmail.com' \
--to=daniel@fooishbar.org \
--cc=airlied@gmail.com \
--cc=andy.shevchenko@gmail.com \
--cc=arnd@arndb.de \
--cc=corbet@lwn.net \
--cc=derek.kiernan@xilinx.com \
--cc=dragan.cvetic@xilinx.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=gregkh@linuxfoundation.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=maciej.kwapulinski@linux.intel.com \
--cc=tzimmermann@suse.de \
/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 NNTP newsgroup(s).