Slashdot Mirror
Transmeta OK'd for Mira Displays
viewstyle writes "eweek is reporting that Transmeta's Crusoe chip has been approved by Microsoft for use in the Mira smart displays -- a move that further edges out AMD and Intel from the mobile processor marketplace."
x86 is only one processor supported by the Mira platform.
To say that Transmeta is edging out a giant like Intel who has the PXA255 is just wrong. Transmeta doesn't have any edge in this market whereas Intel certainly does.
I have been pwned because my
The Transmeta chip being approved for use in one device hardly qualifies as edging anyone out of the mobile chip market, much less "further" edging them out. Besides, what do we (read as: geeks) have against AMD? I thought we liked them, no?
With the dot-com bust giving Transmeta a hard time, an approval by Microsoft could be just what they need to get back on their feet and out of bankruptcy.
Maybe this will be a lesson learned for the GNU/Linux community as well, to support the companies that use open source and contribute back to the movement. Everyone's all about freedom, but nobody seems to want to pay the price for it. Apparently Transmeta has seen this and is moving on to greener and more friendly pastures. Good for them I say, and I'm looking forward to the product.
Think about that the next time you Bittorrrent the latest release of Redhat instead of purchasing it. What would you do if they decided to switch over to Microsoft? You'd all be screwed then.
... edged right out of their tiny 15% and 80% market niches. Tremble, AMD and Intel, tremble.
Intel has the new XScale, and AMD has their Alchemy processors (MIPS32) for the Mira displays. AMD demoed the Mira running with the Au1500 recently, and ran better/faster than the PXA stuff from Intel. I saw it, it was reasonably cool.
The sad thing is, they're just re-inventing X-terminals. Mira is just a dumb display, the magic still happens on some big, beefy iron in the background. Everything old becomes new.
Going for the mobile market is indeed a mistake.
They went for the blade-server market, which was not a mistake.
Embedded (under certain circumstances) is not a mistake either, but it's a very competitive market.
They've ignored the quiet PC market, and that was a huge mistake.
OK, here comes the clarification. The power savings of TMTA chips don't matter much for mobiles, because the power profile is dominated by the monitor and other devices, not the CPU. The one advantage the chip does offer for mobiles is low heat disipation, but they didn't pitch that. Instead, they marketed enhanced battery life and when that (predictably) didn't materialize, they started to look like Enron execs.
The power savings matter a great deal for server farms, because when you jam 1000s of CPUs into a room, all those watts add up quickly, and you also reduce the cost of cooling the room. In general, when TMTA's chips are scattered all over the place, the heat dissipation and power savings don't matter that much. When the chips are packed together in small spaces, then it matters, which brings us to...
...The quiet PC. The low heat allows TMTA chips to run FANLESS. This wasn't emphasized enough by their marketing or engineering departments, and as a result there are virtually NO PCs being made with TMTA's chips, and there are no inexpensive motherboards using a TMTA chip. That market has been conceded entirely to VIA's mini-itx boards. This is a real shame, because apparently the TMTA chips can (in theory) perform better than VIA's chips and do so without a fan. Also, TMTA squandered its geek appeal by making it very difficult for anybody to experiment with their chips. I haven't seen any SBCs with a TMTA chip in them for under $600. The chip itself is available for $90 in single units, but there is nothing to plug it into. Since Intel chips cost about the same, I don't see any argument for why TMTA can't get somebody to make inexpensive MoBos. That is an oversight that must be corrected.
If you look at the mini-itx.com site, you'll see that most of the projects are hobbiest kitsche. I imagine that most of the projects at the original Homebrew Computer Club were like that too; but one of them was the first Apple. Unless TMTA gets its /\55 in gear, it will end up like Zilog instead of Motorola.
It's very, very ironic that Linus works for a company that has failed so miserably to understand the importance of hackers and the "garage" in driving the industry. I haven't heard him say anything about advocating within TMTA to make their hardware more open. Notice, I mean open in the sense that you should be able to buy components separately and tinker with them; I'm not asking TMTA to give up any copyrights or patents.
For example, in theory, you can emulate any CPU by writing your own code-morphing software. Now, I defy anybody to find the documents that tell you how to do that.
For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
Good god! Do people STILL believe in this myth of some "Wintel" connection?! Microsoft and Intel have never been more than forced friends, and for the most part they've been more like advisaries!
The only company that Microsoft can't push around in the PC business is Intel, and the only company that Intel can't push around in the PC business is Microsoft. For this reason alone the two have been trying their damndest to find alternative suppliers. Both companies realize that they are heavily dependant on the other for their own success, but given half a chance to support a third party, they always jump all over it. Intel has been a strong supporter of Linux, while Microsoft has gone out of their way to support AMD and now Transmeta.
The whole "Wintel" thing exists only because Microsoft and Intel have pretty much been forced to work together, whether they like it or not.