WebML CG Teleconference – 3 October 2019 - meeting minutes from Kostiainen, Anssi on 2019-10-03 (public-webmachinelearning@w3.org from October 2019)

From: Kostiainen, Anssi <anssi.kostiainen@intel.com>
Date: Thu, 3 Oct 2019 15:19:19 +0000
To: "public-webmachinelearning@w3.org" <public-webmachinelearning@w3.org>
Message-ID: <BE013CA0-4D9D-436F-8AFF-0027EF38FED8@intel.com>

Hi All,

Draft minutes from today's call at:

https://www.w3.org/2019/10/03-webmachinelearning-minutes.html

We agreed to come up with proof-of-concept requirements (incl. proposed target hardware and platform) for evaluating performance of buffer sharing between dedicated ML hardware and GPU in order to address concern raised in a data-driven manner. Ningxin leads, Paul contributes (thanks!). Proposal to be posted on GH for the group to review.

Thanks,

-Anssi (WebML CG chair)

Received on Thursday, 3 October 2019 15:19:49 UTC