Transmeta Mini-ITX Board Reviewed

NobodyButMe writes "Transmetazone.com has posted a link to a 'world-exclusive' IBASE MB860 review on EpiaCenter.com. This appears to be the first review of a Mini-ITX board built around Transmeta's efficeon technology. Transmeta has also approved this board to be the official reference platform for the TM8600 processor and if you take a look at the benchmark results in the review (page 4) then you'll understand why as VIA's EPIA-M10K board looks quite pale in comparison. The review also adresses issues such as power consumption, temperatures and thermal throttling - three very interesting points when looking at the Efficeon processors. If the MB860 weren't so expensive (~500$ or something as it's aimed at the 'industrial market') then this could easily beat the EPIA boards (IMHO)."

Comments

Beat, by one measure

[Profane MuthaFucka]

Transmeta's mini-ITX board might beat a VIA board in speed, but VIA still has price going for them. And speed isn't that important to VIA's strategy, since their CPUs are meant to be fast enough for most jobs, but not the fastest CPUs available. They concentrate on power consumption. If Transmeta can lower the board price to $175, they would really have something good.

Re:Beat, by one measure

[bhtooefr]

VIA's trying to be all-around. Cheap (so the masses can buy - actually, this is goal #1 - the division that makes their CPUs was started by a couple former Dell engineers who were ticked that Intel was charging so much for their 386), small (look at Nano-BGA - smallest x86 around), fast enough, and cool.

Transmeta, on the other hand, is trying to be ice cold, and more importantly, low power. They're actually one of the physically LARGEST x86 solutions, and they're not cheap. Sure, they might be faster than VIA, but not by too much (I actually had a chance to read the article on Epiacenter several hours before the /.ing, but only glanced at it).

Transmeta vs AMD vs Intel

[fembots]

Is Transmeta the new AMD in terms of innovation and catering the real consumer needs?

I wonder if/when Transmeta's price has come down to $100-$200 mark, will it start to attract more users?

Re:Transmeta vs AMD vs Intel

[RAMMS EIN]

``Is Transmeta the new AMD in terms of innovation and catering the real consumer needs?''

err, well. What are the ``real consumer needs''?

If you mean ``what consumers want'', then TM is definitely not it. Consumers want higher, bigger, larger, and the Efficeon doesn't give them that with it's 1 GHz - which doesn't even go as fast as an 1 GHz Athlon, as the x86 crap^H^H^Hode is emulated.

If you mean ``what geeks want'', then the Efficeon might go somewhere. There's certainly interesting technology in there, and there are probably lots of geeks who care about power efficiency.

I am afraid, though, that the Transmeta we know is not going to make it. There simply does not appear to be too much of a market for their kind of chip. Perhaps a transformation, like making CPUs with programmable code morphing (so they can emulate pretty much any architecture) could save them. But then, they run the risk to become Jack of all trades, master of none.

I still don't really see why their chips are fundamentally better than other manufacturer's. Surely one could layer code morphing (which is, after all, software) on top of mostly any other CPU? And don't go telling me that there aren't other energy-efficient CPUs out there. It seems to me that other manufacturers could release Efficeon-like processors if they wanted to - and will if Efficeon turns out to be a success.

Transmeta has no direction.

[Thai-Pan]

Is it just me, or does Transmeta seem to be completely dropping the ball when it comes to catering their product to their own key demographic?

It's essentially built like a normal computer motherboard, but who in their right mind is using a low power embedded solution like this for a desktop? Really, people are using Transmeta's projects for places where low power consumption and small size are key. Like home theatre PCs, car PCs, and so forth.

Transmeta needs to get smart and produce products directly targeted at these embedded solutions; not vague products which could possibly be contributed towards them. If you want to build a home theatre PC, you need to hunt around for the motherboard, CPU, etc. from a normal computer, plus the chore of getting together a remote control system, small quiet power supply, suitable case that doesn't look like a budget computer from 1996, a fancy home audio sound card, etc.. If you want a car PC, you're going to be hunting for some very specialized input devices, screens, power supplies, etc. Why isn't anybody producing proper kits for these uses?

Just a little bit about Xenarc displays...

[apharov]

I work part-time for an embedded hardware manufacturer, and I've had a really bad experience about Xenarc displays. We got two pieces of some ~6" displays for a project, one of them didn't light up at all and the other one was somewhat funny looking. Being hardware guys we disassembled them of course.

The result: both of the displays had bent pins on chips and mauled PCB's. It seemed like someone had been intent on destroying the display internals with a screwdriver. In addition to that the working display showed an interesting fading pattern of something like Bubble-Bobble characters (the pattern appeared when TTL level control signal was cut off but backlight remained on). I googled a bit with the LCD panel partnumber and the only result I could find was some asian company selling really cheap panels recycled from some kind of gaming machines. Xenarc quality indeed.