The mystery of whether Blaze DFM had closed down or not is over. It has been acquired by Tela Innovations mainly, it would seem, for the PowerTrim technology that had been licensed by TSMC. A major strategic relationship between Tela and TSMC was also announced. Both these are interesting developments..
Tela acquiring Blaze is interesting primarily because neither company is public: one private company is acquiring another (Tela is not venture-backed but funded privately, which is a story for another time). I think that this is something that we will see more of in future, especially in the year ahead, along with other acquisitions as small companies run low on cash and need to find partners. Historically small EDA and semiconductor companies would either get large enough to go public or be acquired by the larger healthy companies in their industry. Obviously for the time being and perhaps for a long time the public markets are closed, and the larger companies are not really healthy enough to be making any major acquisitions (despite Synopsys’s better than expectation results) nor do they have the cash to pay up. And nobody is borrowing anything anytime soon (with the federal government wanting to borrow a trillion dollars why would people with cash lend to anyone else).
The other interesting aspect of this is TSMC partnering with Tela over not just Blaze’s technology but their own. The basis of this technology was described in a guest blog last week by Mike Smayling. This was coincidental since I didn’t know about this announcement then. It just seemed a good time to talk about lithography since this week it is SPIE, the big conference in lithography. And remember, Moore’s law is largely about driving lithography and its ecosystem forward.
TSMC will likely be using Tela’s technology to create value added options for their foundry users, They had previously done this with the Blaze technology, supplying it as a premium service with a wafer royalty. Now, although I don’t know the financial details of the details, presumably doing the same with Tela but on a much larger scale. Tela’s technology is much more widely applicable and the alternatives are much more limited, especially at the nodes below 45nm where current OPC techniques run out of steam and some form of restricted design rules becomes essential. You simply can’t print arbitrary layout any more.
Going back to beginning, and whether private companies or public companies, are the most likely eventual home for venture-backed EDA more than EDA companies companies, I’d have to offer pretty good odds on TSMC acquiring Tela at some point in time. But Tela also has Qualcom, Intel, KLA-Tencor and Cadence as strategic investors, who are all also potential suitors or blockers. Of course a TSMC acquisition might cut off Tela revenue from the other major process hubs (Intel, UMC, Japan Inc, IBM/Samsung) but from TSMC’s point of view that would probably be a good thing.