Will fpga designers be spoilt for choice?

1 min read

The programmable logic market is notable not only for its products, but also for the large number of companies which have tried to get a bit of the action. Over the 30 years or so of programmable logic, some 60 companies have tried to break into the fpga business – including some 'big names' – only to fall by the wayside. Of this extensive list, only four could be deemed as having succeeded in any meaningful fashion – in alphabetical order, Altera, Lattice, Microsemi (the new owner of Actel) and Xilinx. The first and last in this list hold the lion's share of the market.

Over the years, these two companies have ratcheted up the pressure; not only on competitors, but also on each other. Both are pushing their technology to limit, designing new devices for leading edge processes and incorporating the highest speed transceivers to meet the seemingly insatiable appetites of the communications industry. And yet, despite the apparently slim chance of success, new companies keep coming to the sector and, until the last couple of years, failing.. But it appears things may be changing; three relative newcomers – Achronix, Silicon Blue and Tabula – are showing what marketeers would call 'traction'. Achronix is set up to take on 'Xitera' head on; its asynchronous fpga technology is aimed four square at the comms market serviced almost entirely by the big two. Tabula is also set to make a pitch at the comms market, noting companies like Cisco Systems as a tier one customer. The third company, Silicon Blue, has been moving along quietly. Launched five years ago, the company came out of the shadows in 2008. Many observers didn't warrant the company a second glance. The reason? It was basing its approach on a patent filed by Xilinx in 1986. At the time, the company said it was taking an old architecture and reworking it from the ground up using new technology. The result was iCE65, a range of ultra low power fpgas manufactured on TSMC's 65nm process. Now, the company has taken the next step with the launch of iCE40, targeted at TSMC's 40nm LP process. It believes its technology will find application in smartphones and tablets. It's interesting to note that, while there is a trend towards fewer, not more, suppliers in many sectors, the opposite is happening in the fpga world. Could it be the programmable logic market is bucking the trend and that designers will be able to have more, rather than less, choice?