Path: ...!2.eu.feeder.erje.net!feeder.erje.net!news.swapon.de!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: John Savard <quadibloc@servername.invalid>
Newsgroups: comp.arch
Subject: Re: Stealing a Great Idea from the 6600
Date: Thu, 20 Jun 2024 11:01:38 -0600
Organization: A noiseless patient Spider
Lines: 34
Message-ID: <9sn87j5e40ike0c7oqd7k05in75faldcuj@4ax.com>
References: <152f8504112a37d8434c663e99cb36c5@www.novabbs.org> <v04tpb$pqus$1@dont-email.me> <v4f5de$2bfca$1@dont-email.me> <jwvzfrobxll.fsf-monnier+comp.arch@gnu.org> <v4f97o$2bu2l$1@dont-email.me> <613b9cb1a19b6439266f520e94e2046b@www.novabbs.org> <v4hsjk$2vk6n$1@dont-email.me> <6b5691e5e41d28d6cb48ff6257555cd4@www.novabbs.org> <v4tfu3$1ostn$1@dont-email.me> <96280554541a8a9b1a29a5cbd5b7c07b@www.novabbs.org> <_vLcO.48542$eX68.22354@fx18.iad> <ObXcO.17998$xTZ2.3297@fx46.iad>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Injection-Date: Thu, 20 Jun 2024 19:01:40 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="c0884c951618dc0b5d1da09bdf5ffb68";
	logging-data="2810442"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1+ZAp06s2SHR8DUSfdbkp8SDLhTknfSYYk="
Cancel-Lock: sha1:hJCkdT3CrED0vhJVj71pq+hScbw=
X-Newsreader: Forte Free Agent 3.3/32.846
Bytes: 3137

On Thu, 20 Jun 2024 10:40:32 -0400, EricP
<ThatWouldBeTelling@thevillage.com> wrote:

>BTW in Jan 2023 Chips-and-cheese did a two part retrospective deep dive
>on Bulldozer microarchitecture compared to Sandy Bridge and others.
>
>It doesn't single out a culprit. It mentions the Bulldozer integer
>scheduler as choosing between either a single oldest entry or based on
>physical location, which sounds like a plain linear priority picker.
>That might cause problems where it can cause saw-tooth performance as a
>stalled dependency chain builds up waiting on one item, then releases.
>That is why I used a circular priority picker, aka a round robin arbiter,
>which makes the last selected slot the lowest circular priority
>and thereby ensures that each scheduler slot is serviced evenly.
>
>But it sounds like there were many interacting issues.
>
>https://chipsandcheese.com/2023/01/22/bulldozer-amds-crash-modernization-frontend-and-execution-engine/
>
>https://chipsandcheese.com/2023/01/24/bulldozer-amds-crash-modernization-caching-and-conclusion/

That's interesting. And here I thought the problem with Bulldozer was
well-known: it tried for efficiency by making the same mistake as the
Pentium 4, having fewer gate delays per cycle so it could have a
faster clock rate.

As those who post here realize, this isn't just for bragging rights:
it means a larger number of instructions can be executing in parallel
at different stages of operation on the pipeline. And the problem with
that, at least for the Pentium 4, was that thermal issues limited
further increases to the clock speed with so many transistors active
at once from the extra instructions.

John Savard