Subject: Re: Fwd: Jim Larus speaking on Friday
From: Craig Chambers (chambers@cs.washington.edu)
Date: Wed Apr 04 2001 - 11:14:46 PDT
I think his use of the word "staged" isn't that much like ours, and refers more
to a job scheduling algorithm (based on some sketchy stuff I heard before about
this project). But it could be worth attending. So how about we delay the
group meeting till 4:30? Or do people want to do TGIF or go home early or
something? I do have some small things I'd like to talk about at the group
meeting (some research ideas inspired by Manuvir's class), so I'm not sure I
want to cancel. Nor can I do the meeting earlier, given my schedule.
-- Craig
Keunwoo Lee wrote:
>
> Should we postpone or reschedule group meeting this Friday? It seems like
> this would be interesting.
>
> ~k
>
> ,--------------- Forwarded message (begin)
>
> Subject: Jim Larus speaking on Friday
> From: Susan Eggers <eggers@cs.washington.edu>
> Date: Tue, 3 Apr 2001 08:57:05 -0700
>
> This Friday Jim Larus from Microsoft Research (and before that
> a professor at the other UW) will be speaking at the systems
> seminar (590S) on a staged model for servers. For this date
> only 590S will be held in **322** and at the usual time, 3:30.
> His abstract is below.
>
>
>
> Enhanced Server Performance with StagedServer
> James Larus
> larus@microsoft.com
> Microsoft Research
>
> Servers--for the web, databases, mail, files, or many other resources--are
> commonly structured as a collection of parallel tasks, each of which runs
> code to process a request. Threads, processes, or event handlers underlie
> these systems' software architecture because these mechanisms offer the
> control independence and dynamic scheduling that hides high latency
> operations such as I/O and communication. Unfortunately, many servers run
> poorly on modern processors. In part, this poor performance is attributable
> to these programs' software architecture, which frequently switches between
> unrelated pieces of code, thereby destroying the program locality that is a
> prerequisite for effective caches, TLB, and branch predictors.
>
> This work propose a software solution to this problem. Cohort Scheduling is
> a technique that increases code and data locality by consecutively executing
> logically related operations across different server requests. To support
> this technique, we developed a programming model, called Staged Computation,
> that offers an abstraction to group related operations and mechanisms to
> implement this scheduling policy. These techniques have been implemented in
> a library called StagedServer. Measurements on realistic systems show that
> Cohort Scheduling can improve both processor performance and server
> throughput. (Joint work with Michael Parkes.)
>
>
> `--------------- Forwarded message (end)
>
> _______________________________________________
> Cecil mailing list
> Cecil@cs.washington.edu
> http://majordomo.cs.washington.edu/mailman/listinfo/cecil
_______________________________________________
Cecil mailing list
Cecil@cs.washington.edu
http://majordomo.cs.washington.edu/mailman/listinfo/cecil
This archive was generated by hypermail 2b25 : Wed Apr 04 2001 - 11:15:04 PDT