Russian wannabe Merced-killer claims Transmeta credit
Saddened by lack of name-check at launch - sabres at dawn impending?
Elbrus International, the Russian chip design outfit that says its E2K chip design would be a Merced-killer if only it could raise the money to build it, is claiming Transmeta's Crusoe family of chips owes a lot to its designs. Crusoe, which was publicly unveiled earlier this week, is intended to combine high performance, low power and low cost with multi-chip compatibility via what Transmeta terms "revolutionary technology." But there are signs that maybe Elbrus reckons some of this might be its revolutionary technology. In a phone interview with our good friends at Dow Jones Newswires, someone referred to as Elbrus director Sergei Babaian is quoted as saying "Transmeta has proved that our idea works." It's maybe early days to be quite so sure about the "works" bit of this, of course, and as Elbrus' own site doesn't seem to mention a Sergei Babaian, we can't help wondering if maybe this mightn't be the rather better known Elbrus director Boris Babayan speaking. That's not of absolutely vital importance, however. Babaian/Babayan says that he's happy for Dave Ditzel, but sorry that he and Transmeta contrived to roll out Crusoe without acknowledging that "he had also learnt a lot from us." This might just be hint at fighting talk. There is at least a connection between Ditzel and Elbrus, as the Russian company worked with him while he was at Sun. Elbrus also lists Transmeta as one of the companies it has co-operated with, although it isn't specific about the nature of this co-operation, and Transmeta is not listed as one of Elbrus' "partner" companies, where the relationship, one presumes, is likely rather stronger. Elbrus announced some details of its E2K processor almost a year ago now, and since then has been seeking, unsuccessfully, funding to take it even to prototype stage. Since Elbrus is only in pursuit of $60 million initially, and Transmeta has already run through several hundred million in startup money, a little envy might be understandable. Elbrus doesn't seem to have made much headway since the E2K announcement, but it may be significant that in November it said: "The work to secure the IP rights for the E2K has been done." At that time Babayan also noted that the disclosure of the "basic parameters" of Merced/Itanium and Transmeta confirmed that Elbrus could still be ahead of these competitors for "a substantial period of time." What happens next may depend first on whether Elbrus' luck turns on the investment front, and second on how securely it has managed to secure its E2K IP. In the first department, the Moscow office of UK outfit Robert Fleming has been on the case without obvious success for a couple of months now, while in the latter..? ® See also: Ex-Soviets seek $$$ for Merced-killer - Transmeta linked? Babayan outlines Elbrus future
Sponsored: Benefits from the lessons learned in HPC