Technology Stocks : Microsoft Corp. - Moderated (MSFT)


 Public Reply | Prvt Reply | Mark as Last Read | FilePrevious 10 | Next 10 | Previous | Next  
To: dybdahl who wrote (17079)10/21/2009 7:37:42 PM
From: QwikSandRead Replies (1) of 18410
 
I think you'd have to elaborate on the kind of applications you're talking about.

I don't believe a cloud office suite's quality would be measured on an isolated metric like bandwidth to the disk...I'm not even sure bandwidth to the disk has a precise meaning in that context. The "cloud" makeup is too complex.

Transaction-per-second benchmarks have been around for decades (TPmC etc.) and can be easily applied to cloud transaction systems in a lab. What I don't know is what kind of crosstalk effect exists within the "cloud" (whatever that means in any given situation) among different kinds of application environments. But I suspect Google and Microsoft can measure that too.

In any case I can't imagine that the performance of applications that mostly perform database transactions is all that difficult to measure, at least in isolation.

Applications that are very compute-intensive or need reliable isochronous-type I/O performance, i.e. music production (not just listening) or video editing (not just viewing) or anything that potentially does heavy computation on concurrent voluminous data streams in real time, I doubt there's a cloud infrastructure for that yet. In those cases the bandwidth to the disk is important but could easily be dwarfed by network latency or hiccups. Those have to stay on the local workstation and are not cloud candidates yet, although clearly that will change.

But you'd have to state specifically what you have in mind before I understand your previous comment.

--QS
Report TOU ViolationShare This Post
 Public Reply | Prvt Reply | Mark as Last Read | FilePrevious 10 | Next 10 | Previous | Next  

Copyright © 1995-2013 Knight Sac Media. All rights reserved.