POWER6/7, compilers and more

By: Jesper frimann (jesperfrimann.delete@this.gmail.com), October 3, 2007 12:18 am
Room: Moderated Discussions
David Kanter (dkanter@realworldtech.com) on 10/2/07 wrote:
---------------------------
>I just ran across a whole bunch of really interesting presentations from folks at IBM on various technologies:
>

Real nice find !

>This one has a roadmap which shows the POWER7 and some vague details, as well as
>a discussion of the POWER6. At long fucking last, they admitted that the POWER6
>was an InO core - it only took them about 2 years after they first started talking
>about it. Hurray for the triumph of technical disclosure over PR and marketing....oh wait.
>
>http://www.spscicomp.org/ScicomP13/Presentations/IBM/HW+SWRoadmap.pdf

It's called stealth marketing :)= a concept developed by
IBM marketing, which by the looks of it consists of former
Homeland Security People.

Well there are some nice info in that presentation as I see it:

POWER6 based p595 will likely be a 4TB RAM machine.(slide 7)
which isn't a surprise really, but will still only be a
64 core machine.

No POWER6 based QCM modules right now (slide 10), which I
think is a big big big mistake.

A POWER6 blade to replace the JS2[01],(slide 12) about time IMHO
that they got a real POWER blade.

p505,p510,p520,p550,p560 are history and replaced by a 4
socket p550, and the js2X POWER6 based blade, also a huge
gamble IMHO. I can see the logic in it because blade servers
are eating up the rack server marked from below. But still....

p575 will go from 16 cores to 32 cores in 2U that'll prob be
something like 900 in specintrate and 1k in fp.

Otherwise it's slide 14 that have the key points.

Partition mobility looks like something that could be
really useful in the datacenter together with partition
hibernation.

Virtual Memory sounds a bid like trouble.

And the Aix 6.1 Workload partitions sounds like a version of
Solaris containers, but with the added feature of you being
able to move your workloads between physical servers.

>
>Here's another presentation that talks about IBM's compiler technology, and one
>rather interesting slide they have (31), shows the performance improvement due
>to compilers. It's roughly 10% annually as measured by SPECfp.
>
>http://www.spscicomp.org/ScicomP13/Presentations/IBM/CompilerRoadmapArchambault.pdf
>
>If you look in the presentations directory, there are several other presentations as well.
>
>DK

Jup..
The Cell Broadband Engine™ Update by Michael Hennecke
have some really nice slides on roadmaps and the future
regarding the SDK and the actual hardware.


// Jesper
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
POWER6/7, compilers and moreDavid Kanter2007/10/02 09:36 PM
  POWER6/7, compilers and moreGabriele Svelto2007/10/02 11:32 PM
    POWER6/7, compilers and moreAnil Maliyekkel2007/10/03 08:23 AM
      POWER6/7, compilers and moreanon2007/10/03 08:28 AM
        POWER6/7, compilers and moreAnil Maliyekkel2007/10/03 02:22 PM
      Can you email me the PPT? (NT)David Kanter2007/10/03 08:56 AM
        Me too if you don't mind Anil (NT)Euronymous2007/10/03 12:33 PM
      POWER6/7, compilers and moreLinus Torvalds2007/10/03 09:34 AM
        POWER6/7, compilers and moreAnil Maliyekkel2007/10/03 02:12 PM
          POWER6/7, compilers and moreAnil Maliyekkel2007/10/03 04:06 PM
  POWER6/7, compilers and moreJesper frimann2007/10/03 12:18 AM
  POWER6/7, compilers and moreMr. Camel2007/10/03 02:33 AM
    POWER6/7, compilers and moreasdf2007/10/03 02:08 PM
  POWER6/7, compilers and moreanon2007/10/03 08:09 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell green?