Mathstar plans to discontinue FPOA development

That is a surprise (to me).

Austin

From their stock call on May 11, 2008:

"First and foremost, the established lead-based programmable logic companies, like Altera and Xilinx, have had nearly 20 years to mature the software tools used by the customers to program their FPGA. MathStar has quality programming tools, but we haven?t yet developed the level of abstraction and automation that FPGA tools have attained. This is by far the biggest issue affecting MathStar?s ability to scale the business.

Customers have struggled to become productive with the FPOA design tools, although our internal teams of application engineers and certified design engineers have become proficient with developing even the most complex video encoder designs using FPOAs."

Seems it was tougher than it looked!

The days when "if you build it, they will come" are long over.

When the manufacturer has to do (all) the work to get the product into customer's sockets, that is a really, really, bad sign....

"scale the business" I suppose is how you say "make money" in this politically correct world?

It appears they had silicon (90nm) and they had customers. They also had a lot of stock-holders now holding pretty worthless stock...

Austin

If they really DID have working silicon, and customers, then you don't pull the plug,

Normally what happens then, is the second-iteration, (next generation) just fails to materialise.

However, if you find your Silicon 'needs another pass', or the customer you thought you had, was Lassoed by the competition 'buying the socket'... - or the customer did the effort/return calculation, and saw the road map of other vendors, and made the decision themselves. [ aka Apple and Freescale/intel?]

-jg

May not be? Sounds like the packet processor "threads" story all over again.

-- Mike Treseler