From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: tuhs-bounces@minnie.tuhs.org X-Spam-Checker-Version: SpamAssassin 3.4.1 (2015-04-28) on inbox.vuxu.org X-Spam-Level: X-Spam-Status: No, score=-0.8 required=5.0 tests=DKIM_SIGNED,DKIM_VALID, DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,HTML_MESSAGE,MAILING_LIST_MULTI, RCVD_IN_DNSWL_NONE autolearn=ham autolearn_force=no version=3.4.1 Received: from minnie.tuhs.org (minnie.tuhs.org [45.79.103.53]) by inbox.vuxu.org (OpenSMTPD) with ESMTP id 2f055cd9 for ; Mon, 18 Jun 2018 18:52:05 +0000 (UTC) Received: by minnie.tuhs.org (Postfix, from userid 112) id 01C25A19D6; Tue, 19 Jun 2018 04:52:04 +1000 (AEST) Received: from minnie.tuhs.org (localhost [127.0.0.1]) by minnie.tuhs.org (Postfix) with ESMTP id C7C01A19D3; Tue, 19 Jun 2018 04:51:38 +1000 (AEST) Authentication-Results: minnie.tuhs.org; dkim=pass (1024-bit key; unprotected) header.d=ccc.com header.i=@ccc.com header.b=JjsHq5d+; dkim-atps=neutral Received: by minnie.tuhs.org (Postfix, from userid 112) id 2BB4AA19D3; Tue, 19 Jun 2018 04:51:35 +1000 (AEST) Received: from mail-io0-f173.google.com (mail-io0-f173.google.com [209.85.223.173]) by minnie.tuhs.org (Postfix) with ESMTPS id 46D5D9EC27 for ; Tue, 19 Jun 2018 04:51:34 +1000 (AEST) Received: by mail-io0-f173.google.com with SMTP id g7-v6so17752535ioh.11 for ; Mon, 18 Jun 2018 11:51:34 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ccc.com; s=google; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc; bh=8fjPGU2aFccXBEifT2SBvgmsw9rphZnOhFnuQEV5HgM=; b=JjsHq5d+aj1ACCqsguRo4dhAjYuIRemoMgBy8FyiBc4X0B5D2Xh5lybZ2YFkjNVkxg TdLMEb8Z7qYCnp32QAs96ngSCn945PT+8H36EsIAJdlOT3TS+/tRFZBVaC03l1sWXWkK qdbaTyE7gyztOTV595RVrzvO17/uJ39Gvhp48= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc; bh=8fjPGU2aFccXBEifT2SBvgmsw9rphZnOhFnuQEV5HgM=; b=bJEW5oglxC7fzdkENPfpVqPYkGmbycfyt5kB7OfJFmAkVpV4Y+ctnYkXecXdJTtYZE x3kk5jxtCitcAjH9EQO04U5e3dl2v/UEOu7+91myVFhCj4sIOIPjYwD+RtThtrjilXvP nrNmYONH2iHVLGXb9GOYy40wy1YKephuHdycB4Aln33GCeMSs2Aauxkxp0mna+mjitMk 9igb9O/sqsLEJpr4lXKg3Y+T31XyF1ZyWf0EorH6JmTLDAqY19WMC05+Rmvi5rsT95S4 yzrTg9s9ME4HOv8cwhpjRhHRrkF+8LBOYeCJ3vcPC4Yuy204NIRztmiJsbKoZL9oY4AV 8r3Q== X-Gm-Message-State: APt69E3/pdf5DA7mmu1xcZrLMgA6o8R/GazNqaXKz+tob4GRQyQBUqKS 7izLTTLsch71Cpld2lDx1lUSbUhF76g/iG7u54pSgw== X-Google-Smtp-Source: ADUXVKJ/x88bWumR3DXFl3tDn85AxbgWm2PaooQqaN8A7LkWxvxCAYbadck6xRE4G16VQxzDMMe73zIsgAhZOxxY3ZU= X-Received: by 2002:a6b:b8d5:: with SMTP id i204-v6mr10811507iof.284.1529347893223; Mon, 18 Jun 2018 11:51:33 -0700 (PDT) MIME-Version: 1.0 Received: by 2002:a4f:ca8a:0:0:0:0:0 with HTTP; Mon, 18 Jun 2018 11:51:02 -0700 (PDT) In-Reply-To: <20180618175638.383D418C086@mercury.lcs.mit.edu> References: <20180618175638.383D418C086@mercury.lcs.mit.edu> From: Clem Cole Date: Mon, 18 Jun 2018 14:51:02 -0400 Message-ID: To: Noel Chiappa Content-Type: multipart/alternative; boundary="000000000000ea1908056eef0d06" Subject: Re: [TUHS] core X-BeenThere: tuhs@minnie.tuhs.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: The Unix Heritage Society mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: The Eunuchs Hysterical Society Errors-To: tuhs-bounces@minnie.tuhs.org Sender: "TUHS" --000000000000ea1908056eef0d06 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Mon, Jun 18, 2018 at 1:56 PM, Noel Chiappa wrote: > > > Just like I retold the Amdahl/Brooks story of the 8-bit byte and > Amdahl > > thinking Brooks was nuts > > Don't think I've heard that one? =E2=80=8BApologies for the repeat, if I have set this to TUHS before (I kno= w I have mentioned it in other places).=E2=80=8B Somebody else on this list mention= ed David Brailsford YouTube Video. Which has more details, biut he had some of not quite right. I had watched it an we were communicating. The actual story behind the byte is a bit deeper than he describes in the lecture, which he thanked me - as I have since introduced him to my old friend & colleague Russ Robelen who was was the chief designer of the Model 50 and later lead the ASC system with John Coche working for him. Brailsford is right about results of byte addressing and I did think his lecture is excellent and you can learn a great deal. That said, Russ tells the details of story like this: Gene Amdahl wanted the byte to be 6 bits and felt that 8 bits was 'wasteful' of his hardware. Amdahl also did not see why more than 24 bits for a word was really needed and most computations used words of course. 4, 6-bit bytes in a word seemed satisfactory to Gene. Fred Brooks kept kicking Amdahl out of his office and told him flatly - that until he came back with things that were power's of 2, don't bother - we couldn't program it. The 32-bit word was a compromise, but note the original address was only 24-bits (3, 8-bit bytes), although Brooks made sure all address functions stored all 32-bits - which as Gordon Bell pointed out later was the #1 thing that saved System 360 and made it last. =E2=80=8BBTW: =E2=80=8B Russ and =E2=80=8B =E2=80=8B Ed Sussenguth invented speculative execution for the ACS system =E2=80=8B a couple of years later=E2=80=8B . I was =E2=80=8B =E2=80=8B harassing him =E2=80=8B last January,=E2=80=8B because for 40 years we have been using his cool idea and it =E2=80=8B =E2=80=8B came back to bite us at Intel. Here is the message cut/pasted from his =E2=80=8B =E2=80=8B email for context: "I see you are still leading a team at Intel =E2=80=8B =E2=80=8B developing super computers and associated technologies. It certainly =E2=80=8B =E2=80=8B is exciting times in high speed computing. =E2=80=8B =E2=80=8B It brings back memories of my last work at IBM 50 years ago on the ACS =E2=80=8B =E2=80=8B project. You know you are old when you publish in the IEEE Annals of =E2=80=8B =E2=80=8B the History of Computing. One of the co-authors, Ed Sussenguth, passed =E2=80=8B =E2=80=8B away before our paper was published =E2=80=8B =E2=80=8B in =E2=80=8B =E2=80=8B 2016. =E2=80=8B =E2=80=8B https://www.computer.og/csdl/mags/an/2016/01/man2016010060.html =E2=80=8B=E2=80=8B Some =E2=80=8B =E2=80=8B of the work we did way back then has made the news in an unusual way with the recent revelations on Spectre and Meltdown. I read the =E2=80=8B =E2=80=8B =E2=80=98Spectre Attacks: Exploiting Speculative Execution=E2=80=99 paper y= esterday =E2=80=8B =E2=80=8B trying to understand how speculative execution was being exploited. =E2=80=8B =E2=80=8B At =E2=80=8B =E2=80=8B ACS we were the first group at IBM to come up with the notion of the =E2=80=8B =E2=80=8B Branch Table and other techniques for speeding up execution. I wish you were closer. I=E2=80=99d do love to hear your views on the state= of =E2=80=8B =E2=80=8B computing today. I have a framed micrograph of the IBM Power 8 chip on the wall in my office. In many ways the Power Series is an outgrowth =E2=80=8B =E2=80=8B of ACS. =E2=80=8B =E2=80=8B I still try to keep up with what is happening in my old field. The =E2=80=8B =E2=80=8B recent advances by Google in Deep Learning are breathtaking to me. =E2=80=8B =E2=80=8B Advances like AlphaGo Zero I never expected to see in my lifetime. =E2=80=8B"=E2=80=8B > > But you can lose with that strategy too. > > Multics had a lot of sub-systems re-written from the ground up over time, > and > the new ones were always better (faster, more efficient) - a common even > when > you have the experience/knowledge of the first pass. > > Unfortunately, by that time it had the reputation as 'horribly slow and > inefficient', and in a lot of ways, never kicked that: > > http://www.multicians.org/myths.html > > Sigh, sometimes you can't win! =E2=80=8BYep - although I think that may have been a case of economics. = Multics for all its great ideas, was just a huge system, when Moores law started to make smaller systems possible. So I think your comment about thinking about what you need now and what you will need in the future was part of the issue. =E2=80=8B I look at Multics vs Unix the same way I look at TNC vs Beowulf clusters. At the time, we did TNC, we worked really hard to nail the transparency thing and we did. It was (is) awesome. But it cost. Tru64 (VMS and other SSI) style clusters are not around today. The hack that is Beowulf is what lived on. The key is that it was good enough and for most people, that extra work we did to get rid of those seams just was not worth it. And in the because Beowulf was economically successful, things were implemented for it, that were never even considered for Tru64 and the SSI style systems. To me, Multics and Unix have the same history. =E2=80=8B Multics was (is) cool; but Unix is similar; but different and t= oo the lead. The key is that it was not a straight path. Once Unix took over, history went in a different direction. Clem --000000000000ea1908056eef0d06 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable


On Mon, Jun 18, 2018 at 1:56 PM, Noel = Chiappa <jnc@mercury.lcs.mit.edu> wrote:
=C2=A0=C2=A0
=C2=A0 =C2=A0 > Just like I retold the Amdahl/Brooks story of the 8-bit = byte and Amdahl
=C2=A0 =C2=A0 > thinking Brooks was nuts

Don't think I've heard that one?
= =E2=80=8BApologies for the repeat, if I have set th= is to TUHS before (I know I have mentioned it in other places).=E2=80=8B=C2= =A0 Somebody else on this list mentioned David Brailsford YouTube Video.=C2= =A0 Which has more details, biut he had some of not quite right.=C2=A0 I ha= d watched it an we were communicating.=C2=A0 =C2=A0The actua= l story behind the byte is a bit deeper than he describes in the lecture, w= hich he thanked me - as I have since introduced him to=C2=A0 my=C2=A0=C2=A0= old friend & colleague Russ Robelen who was was t= he chief designer of the Model 50=C2=A0and later lead= the ASC system with John Coche working for him.=C2=A0 Brailsford is right = about results of byte addressing and I did think his lecture is excellent a= nd you can learn a great deal.=C2=A0 =C2=A0That said, Russ tells=C2=A0the details of story like this:

Gene Amdahl wanted the byte to be 6 bits and felt=C2=A0that 8 bits was 'wasteful' of his hardware.=C2= =A0 =C2=A0Amdahl also did not see why more=C2=A0than = 24 bits for a word was really needed and most computations used words of co= urse.=C2=A0 4, 6-bit bytes in a word seemed satisfactory to Gene.=C2=A0 =C2= =A0Fred Brooks kept kicking=C2=A0Amdahl out of his of= fice and told him flatly - that until he came back with=C2=A0things that were power's of 2, don't bother - we couldn'= ;t program it.=C2=A0 =C2=A0The=C2=A032-bit word was a= compromise, but note the original address was only 24-bits (3, 8-bit bytes= ),=C2=A0although Brooks made sure all address functio= ns stored all 32-bits - which=C2=A0as Gordon Bell poi= nted out later was the #1 thing that saved System 360 and=C2=A0made it last.
<= div class=3D"gmail_extra">

=E2=80=8BBTW: = =E2=80=8B
Russ and
=E2=80=8B =E2=80=8B
Ed Sussenguth in= vented speculative execution for the ACS system
=E2=80=8B = a couple of years later=E2=80=8B
.=C2=A0 =C2=A0I was
= =E2=80=8B =E2=80=8B
harassing him
=E2=80=8B last January,=E2= =80=8B
because for 40 years we have been using his cool idea and it=E2=80=8B =E2=80=8B
came back to bite us at Intel.=C2=A0 Here= is the message cut/pasted from his
=E2=80=8B =E2=80=8Bemai= l=C2=A0for context:

"I see you are still leading a team at Intel
=E2=80= =8B =E2=80=8B
developing super computers and associated technologies. It cert= ainly
=E2=80=8B =E2=80=8B
is exciting times in high speed co= mputing.
=E2=80=8B =E2=80=8B
It brings back memories of my l= ast work at IBM 50 years ago on the ACS
=E2=80=8B =E2=80= =8B
=E2=80=8B =E2=80=8B
the History of Computing. One of the co-auth= ors, Ed Sussenguth, passed
=E2=80=8B =E2=80=8B
away before o= ur paper was published
=E2=80=8B =E2=80=8B
in
=E2=80=8B =E2=80=8B
2016.
=E2=80=8B =E2=80= =8B
=C2=A0https://www.computer.og/csdl/mags/an/2016/01/= man2016010060.html
=E2=80=8B=E2=80=8B
=C2=A0Some
=E2=80=8B =E2=80=8B
of the work we did way back then has made the new= s in an unusual way
with the recent revelations on= Spectre and Meltdown. I read the
=E2=80=8B =E2=80=8B=E2=80= =98Spectre Attacks: Exploiting Speculative Execution=E2=80=99 paper yesterd= ay
=E2=80=8B =E2=80=8B
trying to understand how speculative = execution was being exploited.
=E2=80=8B =E2=80=8B
A= t
=E2=80=8B =E2=80=8B
ACS we were the first group at IBM to = come up with the notion of the
=E2=80=8B =E2=80=8B
Branch Ta= ble and other techniques for speeding up execution.=C2=A0

<= /div>
I= wish you were closer. I=E2=80=99d do love to hear your views on the state = of
=E2=80=8B =E2=80=8B
computing today. I have a framed micr= ograph of the IBM Power 8 chip on
the wall in my o= ffice. In many ways the Power Series is an outgrowth
=E2= =80=8B =E2=80=8B
of ACS.
=E2=80=8B =E2=80=8B
I still try to ke= ep up with what is happening in my old field. The
=E2=80= =8B =E2=80=8B
recent advances by Google in Deep Learning are breathtaking to = me.
=E2=80=8B=C2=A0 =E2=80=8B
Advances like AlphaGo Zero I n= ever expected to see in my lifetime.
=E2=80=8B"=E2=80= =8B
=C2=A0


But you can lose with that strategy too.

Multics had a lot of sub-systems re-written from the ground up over time, a= nd
the new ones were always better (faster, more efficient) - a common even wh= en
you have the experience/knowledge of the first pass.

Unfortunately, by that time it had the reputation as 'horribly slow and=
inefficient', and in a lot of ways, never kicked that:

=C2=A0 http://www.multicians.org/myths.html

Sigh, sometimes you can't win!
=E2=80=8BYep - although I think that may have bee= n a case of economics.=C2=A0 =C2=A0 Multics for all its great ideas, was ju= st a huge system, when Moores law started to make smaller systems possible.= =C2=A0 So I think your comment about thinking about what you need now and w= hat you will need in the future was part of the issue.=C2=A0 =C2=A0=E2=80= =8B
=C2=A0
=
I look at Multics vs Unix the same way I loo= k at TNC vs Beowulf clusters.=C2=A0 =C2=A0At the time, we did TNC, we worke= d really hard to nail the transparency thing and we did.=C2=A0 =C2=A0It was= (is) awesome.=C2=A0 But it cost.=C2=A0 =C2=A0Tru64 (VMS and other SSI) sty= le clusters are not around today.=C2=A0 The hack that is Beowulf is what li= ved on.=C2=A0 =C2=A0The key is that it was good enough and for most people,= that extra work we did to get rid of those seams just was not worth it.=C2= =A0 And in the because Beowulf was economically successful, things were imp= lemented for it, that were never even considered for Tru64 and the SSI styl= e systems.=C2=A0 =C2=A0 =C2=A0To me, Multics and Unix have the same history= .
=E2=80=8B=C2=A0 =C2=A0Multics was (is) cool; but U= nix is similar; but different and too the lead.=C2=A0 =C2=A0The key is that= it was not a straight path.=C2=A0 Once Unix took over, history went in a d= ifferent direction.

Clem
--000000000000ea1908056eef0d06--