[Duncan's Home] Duncan's Jotter
faq -  feedback -  home 
Members
Logon   -   Sign Up
smeed >> duncan.smeed.org | Duncan's Jotter
duncan.smeed.org | Duncan's Jotter

Day Link Icon 5/3/2005

Ultra-thin client to close digital divide

(by Duncan, @ 1:54 PM)

Ultra-thin client to close digital divide | The Register

"The Nivo (network in, video out) box is a sub-£100, ultra-ultra-thin client that can be networked along with several others to a central PC/server. It is cheaper, more accessible and more environmentally friendly than a PC, Ndiyo argues, needing much less in the way of raw materials to build, and consuming far less power. It is also based on non-proprietary standards and open source software."...

In addition to The Register article Ndiyo! - The Ndiyo Project has more details.

Comments: 0 | Reply | Categories: CAD



Day Link Icon 3/8/2005

Cache Profiling and the SPEC Benchmarks: A Case Study

(by Duncan, @ 11:40 PM)

"As VLSI technology improvements continue to widen the gap between processor and main memory cycle times, cache performance becomes increasingly important to overall system performance. Cache memories help alleviate the cycle time disparity, but only for programs that exhibit sufficient spatial and temporal locality. Programs with unruly access patterns spend much of their time transferring data to and from the cache. To fully exploit the performance potential of fast processors, programmers must explicitly consider cache behavior, restructuring their codes to increase locality. As these fast processors proliferate, techniques for improving cache performance must move beyond the supercomputer and multiprocessor communities and into the mainstream of computing.

In this paper, we examine some of the techniques that programmers can use to improve cache performance. We show how to use CProf, a cache profiler, to identify cache performance bottlenecks and gain insight into their origin. This insight helps programmers understand which of the well-known program transformations are likely to improve cache performance. Using CProf and a "cookbook" of simple transformations, we show how to tune the cache performance of six of the SPEC92 benchmarks. By restructuring the source code, we greatly improve cache behavior and achieve execution time speedups ranging from 1.02 to 3.46."

Comments: 0 | Reply | Categories: CAD

Optimization of Computer Programs in C

(by Duncan, @ 11:37 PM)

Optimization of Computer Programs in C
"This document describes techniques for optimizing (improving the speed of) computer programs written in C. It focuses on minimizing time spent by the CPU and gives sample source code transformations that often yield improvements. Memory and I/O speed improvements are also discussed."

Comments: 0 | Reply | Categories: CAD

My quick hacks to tax cache

(by Duncan, @ 4:49 PM)

My students are reporting some interesting results from my couple of C quick hacks to tax cache that:
"exposes more extreme pathological negative performance characteristics because of poor memory access patterns

They may be hacks but they are well-documented ;-). They also underline the current lecture material for my computer architecure and design class.

Comments: 2 | Reply | Categories: CAD



Copyright © 1999-2005, Duncan Smeed. All rights reserved.
 
May, 2005
Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31  
Mar  Sep
Tell ICANN to keep their hands off .org!


Run the HTML validator for this page
Webmaster: web at smeed.org