Intent to Ship: WebAssembly SIMD

342 views
Skip to first unread message

Zhi An Ng

unread,
Feb 22, 2021, 1:59:23 PM2/22/21
to blink-dev
Contact emails

gde...@chromium.org, natt...@chromium.org, zh...@chromium.org 


Explainer

https://github.com/WebAssembly/simd/blob/master/proposals/simd/SIMD.md

https://github.com/WebAssembly/simd/blob/master/proposals/simd/W3CTAG-SIMDExplainer.md


Specification

https://github.com/WebAssembly/simd/blob/master/proposals/simd/SIMD.md


Design docs

https://github.com/WebAssembly/simd/tree/master/proposals/simd


Summary

WebAssembly SIMD will expose hardware SIMD instructions to WebAssembly applications in a platform-independent way. The SIMD proposal introduces a new 128-bit value type that can be used to represent different types of packed data, and several vector operations that operate on packed data. 

SIMD can boost performance by exploiting data level parallelism and is also useful when compiling native code to WebAssembly

WebAssembly SIMD successfully progressed to Phase 4 in the WebAssembly W3C process (https://github.com/WebAssembly/meetings/blob/master/main/2021/CG-02-16.md#fixed-width-simd-zhi-an-ng---slides10-min).


Blink component

Blink>JavaScript>WebAssembly


Search tags

WebAssembly, wasm, SIMD, SIMD.js


TAG review

https://github.com/w3ctag/design-reviews/issues/487


TAG review status

Issues addressed


Risks

Interoperability and Compatibility

Gecko: In development (https://bugzilla.mozilla.org/show_bug.cgi?id=1625130)

WebKit: No signal (https://lists.webkit.org/pipermail/webkit-dev/2021-February/031696.html)

Web developers: Positive

Is this feature fully tested by web-platform-tests?

No, but covered by WebAssembly spec tests (https://v8.dev/docs/feature-launch-process#tests and https://github.com/WebAssembly/simd/tree/master/test/core/simd).


Tracking bug

https://bugs.chromium.org/p/v8/issues/detail?id=6020&desc=2


Link to entry on the Chrome Platform Status

https://www.chromestatus.com/feature/6533147810332672


Links to previous Intent discussions

Intent to Experiment: https://groups.google.com/a/chromium.org/g/blink-dev/c/y3z1Q4HnpgY/m/SURHtNTJAQAJ

Intent to Extend Origin Trial 1 (M87-M88):

https://groups.google.com/a/chromium.org/g/blink-dev/c/6ZUHoEUQeEI/m/uzivlQFUAgAJ

Intent to Extend Origin Trial 2 (M89-M90):

https://groups.google.com/a/chromium.org/g/blink-dev/c/bw_RgyH4Xss/m/hPtsEbwRAQAJ

Requesting exception to required breaking period

Developer engagement

https://blog.tensorflow.org/2020/09/supercharging-tensorflowjs-webassembly.html

https://brionv.com/log/2019/03/03/simd-in-webassembly-tales-from-the-bleeding-edge/

https://www.youtube.com/watch?v=9jh4OBeKENs

https://ai.googleblog.com/2020/10/background-features-in-google-meet.html


This intent message was generated by Chrome Platform Status.

Yoav Weiss

unread,
Feb 24, 2021, 9:28:34 AM2/24/21
to Zhi An Ng, blink-dev
That doesn't seem accurate, as I don't see a ton of actual activity on that bug.
have you tried reaching out? Alternatively, were they involved in the WG decision to move this to phase 4?
--
You received this message because you are subscribed to the Google Groups "blink-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/e9438e74-bd3a-4c27-9055-dc34e9fddb3dn%40chromium.org.

Ben Kelly

unread,
Feb 24, 2021, 10:20:43 AM2/24/21
to Yoav Weiss, Zhi An Ng, blink-dev
On Wed, Feb 24, 2021 at 9:28 AM Yoav Weiss <yo...@yoav.ws> wrote:
On Mon, Feb 22, 2021 at 8:00 PM Zhi An Ng <zh...@chromium.org> wrote:
Interoperability and Compatibility

Gecko: In development (https://bugzilla.mozilla.org/show_bug.cgi?id=1625130)

That doesn't seem accurate, as I don't see a ton of actual activity on that bug.
have you tried reaching out? Alternatively, were they involved in the WG decision to move this to phase 4?

FWIW you have to look at the dependencies on that bug.  Last I checked WASM SIMD was working in firefox nightly.  (But, yes, asking on mozilla/standards-positions would be good.) 

Yoav Weiss

unread,
Feb 24, 2021, 10:41:49 AM2/24/21
to Ben Kelly, Zhi An Ng, blink-dev
Good point, I missed that... 

Zhi An Ng

unread,
Feb 24, 2021, 11:59:52 AM2/24/21
to blink-dev, yo...@yoav.ws, zh...@chromium.org, blink-dev, wande...@chromium.org
On Wednesday, February 24, 2021 at 7:41:49 AM UTC-8 yo...@yoav.ws wrote:
On Wed, Feb 24, 2021 at 4:20 PM Ben Kelly <wande...@chromium.org> wrote:
On Wed, Feb 24, 2021 at 9:28 AM Yoav Weiss <yo...@yoav.ws> wrote:
On Mon, Feb 22, 2021 at 8:00 PM Zhi An Ng <zh...@chromium.org> wrote:
Interoperability and Compatibility

Gecko: In development (https://bugzilla.mozilla.org/show_bug.cgi?id=1625130)

That doesn't seem accurate, as I don't see a ton of actual activity on that bug.
have you tried reaching out? Alternatively, were they involved in the WG decision to move this to phase 4?

Yes, they were involved in the WG decision, in particular Lars Hansen was present at that CG meeting (https://github.com/WebAssembly/meetings/blob/master/main/2021/CG-02-16.md#introduction-of-attendees) and has been actively involved in the entire process (he is present in almost all of the meetings, it's a bit manual but you can click into each issue in https://github.com/WebAssembly/simd/issues?q=is%3Aissue+is%3Aclosed+agenda and look at the notes to see the list of participants, or this shows his activity on the issue tracker https://github.com/WebAssembly/simd/issues?q=is%3Aissue+commenter%3Alars-t-hansen).

I've filed an issue on mozilla/standards-positions https://github.com/mozilla/standards-positions/issues/491.

Zhi An Ng

unread,
Feb 24, 2021, 4:31:44 PM2/24/21
to blink-dev, Zhi An Ng, yo...@yoav.ws, Zhi An Ng, blink-dev, Ben Kelly
As a follow up:

Gecko: Shipping, "We support this and will be shipping it, initially for Intel CPUs, almost certainly in FF89; ARM64 support to follow, likely FF90." (https://github.com/mozilla/standards-positions/issues/491#issuecomment-785303882)
Webkit: Positive, "We have not yet implemented the proposal but we generally have a positive opinion of it." (https://lists.webkit.org/pipermail/webkit-dev/2021-February/031705.html)

Yoav Weiss

unread,
Feb 24, 2021, 4:39:02 PM2/24/21
to Zhi An Ng, blink-dev, Zhi An Ng, Ben Kelly
Thanks for collecting those signals!

LGTM1

--
You received this message because you are subscribed to the Google Groups "blink-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.

Manuel Rego Casasnovas

unread,
Feb 25, 2021, 12:32:04 PM2/25/21
to Yoav Weiss, Zhi An Ng, blink-dev, Zhi An Ng, Ben Kelly
LGTM2

On 24/02/2021 22:38, Yoav Weiss wrote:
> Thanks for collecting those signals!
>
> LGTM1
>
> On Wed, Feb 24, 2021 at 10:31 PM Zhi An Ng <zh...@chromium.org
> <mailto:zh...@chromium.org>> wrote:
>
> As a follow up:
>
> Gecko: Shipping, "We support this and will be shipping it, initially
> for Intel CPUs, almost certainly in FF89; ARM64 support to follow,
> likely FF90."
> (https://github.com/mozilla/standards-positions/issues/491#issuecomment-785303882
> <https://github.com/mozilla/standards-positions/issues/491#issuecomment-785303882>)
> Webkit: Positive, "We have not yet implemented the proposal but we
> generally have a positive opinion of it."
> (https://lists.webkit.org/pipermail/webkit-dev/2021-February/031705.html
> <https://lists.webkit.org/pipermail/webkit-dev/2021-February/031705.html>)
>
> On Wednesday, February 24, 2021 at 8:59:52 AM UTC-8 Zhi An Ng wrote:
>
> On Wednesday, February 24, 2021 at 7:41:49 AM UTC-8
> yo...@yoav.ws wrote:
>
> On Wed, Feb 24, 2021 at 4:20 PM Ben Kelly
> <wande...@chromium.org> wrote:
>
> On Wed, Feb 24, 2021 at 9:28 AM Yoav Weiss
> <yo...@yoav.ws> wrote:
>
> On Mon, Feb 22, 2021 at 8:00 PM Zhi An Ng
> <zh...@chromium.org> wrote:
>
> Interoperability and Compatibility
>
> Gecko: In development
> (https://bugzilla.mozilla.org/show_bug.cgi?id=1625130
> <https://bugzilla.mozilla.org/show_bug.cgi?id=1625130>)
>
> That doesn't seem accurate, as I don't see a ton of
> actual activity on that bug.
> have you tried reaching out
> <https://bit.ly/blink-signals>? Alternatively, were
> they involved in the WG decision
> <https://github.com/WebAssembly/meetings/blob/master/main/2021/CG-02-16.md#fixed-width-simd-zhi-an-ng---slides10-min> to
> move this to phase 4?
>
>
> Yes, they were involved in the WG decision, in particular Lars
> Hansen was present at that CG meeting
> (https://github.com/WebAssembly/meetings/blob/master/main/2021/CG-02-16.md#introduction-of-attendees
> <https://github.com/WebAssembly/meetings/blob/master/main/2021/CG-02-16.md#introduction-of-attendees>)
> and has been actively involved in the entire process (he is
> present in almost all of the meetings, it's a bit manual but you
> can click into each issue in
> https://github.com/WebAssembly/simd/issues?q=is%3Aissue+is%3Aclosed+agenda
> <https://github.com/WebAssembly/simd/issues?q=is%3Aissue+is%3Aclosed+agenda>
> and look at the notes to see the list of participants, or this
> shows his activity on the issue tracker
> https://github.com/WebAssembly/simd/issues?q=is%3Aissue+commenter%3Alars-t-hansen
> <https://github.com/WebAssembly/simd/issues?q=is%3Aissue+commenter%3Alars-t-hansen>).
>
> I've filed an issue on mozilla/standards-positions
> https://github.com/mozilla/standards-positions/issues/491
> <https://github.com/mozilla/standards-positions/issues/491>.
>
>
> FWIW you have to look at the dependencies on that bug.
> Last I checked WASM SIMD was working in firefox
> nightly.  (But, yes, asking on
> mozilla/standards-positions would be good.)
>
>
> Good point, I missed that...
>
> --
> You received this message because you are subscribed to the Google
> Groups "blink-dev" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to blink-dev+...@chromium.org
> <mailto:blink-dev+...@chromium.org>.
> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/21a8098f-0665-46cc-8be8-3d71343e4f2cn%40chromium.org?utm_medium=email&utm_source=footer>.
>
> --
> You received this message because you are subscribed to the Google
> Groups "blink-dev" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to blink-dev+...@chromium.org
> <mailto:blink-dev+...@chromium.org>.
> To view this discussion on the web visit
> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACj%3DBEixipxR5vBgy-Xk_xX_r19d79pzr7rJEOWjbBRYrgShgA%40mail.gmail.com
> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACj%3DBEixipxR5vBgy-Xk_xX_r19d79pzr7rJEOWjbBRYrgShgA%40mail.gmail.com?utm_medium=email&utm_source=footer>.

Mike West

unread,
Feb 25, 2021, 2:14:39 PM2/25/21
to blink-dev, Manuel Rego, blink-dev, Zhi An Ng, Ben Kelly, yo...@yoav.ws, Zhi An Ng
LGTM3.

Zhi An Ng

unread,
Feb 25, 2021, 4:41:17 PM2/25/21
to blink-dev, Mike West, Manuel Rego, blink-dev, Zhi An Ng, Ben Kelly, yo...@yoav.ws, Zhi An Ng
Thank you API Owners for the reviews and LGTMs.
Reply all
Reply to author
Forward
0 new messages