The problem for the FPGA companies is growth. They’ve been stuck in a $3 billion niche since the year 2000 and the only significant progress for their products has been shrinking geometries and adding blocks.
The obvious way to go to make FPGAs more widely used, is to do something to improve the two great drawbacks of FPGA – cost and power.
And the obvious way to do that would be to try and find a programmable DSP approach to programmable logic.
Is Altera following that route?
“The ROI for developing programmable DSP processors can be challenging”, David Greenfield, senior director of Hardcopy at Altera, told me last week.
“We continue to look at it and, if the right business opportunity presents itself we will look into the different options”, he added.
“We are assessing both development efforts in-house and looking at other technology”, said Greenfield, “we are paying attention to third parties. We have involved our university partners.”
So is Altera going for a programmable DSP approach to FPGA? Well that’s not exactly a Yes but, on the other hand, it’s not entirely a No.