Slashdot Mirror

Caino

Don't touch my .sig there!

Caino

Don't touch my .sig there!

I like "FillIn" as the name of the Midori network upgrade system.

... it must be cool then, eh ? ;)

http://midori.transmeta.com/midori_logo_face.jpg

There is no such thing as an "Intel binary", if there were we could run windows apps on linux, vice versa, and whatnot.

Granted the assembly language to run on x86 machines is quite differerent from that to run PPC, Alpha, Sparc and others (and much less efficient too), so you could have meant that the "Intel instruction set", but that's a different thing. Assuming you're writing in a cross platform language (C/C++/Java) like a normal person, going from Linux to Windows isn't any harder than Linux to MacOS, or Solaris or whatever, all you have to do is recompile.

Although this qualifies as a commercial implementation, the instruction translation of which you speak is the basis of what Transmeta's doing with their code morphing stuff, but that's a bit offtopic.

--

We just discussed about power shortages and you keep your hungry hardware ?
Come on !
i just think we should take a deeper look to the low-consumption alternatives around, like this, or this.
What ? Vapourware. Nope. I own many machines running these processors and my brother just bought a transmeta laptop which he's in love with.
Don't believe the hype and aim your purchases towards a brighter future.
Intel's selling radiators, so is Nvidia.
--

You can download Darwin for x86. But that won't give you Quartz/Carbon/Cocoa (Mac OS X userland libraries) which are all proprietary binaries for PowerPC machines; good luck finding a machine that can code-morph into both an x86 (for the games on your Windows partition) and a PowerPC (for Mac OS X's userland). -- Dr. Mario game for PC based on Mac OS X's Aqua theme
Like Tetris? Like drugs? Ever try combining them?

Who would run Gnome or KDE when they can have Aqua

People who also want to play games. We are in the phase of the gaming market when the most heavily marketed games are on the x86 PC because that's the largest audience; the new consoles (specifically Sony PS2) do not have a foothold. It was similar in the Doom/Duke/Quake days just before the PSX and N64 came out.

The parts of Mac OS X that run on top of the Darwin kernel are only available as binaries for PowerPC-based Macintosh computers. I haven't seen a CPU that can code-morph into both an x86 (for games) and a PowerPC (for Mac OS X).

Yes. Sony has a VAIO laptop that includes a Crusoe. Check it out.

Not sure if you are joking, so I'm answering seriously. It's clear to me they had the mobile market in mind all along; their design is tiny and dissipates a very small amount of heat, and they have a feature for dynamically changing the clock rate. Existing "mobile" chips have a feature for idling the CPU to save power, but Transmeta can actually slow the clock rate down. More on this, plus cool infrared photos of heat dissipation, here.

I would seriously love to have a desktop computer with two or four Crusoe chips in it. I dream of having a computer running Linux quickly yet as quiet as the Atari 520 ST we used to have. (The Atari had no cooling fans, and no hard disk; if the floppy disks were idle, it was silent.) If a single 600 MHz Crusoe runs about as fast as a Celeron/300, two of them ought to be plenty for reading my mail and such. And it should be possible to cool them with just heat sinks.

steveha

Digital (Compaq) developed an x86 Dynamic Binary Translator running on Alpha called FX!32. FX!32 won Byte Magazine's "Best Technology" award at Fall Comdex '95.

Dynamic in this case means that some code is emulated on the fly, and some is translated. This approach was pioneered for bytecode systems in Smalltalk implementations in the 80's, and of course is now used in Sun's HotSpot and other dynamic adaptive JVMs.

Static binary translators have been around for even longer, and were used (among other things) for running VAX programs on Alpha. A useful overview of this sort of technology appeared in the Digital Technical Journal 4:4 (1992). HP also performed binary translation between the HP3000 and the Precision architecture, but I can't find on-line info on that, just a citation to a paper article (1987). There is also a useful survey article on static and dynamic binary translation.

What is presumably novel in Transmeta's approach is that their instruction set architecture (ISA) is tuned specifically for dynamic translation (see page 12ff of Transmeta's paper The Technology Behind Crusoe Processors . Some microcode architectures have been designed specifically for general emulation (most have been tuned for a particular macroinstruction ISA), e.g. the early Lisp Machines (1976-81).

The Crusoe's not really aimed at the PDA market. It's way too overpowered and as energy stingy as it is, it doesn't come anywhere near todays PDA processors.

The Crusoe TM 3200 - 333mhz (Transmeta's most "mobile" processor) consumes ~15mW of power in it's most power conservative state (deep sleep idle) and averages 1.4W (mp3 playback) (both of these figures include the Northbridge chipset power consumption as well).
By comparison, an entire Palm Pilot Pro (using a 16mhz Dragonball eats ~26mW in it's idle state and consumes ~160mW when running CPU intensive applications.

While you could leave your Crusoe on the shelf and outlast a Palm, It wouldnt take very long for 1.4W to drain a pair of AAA batteries once you start doing something with the unit.

The Crusoe's not really aimed at the PDA market. It's way too overpowered and as energy stingy as it is, it doesn't come anywhere near todays PDA processors.

The Crusoe TM 3200 - 333mhz (Transmeta's most "mobile" processor) consumes ~15mW of power in it's most power conservative state (deep sleep idle) and averages 1.4W (mp3 playback) (both of these figures include the Northbridge chipset power consumption as well).
By comparison, an entire Palm Pilot Pro (using a 16mhz Dragonball eats ~26mW in it's idle state and consumes ~160mW when running CPU intensive applications.

While you could leave your Crusoe on the shelf and outlast a Palm, It wouldnt take very long for 1.4W to drain a pair of AAA batteries once you start doing something with the unit.

Transmeta.com

Even though the story was rejected, one piece of news today on Yahoo Business was that the US Army has announced they will be using Transmeta chips for their battlefield backpack computers.

Who needs Compaq anyway ...

If it's so easy to match the power/performance of the TM chip, why is it only now (w/ TM chips) we are seeing laptops from NEC, Fujistu, etc. with 8+ hour battery life claims?
NEC fujitsu

If it's so easy to match the power/performance of the TM chip, why is it only now (w/ TM chips) we are seeing laptops from NEC, Fujistu, etc. with 8+ hour battery life claims?
NEC fujitsu

Six hours is OK, but in cases when I've been six hours from an outlet, I've not been able to use a notebook at all, and resorted to handwriting notes. However, the Crusoe will enable people like me to get out into the sticks and work:

http://www.transmeta.com/crusoe/lowpower/

Using the same battery-saving measures, I'd get another two hours over your iBook, enabling me to work in very remote locations.

As Treasurer of the Bucknell University ACM (Association for Computing Machinery), myself and the other officers help to persuade industry, faculty, and students computer experts or evangelists to (of OOP, OSS, Linux, etc) come to Bucknell to give a presentation. In the past year or so, we've had guests like Dan Quinlan of Transmeta, speaking on the Linux Standards Base, Ralph Droms (inventor of DHCP), a faculty member at Bucknell, John 'Maddog' Hall (Linux International executive director) on the Flexibility of the Linux OS, and many others. Currently, Eric S. Raymond has added us to his mailing list and will probably come Spring semester to talk about his ideals and beliefs when it comes to software.

What are our methods of obtaining guests? First, it helps to have some connections with someone related to the person you'd like a have speak at your school. Second, being at a top-notch college like Bucknell University tends to give some incentive, perhaps, for people to visit. Finally, persistance does pay off occassionally; if there's someone you really want, make sure you remind them via email or vmail every so often that you'd be absolutely delighted to have them grace you with their presence ;-D

Good luck!

Michelle wrote:
Plans are also in the works for a model integrating a low-power 600Mhz chip and 128MB of RAM. (Now from where does that sound familiar?) The folks at VIA promise an announcement about that new model at Comdex, but there aren't that many lines to read between here............
What does that mean? I need some help reading between the lines! Michelle

The the low-power chip would be possibly Intel's new low-power 600Mhz Pentium III or more probably the revolutionary Crusoe processor made by Transmeta. Most Slashdotters are big fans of Transmeta, which only recently began shipping products.

Use the search tool at the bottom of the Slashdot homepage to search for Transmeta articles.

The server is not responding at the link you provide, so I can't see what it is you're trying to point to. In the meantime, you may want to reviewthispa ge, and also thiswhit e paper(PDF). I'm not saying Dr. Pratt is wrong (that'd be kind of difficult, since I can't even find out what his claims are), but the information presented here directly contradicts your claim, so SOMEONE'S off the mark. I'll make my own judgements once I'm able to read what he has to say. If you know of another site where whatever it is you're trying to link to is available, or can quote it here, please do.

The server is not responding at the link you provide, so I can't see what it is you're trying to point to. In the meantime, you may want to reviewthispa ge, and also thiswhit e paper(PDF). I'm not saying Dr. Pratt is wrong (that'd be kind of difficult, since I can't even find out what his claims are), but the information presented here directly contradicts your claim, so SOMEONE'S off the mark. I'll make my own judgements once I'm able to read what he has to say. If you know of another site where whatever it is you're trying to link to is available, or can quote it here, please do.

The Crusoe chip may not be using any new ideas but it does take things farther than anyone else has tried. the PowerPC was to be a risc chip that could emulate cisc. crusoe is a riscish chip that emulates the x86 instruction set.

Actually if i remember correctly, the crusoe is a VLIW core, source:DITZEL.PDF pg10, when the chip was first released. The Document may be here somewhere


How every version of MICROS~1 Windows(TM) comes to exist.

I know Intel's chip naming is awful, but we should at least cope with it.

I miss the days of Z80A, 65C02 and TI9929 but hopefully the guys at Transmeta are working so we can again compute with TM3200 and TM5400 instead of Celerons, Pentiums, Durons and Weirdiums. Meanwhile, K6-3's and K7's are still fine. :)

info on crusoe chip:

Technology used
A whitepaper. The Technology Behind Crusoe(TM) Processors (pdf)
The processor's

from this info on transmeta's website the crusoe chip don't look too bad. it appears as if they've simply created a pretty good chip not based on x86 and then use the software as an interface between software and hardware.
they say:

The Code Morphing software can evolve separately from hardware. This means that upgrades to the software portion of the microprocessor can be rolled out independently of hardware chip revisions
and:
Crusoe processors provide speeds of up to 700MHz in mobile platforms

Not a bad idea really... so long as it works.

tahpot
on /.

info on crusoe chip:

Technology used
A whitepaper. The Technology Behind Crusoe(TM) Processors (pdf)
The processor's

from this info on transmeta's website the crusoe chip don't look too bad. it appears as if they've simply created a pretty good chip not based on x86 and then use the software as an interface between software and hardware.
they say:

The Code Morphing software can evolve separately from hardware. This means that upgrades to the software portion of the microprocessor can be rolled out independently of hardware chip revisions
and:
Crusoe processors provide speeds of up to 700MHz in mobile platforms

Not a bad idea really... so long as it works.

tahpot
on /.

info on crusoe chip:

Technology used
A whitepaper. The Technology Behind Crusoe(TM) Processors (pdf)
The processor's

from this info on transmeta's website the crusoe chip don't look too bad. it appears as if they've simply created a pretty good chip not based on x86 and then use the software as an interface between software and hardware.
they say:

The Code Morphing software can evolve separately from hardware. This means that upgrades to the software portion of the microprocessor can be rolled out independently of hardware chip revisions
and:
Crusoe processors provide speeds of up to 700MHz in mobile platforms

Not a bad idea really... so long as it works.

tahpot
on /.

If you visit their homepage you will not have to guess. Crusoe processors do indeed translate instructions in software, but they also profile and optimize the code during run-time. Specialized hardware helps them achieve this more efficiently than ordinary processors, like the PPC. This kit is tailored, hardware and software, for emulating x86.

As for performance, check out the same homepage for biased but probably to some extent correct benchmarks, both energywise and performancewise.

If you don't know, why post? Transmeta has a FAQ on their website. You might try reading it.

10. Are Crusoe processors available now? The first Crusoe processors, the TM3200 and TM5400 are available and shipping now.

From their own homepage: Founded in 1995

2000-1995 = 5 years (okay, so maybe 4.5 or something), so what are you talking about? Maybe it's that Pentium rounding problem again...

From Crusoe LongRun Technology page:

"In contrast, the TM5400 can adjust its power consumption without turning itself off - instead, it can adjust its clock frequency on the fly. It does so extremely quickly, and without requiring an operating system reboot or other slow and involved OS or BIOS operation. As a result, software can continuously monitor the demands on the processor and dynamically pick just the right clock speed (and hence power consumption) needed to run the application - no more and no less - so no power is wasted. Since the switching happens so quickly, it is not noticeable to the user."

It's not just that it can down into Energy Save Mode, but that it can dynamically chagnge the processor speed (and thus voltage) to meet the demands of the applications. It uses the minimum amount of power needed to get the job done. That and the fact that it is freakin small( ~1cm X ~1cm) and freakin fast (supposedly) that keeps me interested in the company.

When I said "STILL no generally available hardware," I meant no IA64. Alpha's existence is obvious. So is PowerPC, but in spite of the fact that PowerPC-64 has been sold and IA64 hasn't, some have the former taking a dive before the marketed might of the latter.

The greatest likelihood of success for the IA64 is in Intel boxes, which 99% of the public are unaware of. Bang for the buck, these machines will see limited service. When Intel & other vendors ever get around to trying to sell VLIW to the public and business they'll find it's like being the last girl scout out to sell cookies, "Sorry, we've already bought some."

The lightweight market will be taken by Crusoe, which is expected to be the market for VLIW for the next year, at least.

Vote Naked 2000

According to Tom's Hardware, the problem with using VLIW in complicated microprocessors is writing a fast compiler, which is really hard since instruction have to be executed concurrently (or if this is impossible you'll lose much performance). Read the article for further information.

This is also why the Crusoe is not as fast as a Pentium III or Athlon (the Crusoe uses VLIW too). However, in the Crusoe case it might not be a problem since Transmeta is targeting another market: PDA's, laptops and such, requiring low power usage and long battery live.

Obviously, the Itanium (stupid name!) is targeted at the server market, where power consumption does not really matter (except when running into cooling problems ofcourse!), so it had better be fast.

And until now it looks like it isn't! Little problem for Intel that has yet to be solved!

Are you sure there wasn't a java emulator running, in turn, on the x86 morphing layer? The morphing info page implies that the only morphing layer available is for x86. Of course, if it runs java code reasonably well while a morphed emulator is playing virtual machine, perhaps that means that when the java itself is morphed, it will run even faster. OTOH, this demonstration still does not take into account the possible overhead of switching and maintaining instruction sets.

Then again, it could have been running a java morphing layer too. If that's the case, then it looks like my reservations have been answered and I'm going to start getting excited about this.

2) I believe that the "code morphing" software is loaded at boot-time, but I could be wrong here. At the very least, all the little caches and everything that cache translated code (and provide a large part of the performance) would have to be flushed every time you switched from user-space to kernel-space in your example. This would (presumably) create a large performance hit.

Anyway, I think the web site has more details.

---

From the article:

Big Blue also provides legal cover from Intel. Because of extensive cross-license agreements, only IBM and ST Microelectronics can manufacture Intel-clone chips for other designers and provide these designers complete legal immunity, said sources close to Intel.

However, from the crusoe tech page on the transmeta web site

The hardware component is a very simple, high-performance, low-power VLIW (Very Long Instruction Word) engine with an instruction set that bears no resemblance to that of x86 processors. Instead, it is the surrounding software layer that gives programs the impression that they are running on x86 hardware. This innovative software layer is called the Code Morphing software because it dynamically "morphs" (that is, translates) x86 instructions into the hardware engine's native instruction set.

So is it really a clone, or an emulator? I'm guessing Crusoe is just playing it safe

From the article:

Big Blue also provides legal cover from Intel. Because of extensive cross-license agreements, only IBM and ST Microelectronics can manufacture Intel-clone chips for other designers and provide these designers complete legal immunity, said sources close to Intel.

However, from the crusoe tech page on the transmeta web site

The hardware component is a very simple, high-performance, low-power VLIW (Very Long Instruction Word) engine with an instruction set that bears no resemblance to that of x86 processors. Instead, it is the surrounding software layer that gives programs the impression that they are running on x86 hardware. This innovative software layer is called the Code Morphing software because it dynamically "morphs" (that is, translates) x86 instructions into the hardware engine's native instruction set.

So is it really a clone, or an emulator? I'm guessing Crusoe is just playing it safe

'nuff said

'nuff said

Go have a look at the SETI@Home Downloads page. Tell me there are no processors in there that might have vector-optimised maths units.

Go have a look at the Crusoe Technology page, or the VMWare site. Tell me that it's impossible to use hardware or software to emulate or translate from one instruction set to another. Besides, what's stopping the KrosnoConv "surplus military" stock being military-spec MIPS or SPARC clones? I didn't read the "about the company" bit, so I expected this was just an American company picking up after the excesses of the United States war machine.

Heck, check any Pentium III and tell me that it's impossible to execute another device's instruction set (8088) natively.

The only points that concerned me about the KrosnoConv boards were the Linux-in-Flash claim, and 32Mb of RAM per processor for less than $US200 (either very slow memory, or only 32Mbit perhaps, co-packaged RAM from the old 8086 days). There are projects out there to put Linux in a PC BIOS, or even an LS120. You can get CompactFlash cards, which behave like very small hard drives (either Flash memory pretending to be an IDE drives, or IBM microdrives really being IDE drives). But they're not cheap.

x86 is not the only architecture that SETI@Home supports. Why shouldn't someone produce an add-in card that uses your existing infrastructure? I would still be interested in getting a (cheap!) board full of heavy-maths processors to do hardware encryption for Virtual Private Networks, or even just a heavy duty key server.

I've already got the expensive bits like hard drives, network cards, monitors, cases and memory. I'd actually love to have a "parasitic" processor running its own OS, where I can download software to it, and have it process data that I store for it in my real RAM. Kinda like a multi-processor machine, where one or more processors are especially suited to encryption math.

Don't just spout "that's not how things work". Because with the introduction of technology like that used in the Crusoe, or even older technology like that used in VMWare, or any C64 emulator, you know this is how things work now.

At PC Expo this past June, Transmeta had web pads running Linux. And they used 802.11 for wireless networking. No cords. They expected something like 8 to 10 hours of running time (with a color LCD screen, no less).

However, the case had a industrial prototype feel to it - no logo or insignia of any sort visible anywhere. I asked the Transmeta booth-babe (who talked quite knowledgably about the pad and what was in it - she sounded like a real geek employee, not your typical "Hi I want to be an actress!" show floor model) about it, and she explained that the product hadn't been announced yet, but that the expectation was that pricing would be very low, because it would come with a ISP or service contract of some sort.

Need an example. Ok, check out Transmeta. The seemed to find a way to use a fraction of the power of an X86 chip, run the same apps, and perform better with a lower clock speed. PowerPC (used by IBM and Apple, duh!) chips do the same thing, only without the X86 compatibility.

Everyone, repeat after me: CISC is DEAD!

"Intel's Speedstep technology allows lower clock speed and lower wattage. Surely what Transmeta is offering is not radically different to that. It will all depend on pricing," he said.

According to Intel's PIII processor thermal design guide, the mobile PIII requires up to 19W at 733 MHz. An activity like playing a DVD, around 3W. With Crusoe, playing a DVD requires 1W!! 66% savings is not radically different?!?! Real world benchmarks suggest similar savings for various applications. Intel has a ways to go to render this difference "not significant", and to deliver all day battery life.

It seems like anyone can call themselves an "analyst" these days... Perhaps his only sources were Intel marketing materials?

> Methinks you have forgotten that
> Transmeta has a non-existent PR
> department.

They have a mailing list for product announcements. They should use it, that's all I ask for.

------------------

This results in massive energy saving abilities in case of the Crusoe, but am I the only one wondering if it can be used for overclocking too? After all, the control of the speed is software based, and therefore it should be possible to hack it, methinks...:-)

In fact, he could have simply walked to civilization, but that wouldn't have been such a good story. I find this to be a decent analogy to what's going on at transmeta.

I must ask:
Where the heck is the generic Evaluation board? I keep going over to the website, waiting for the Transmeta development stuff to become available, along with the generic $300 internet appliance development board, or the wearable computer development board, or even the somewhat risqué Gal Friday Personal Companion development board, which is being eagerly awaited by the guys over at Real Doll. I know a lot of technical geeks who would much rather spend several years hacking hardware and hydraulics and software to make a girlfriend who will put out after a romantic evening of watching deep space nine and listening to long winded rants about how stupid their boss is.
(I got married early, so I skipped the need to learn hydraulics...)

Seriously, though:
Is it just me, or has anybody else noticed that the draw for transmeta seems to be variable power draw more than any inherent improvements in speed?

I want my developer tools! (and yes, I understand that keeping the tool kits hidden may slow down the competition - but I want to be the first kid on my block with a transmeta pda... which I built myself.

I think I see slashdot at 1024x768 on that Quanta Pad.

What rock've you been hiding under?

Kean
http://home.san.rr.com/dlacey/

B1ood