Slashdot Mirror
Transmeta To Release Next Generation CPU
CodeShark writes: "According to this story at CNN, Transmeta is set to release their new TM6000 microprocessor this afternoon. The chip apparently incorporates some of the functions usually provided by high-performance (and high price!) chip sets. Transmeta is reporting a further reduction in power requirements by 44% and sees the laptop and sub-laptop markert as the primary markets for their new CPU. Intel and AMD claim to be catching up with the Transmeta chips in terms of power requirements, I'd be curious to find out what the real world comparisons might make of those claims ..." If anyone out there is at Microprocessor Forum, please say in comments any further details that are made clear there.
They're announcing it, not releasing it. Production won't begin until the second half of 2002.
I'm curious about what information AMD and Intel have to back up their claims that they're catching up to Transmeta in power requirements.
I just got a toshiba laptop earlier this year with a 700mhz celeron. I love it but I rarely use it without being plugged into the wall, as from my experience it only lasts about 2-3 hours.
I remember seeing stuff saying a laptop with a transmeta chip can have a battery life of about 8 hours.
Assuming that is true, how could Intel and AMD possibly say they are catching up? I mean mine is a celeron, not even a pentium III or anything and it sucks up power like I would have never imagined. I hope Intel isn't talking about their powerstep technology, that is just a freaking joke.
Anyone with more information on power consumption among the different chips, I would think Transmeta would have tons of information about this since it's really their main selling point isn't it? I better go check their site.
FiGZ.COM - A waste of perfectly good web space
As is often the case, the Register has some really interesting comments on this story here. Apparently this release has a lot of market control and damage control related to it. There is a class action suit going due to previous claims of high speed chips. Anyways, read the Register article for more details.
Transmeta refuses to release any industry standard benchmark results for their CPUs.
Ask them. If you get something other than FUD back, please post it.
Why won't they run the SPEC int and FP tests??
They try and hide behind low power claims and can spin FUD with the best of'em. Low power means absolutely nothing unless you know how much WORK it can do.
They will give you benchmark results only if you sign an NDA and promise not to tell anyone how slow their chips are. Most companies who sign the NDA decide not to use their product. What does that say?
I'd really like to see these guys compete with Intel/Rambust, but I have no respect for companies built on FUD, regardless of who is involved.
Part of the deal with the Crusoe-chips is that they do "code-morphing" and morphs x86 instructions into something the crusoe can handle.
What if the crusoe chip could do the same to PowerPC-code?
Imagine dual-booting MacOSX with Linux x86 and Windows.
Now, that would be interesting, (and probably not something Apple would like).
One concern that goes through my mind when I look at the not very stunning performance of Crusoe is the effectiveness of VLIW (very long instruction word) processors.
Both Transmeta and Intel have bet that VLIW processors are the way forward. Intel's Itanium and Transmeta's Crusoe are both based around the VLIW concept. Transmeta hides the VLIW nature of Crusoe behind the 'Code Morphing' software that allows the chip to be IA32-compatible - Intel's IA64 architecture gives compilers raw access to the VLIW nature of the processor, and has (very slow) on-chip emulation of IA32.
Between them, they make up the only commercial VLIW processors around, and both are very poor in terms of performance compared to more conventional modern processors, whilst at the same time introducing some enormous obstacles to overcome - IA64 requires some very major changes to the way compilers work, and Crusoe requires major extra complexity in the form of the Code Morphing translation layer.
I don't wish to jump the gun, but I think this means things don't look too bright for the VLIW concept. Evolutionary enhancements to conventional RISC/CISC processors appear able to continue Moore's Law for many years yet. AMD has outright rejected VLIW for its future 64-bit strategy (x86-64) and none of the other major CPU manufacturers seem to be jumping on board either.
Have Transmeta and Intel made a very large strategic mistake? VLIW looks good on paper, but is it effective on a practical level?
It will certainly be interesting to see what happens with future Crusoe and IA64 processors.