As AJ intimated, when considering the power and constraints of the cloud relative to EDA, it's really in the "eye of the beholder".
For back-end tools we are challenged on two fronts - the amount of data to be transferred (bandwidth) and potentially the responsiveness of manipulations of graphical models (latency).
The closer we move up towards the front-end of the design flow, the less relevant these concerns become.
Indeed, the power of spinning up 1,000 servers to perform verification regressions in a day rather than a month is powerful, with no latency constraints of significance wrt the UI. Also, at higher levels of design abstraction, how much data really needs to be transferred?
The business model is likely to be the only real hurdle to rolling out a commercial SaaS solution. But in reality, that could likely be boiled down to mathematics and product packaging.
Xuropa.com is an EDA cloud computing service that is being used by its customers (including Synopsys and Cadence) for marketing and limited evaluations (in an enclosed, online environment).
It's hard to see EDA tools operating in an open, public environment like GoogleDocs but internal clouds for large enterprises should provide enough of a secure environment, I would imagine.
My Mom the Radio Star Max MaxfieldPost a comment I've said it before and I'll say it again -- it's a funny old world when you come to think about it. Last Friday lunchtime, for example, I received an email from Tim Levell, the editor for ...
A Book For All Reasons Bernard Cole2 comments Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...