IBM regains supercomputer bragging rights

TalkBack 9 of 11:

Next »

« Previous

• Thread View
• Flat View

Linearity

Just because you double the # of processors and don't get double the performance doesn't mean that it's not linear. Given the numbers you quoted, the performance would be linear with an equation of:
P = X * Q * N where N > 1
X = Scaling factor (0.8 in example quoted)
P = Performance of the system
Q = Performance of the processor
N = # of processors

Ideally, X should be one or more. Unfortunately, real-world things like interprocessor communications and memory contention limit X to less than one.

I think that getting a scaling factor of 0.8 is pretty darn good. Especially given the number of processors involved in these machines.

Posted by: Letophoro   Posted on: 06/21/04 You are currently: a Guest | Members login | Terms of Use

Alert moderator to an offensive message

Subscribe to this discussion via Email or RSS

SponsoredWhite Papers, Webcasts, and Downloads

• Three Steps You Need to Know to Stop Data Loss Varonis Sensitive data exposed to misuse or loss... it is the stuff of nightmares ... Download Now
• The Impact of Virtualization Software on Operating Environments VMware Today's use of virtualization technology allows IT professionals to ... Download Now
• Virtualization: Architectural Considerations And Other Evaluation Criteria VMware Of the many approaches to x86 systems virtualization available in the ... Download Now