•  
• EDN Europe MAGAZINE
•  
• Free SUBSCRIPTION
• Subscription Renewal
• View Media Kit

• EDN
• EDN Australia
• EDN Asia
• EDN China
• EDN Japan

EDN Europe's Editor Graham Prophet posts a selection of comments and insights prompted by the many items of industry news and rumour that cross the editorial desk or are gathered on his frequent travels to interviews, press conferences and events around Europe - and further afield - and somehow never find their way to the magazine or the web site, recovering some of the information otherwise lost in the noise level...

Tuesday, March 16, 2010

Are we there yet?

| Permalink | Email This | Comments (0) |

| Digg This | Slashdot This | add to Del.icio.us |

I posted yesterday an update on the product plans of the new FPGA market entrant Tabula; I call it new, but the gestation of the company has in fact been prolonged, and it has figured for a while in lists of potential new players in the sector. (As in; “….and then there’s Tabula but no-one knows quite what they are up to….”). Technically, the concept is challenging, and the challenge for Tabula in gaining entry to many markets will extend to convincing some potential users about the viability of programmable logic, all over again. Those users needed to be convinced first time around that they could depend on logic whose function was only retained by values held in SRAM memory cell: they may take some further convincing that logic whose existence is even more ephemeral will be something they can build a product on.
And then, there’s the problem of breaking into a market so dominated by two major players. For sure, Tabula has been well funded: but I offer you a quote from another player in the same market, who should probably remain anonymous; “Do you know how much venture capital has gone into programmable-logic startups in the last few years? At least $1.5 billion. And with what market impact? None at all.”
As with every new entrant in the field, I have to ask,, “Does this development constitute programmable logic finally living up to its ultimate promise?” Thus far, the answer has been, “no” – but I live in hope. Let’s go back a decade and revisit some of the directions in which the industry seemed to be heading. All the talk then was of the virtual corporation: re-run the last ten years, and we should all now be working from our garages, forming ad-hoc groups of designers working on innovative and advanced, mass-market product concepts. We would be buying-in multiple pre-qualified blocks of IP, assembling them in a virtual prototype, and the EDA tools would quickly reduce them to design files for a single-chip SoC device.
Of course, that vision never came about; large-scale ASICs and SoC devices became impossibly expensive, in terms of design- and pre-manufacturing costs, for any but the largest projects and companies to contemplate; and the largest FPGAs, that would support designs on the SoC scale, continue to cost too much to be the basis of high-volume production runs. That’s not to deny that each has well-established market successes: simply that the technology to implement that grand late-90s vision has yet to emerge. The IP market, too, never matured in the way that its proponents hoped; you could argue – perhaps a little more tenuously – that the absence of an affordable implementation technology held back growth there as well.
Now comes this new contender promising system-scale logic resources at volume-market prices (or nearly so, in the 40-nm generation). Is the idea of complex, but affordable, custom silicon-for-all, finally within our grasp? I’m not about to make that prediction: but there appears to be enough about this offering to deserve close study, at the very least.
Now, how about a few programmable-analogue tiles on the same chip to make it truly the basis of a single-chip product? I asked Tabula CTO Steve Teig that question… he appeared to think he had enough to deal for now, introducing the logic device – but he didn’t rule it out.

Related entries in: EDA/ IP | Programmable Logic/ Memory |

(Add your comments)

Post a comment

* Display Name
* Comments
* Anti Spam security code	To pervent spamming, please enter the anti-spam code as shown above to proceed with the submission