Slashdot Mirror
New Transmeta Chip: "Efficeon"
ddtstudio writes "Oh, "Astro" was such a friendly name -- but it probably had trademark issues. So the alphabet blender came up with "Efficeon" instead. This eWeek story gives the lowdown on what Transmeta is doing apres Linus. There's also a writeup on ExtremeTech."
Does anyone besides me read this as "F-ing Con"? Maybe my problem is that seems like a good description of Transmeta's current business model.
Why do companies like Intel, AMD, and Transmeta keep having to name processors so they sound like they came straight from the mouth of Rob Schneider's SNL character The Richmeister?
The Celer-on, the Opter-on, the Athl-on, the Effice-on.
It's not good for marketing, guys! Everybody hates Rob Schneider!
Until Slashdot fixes the funny modifier, use insightful or interesting. The poster knows your intentions.
Transmetta specializes in low power computing. That's their niche - processors that don't eat up much energy. This really isn't a direct-sale market. The stuff they make is built into other devices. So their job is to convience sony and/or palm or whoever makes the low power (read - portable) devices that they are the ones to talk to. So I don't think there's really a need to become well known outside the computing industry.
To make laws that man cannot, and will not obey, serves to bring all law into contempt.
--E.C. Stanton
You can't copyright a name, but you can trademark it.
47% of all statistics are made up on the spot.
I propose we stop with these new-fangled blender names. I propose Transmeta call this new chip "The Scotsman."
Nothing is cheaper with the power than "The Scotsman!" Cue intel-sounding theme, but with bagpipes.
Transmeta chips are essentially a VLIW CPU, surrounded by a sophisticated JIT compiler that translates and compiles x86 code to the native instruction set of the VLIW CPU, and then runs the code.
This allows Transmeta to save on die space. Smaller die = less transistors = less heat = less power.
Unfortunately, this approach so far has yielded substandard performance. And even though power consumption was better than Intel's mobile processors for awhile, Intel quickly geared up, threw money and engineers at the problem, and came out with the Pentium M.
Arstechnica.com has speculated before that Transmeta could easily use the same approach to optimise for speed/performance as opposed to power consumption : I'm hoping they do.
Be kind. There are too many mean people out there already.
$350 million buys a lot of presence in magazines etc... Similar problem for linux in trying to get past the reams of Microsoft bought advertorials etc. in the magazines as well...
Transmeta's Efficeon will have to compete on it's technical merits, and when people who matter realise that it offers a much better power consumption, lower temperatures and much longer battery life they'll start to take it up. 3 hours or so with Intel Celeron 1500 is just not on when I was used to some 24 hours or so battery life on my old 8086. Hopefully, the Efficeon will enable them to make notebooks that can cope with a complete working day or more away from the mains outlet... RAM's cheap enough these days to enable them to give it a seriously large cache so as to minimise HD usage, and sticking the OS in a bootable flash disk will improve matters as well. Now we just need a very low power display technology such as high res colour "electronic ink" based thin displays
Donald 'Duck' Dunn: We had a band powerful enough to turn goat piss into gasoline.
Transmetta specializes in low power computing. That's their niche - processors that don't eat up much energy. This really isn't a direct-sale market. The stuff they make is built into other devices. So their job is to convience sony and/or palm or whoever makes the low power (read - portable) devices that they are the ones to talk to. So I don't think there's really a need to become well known outside the computing industry.
True, but companies who make consumer devices would be more inclined to listen to Transmeta if consumers were demanding devices with Transmeta processors. The way to convice consumers is with direct marketing. If you were correct, Intel wouldn't need to targer consumers either, since most of Intel's sales are to OEM, not direct either.
0xfeedface
Instead of doing this translation in hardware, Transmeta does this in software, and it enables a lot of optimization while (at the same time) vastly reducing the amount of hardware resources required to do wide, out-of-order execution.
They get varied results -- some things go much, much faster on the Transmeta, but it's very bad at doing other things (especially things like self-modifying code).
The internal architecture is also very geared towards translation and running translated code. There are features that allow it to run a bunch of code in a translation that is fast, but not safe. If there is a problem with this unsafe translation (memory exception or something) the execution can be rewinded (rewound?) into a known-good state and a slower translation or interpretation can be used.
Transmeta has released some good papers on this whole thing. If you're interested in this kind of thing, you might want to also check out HP's Dynamo and Intel's DAISY.
Yay, clever computer architecture!
-- Erich
Slashdot reader since 1997