Re:Relative performance? (Score 1)

I am pretty sure that relative performance of their 400-700Mhz CPU (which btw does not exist yet, they are working on the prototype... Hmm.. paper CPU, how nice...) Will compare to at most a pentium 160Mhz if that. "Caching blocks of x86 instructions" - WHAT? In order to determine if a fetched "block" is a hit or miss they need to compare this entire "block" to contents of their "translation table" This comparison already increases this poor "block"'s processing time by number of entries in this "translation table"... Please... Also these transmeta guys openly lie about Intel CPUs in their white paper saying that intel translates x86 instructions before executing them. This is true for the 16bit x86 instructions, but not for 32bit... Or do they mean that they only emulate pure x86? Hmmm... I guess it's back to the DOS/win3.1 days for them then... Anyway this whole thing is a bull, since when do emulators run things better or fastern then the real thing? Oh well, I guess slashdotters will be content with a cpu run on water as long as Linus had his hands in it...