Edge Rewrite
// HTMLRewriter · presentation

This page was redesigned at the edge.

Cloudflare fetched the original article and streamed it through HTMLRewriter to apply an entirely new visual system without rebuilding the source page.

// request.cf · coarse context

A page that knows where it met you.

Only coarse request metadata is shown. This demo does not display or persist visitor IP addresses.

Country
US
Cloudflare location
CMH
Connection
HTTP/2
Language
Not provided

Ray ID: a21f66282fc49318

Jump to content

Talk:OneAPI (compute acceleration)

Page contents not supported in other languages.
Add topic
From Wikipedia, the free encyclopedia

The {advert} tag on this article is inappropriate and should not have been added.

[edit]

The article (as of now) states what the initiative is, who started it, and what its goals are. Those are statements of fact. It doesn't say it's "good" or "better than" something else. Longitude2 (talk) 14:39, 8 February 2020 (UTC)Reply

I am a user of Intel OneAPI, and until I came across this page I did not know of any "general use" of it, beyond the Intel tools. While I (very) like the tools, I have severe reservations about this page. It claims to be a general, industry-wide standard but I cannot find enough cites to support this.
IMHO, for this page to continue to exist it needs to demonstrate that it is a true multi-vendor software. And, ones which are not really multi-vendor (e.g. the Intel-Heidelberg) need to be de-emphasized. Ldm1954 (talk) 19:29, 10 April 2023 (UTC)Reply
An important addendum. The term OneAPI is apparently an Intel trademark, see https://www.fluidnumerics.com/legal/trademarks. The advert marking is highly appropriate. Ldm1954 (talk) 02:33, 12 April 2023 (UTC)Reply
I did a copyedit pass and removed the advert and COI tags, since this article meets the factual/objective standard now. However, the notability is still in question: it's not clear to me how much impact this standard has in the industry, and if it's effectively just an internal Intel tool that's been made open-source it probably doesn't deserve a page. The article does list implementations released by non-Intel actors, so I left the page up. Tangledyarn (talk) 18:31, 24 January 2024 (UTC)Reply

Non-notable label - citations added

[edit]

This is an open specification not a single vendor solution. Additional references have been added and cite multiple sources and companies. This is fairly new, additional sources and industry engagement will become available as the project evolves.  Preceding unsigned comment added by GoBuck76 (talkcontribs) 17:46, 10 February 2020 (UTC)Reply

So, what exactly is oneAPI?

[edit]

240F:60:1125:1:3151:7C6:CC82:DD2B (talk) 13:13, 10 April 2020 (UTC)Reply

Advice request to remove maintenance header

[edit]

Hello, I tried to keep my edits as factual and free from bias as possible. Can you advise me on how to improve it so that the maintenance header can be removed? Thanks  Preceding unsigned comment added by Kgmoffat (talkcontribs) 16:50, 8 October 2021 (UTC) Kgmoffat (talk) 15:17, 11 October 2021 (UTC)Reply

Misleading sentence

[edit]

This sentence "University of Heidelberg has developed a SYCL/DPC++ implementation for both AMD and Nvidia GPUs" implies that SYCL is somehow a subset of DPC++. The reference is also not appropriate. It should point to the developer page (https://github.com/OpenSYCL/OpenSYCL)  Preceding unsigned comment added by 2001:708:10:4008:B:0:0:C (talk) 08:43, 19 May 2023 (UTC)Reply

computer.org reference removed

[edit]

@0xDeadbeef: Can you help me to understand this edit? I think that putting <ref></ref> around the [...] would have made it into an acceptable link. Thank you in advance! Dimawik (talk) 21:10, 10 January 2024 (UTC)Reply

Does it need to be a reference? It doesn't seem needed to support any claim. 0xDeadbeef→∞ (talk to me) 22:47, 10 January 2024 (UTC)Reply