Account Options

  1. Sign in
The old Google Groups will be going away soon, but your browser is incompatible with the new version.
Google Groups Home
« Groups Home
comp.arch
+ new post
About this group
Subscribe to this group
This is a Usenet group - learn more
Message from discussion Beyond multicore
The group you are posting to is a Usenet group. Messages posted to this group will make your email address visible to anyone on the Internet.
Your reply message has not been sent.
Your post was successful
 
From:
To:
Cc:
Followup To:
Add Cc | Add Followup-to | Edit Subject
Subject:
Validation:
For verification purposes please type the characters you see in the picture below or the numbers you hear by clicking the accessibility icon. Listen and type the numbers you hear
 
Robert Myers  
View profile  
 More options Mar 23 2007, 11:17 am
Newsgroups: comp.arch
From: "Robert Myers" <rbmyers...@gmail.com>
Date: 23 Mar 2007 08:17:35 -0700
Local: Fri, Mar 23 2007 11:17 am
Subject: Re: Beyond multicore
Print | View thread | Show original | Report this message | Find messages by this author
On Mar 22, 5:12 pm, Del Cecchi <cecchinos...@us.ibm.com> wrote:

> Actually you pretty much described Blue Gene, at least at a conceptual
> level.

> The problem with this as a desktop/side model is there would have to be
> customers and software.

> If you have a plan for those, the sandhill road boys would take care of
> you.

I thought the 80-core model displayed by Intel already resembled Blue
Gene: each core had it's own router chip.  Not nearly as much memory
per processor, of course, and I'd be curious to know about the
bisection bandwidth of the chip (one track mind).

What are we going to be using any of this junk for, except animation
and problems in computational physics with a certain kind of
structure?

As to the illusion that we're running out of things for processors to
do, I have a new name for the user interaction model.  I call it
"point and wait."  Since memory requirements are going up well faster
than single thread processor capability, I assume that what I'm
waiting on is a von Neumann bottleneck (updating the display) that is
created by software that just keeps getting ever glitzier and ever
more resource intensive.

Robert.


 
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
©2013 Google