I'll don't see transmeta as any kind of a sure thing. A repeat of three past posts.
Message 14653567
my feeling on transmeta is one of that's nice. The Benchmarking of the tranmeta in product as
being about a 300mhz pentium 3 is not impressive. My feeling is that all designs will be small with
800x600 screen and a small pain in the ass to use keyboard. I don't think any designs will have
1280x1024 or 1600x1200 size displays. Battery life is also most likely factored in screen size. I don't
know considering small size factors vs extended battery life if that transmeta laptops will be all that
desireable. As msft oses are piggy going on piggier, over time unless the performance of the
processor increases the perceived slowness of the tranmeta system may not be very attractive.
Within 2 or 1 year no doubt Intel will have incorporated many of the functional power saving
techniques of the transmeta chips. I applaud the technology but I not impressed with the products
so far.
Message 14678102
I don't see transmeta as disruptive. I think of the transmeta innovation in two areas. One is an end
to end analysis of user's need and power use. The other is the emulation engine or whatever you
want to call. Emulation clock for clock cannot compete with direct coding. The max performance of
the transmeta chips is a testament to this.
The innovative power saving techniques all can be copied partly and for the most part in whole by
amd, Intel, whomever over time. The emulation will never compete clock for clock with direct
code. So amd and intel chip users will be able to implement systems with far more dynamic range of
compute power which yes will impact battery life when needed but be just as stingy when not. User's
could also be given nice choice, user selects max cpu speed to extend battery and loses response.
Message 14679320
cheaper, simpler, smaller, and, frequently, more convenient to use."
Cheaper, Intel has own fab,
Simple, emulation, whatever you want to call it is an entire level of complexity. everything to
everyone.
smaller, ultimately a processor designed to task uses fewer transistors and fewer leads to smaller in
time.
More convenient, with twice the power I can do in half the time. Yes my battery won't last. But as I
mentioned earlier much of the innovation of the crusoe is just a program or process that
intelligently determines what power make sense and this feature can be embedded in the bios or OS
of any system.
The above is why I don't consider crusoe a disruptive technology. The continuing shrinking of
feature size in IC's means more and more surplus processing power at less and less power
consumption.
Now laptop's is one area of the market. If one looks to the embedded market, ac available
applications are numerous and power is of little import. In this space designing a processor core like
mips make for a cleaner design. Also mips for example has lots of folks who know how to program
it and their is a tremendous established infrastructure of prototype and software development and
simulation tools.
In mobile space me thinks that mips core can be fabed in a technology that is very low power and
you still have all the support in place.
I just used mips as an example. Yes I do own mips stock.
Tom Watson tosiwmee |
