Saturday Night Blues(pec)

There was an interesting article by one of the guys of BlueSpec, about how the threading model is deficient.

 It is a topic often touched upon by George Harper in several of his articles (and my comments! )

‘Tis but a simple enough approach - Let thy not believe in thy own scheduling. Ask for what you want, and the compiler shall provide.

The problem with this is not the tech - which I wont argue for or against - but rather the value and business proposition of it.  The EDA industry is built on trust (or lack thereof). Design companies have teams which even work on the same problem simultaneously  to cut down on the chance of error. 

The problem is in trusting the tool to generate good glue/scheduling logic. Apart from the obvious need for a system level sequential verification tool (which I wont harp upon ), there is the obvious problem on selling this proposition.

 I do hope they have good FAEs!! Tags: eda startup


Saturday Night Blues(pec)


June 03, 2006

Find me on Twitter @sandeepssrin

Did i make any mistake? Please consider sending a pull request.