Slashdot Mirror
Compaq Holds Off On Crusoe
scott1853 writes "Accroding to this article on ZDNet, based on this article at G2News, Compaq has decided to put off use of Crusoe in favor of Intel's P3. Unlike IBM, Compaq isn't claiming to put the processor on hold, they appear to have made their final decision not to use it. Could this be Intel flexing it's influential muscles, or is Transmeta not being competitive enough price-wise?"
This decision only confirms the wavering they've been having since at least June. They always wanted to see the goods before they committed themselves to shipping a product, and I guessed they aren't impressed. And unlike IBM, Compaq has never even demoed a transmeta prototype, so the decision is even easier for them to make without losing complete face.
-- Anne Marie
The DOJ and FTC have been breathing down Intel's neck for more than a decade now, but unlike Microsoft, Intel is bright enough not to stick its neck out far enough to get it chopped off. (Here's a hint to all you aspiring Bill Gateses at home: burn all your emails before the feds bang on your door.)
-- Anne Marie
Even worse is to hear "speeds up to 700 mHz". The 700 mhz Pentium chip, or AMD's chips don't have the "up to" tag that consumers, over the years, have equated to 'pie in the sky bullshit' speak. They are 700 mhz, period. Does this matter in real world applications? Maybe not, but it sure won't help it get sold.
There's already a big enough problem whereby Intel has, despite how evil you (or I) might think they are, built a really solid brand that consumers seem to trust (rightfully or not). But when you have to start telling them how the code morphing gizmos work it gets really confusing really fast, and the vast majority of people are going to stick with Intel..or at least get AMD processors where the mapping of performance is relatively straightforward (yes, I realize there are cache differences, differences in FPU performance, etc, but compared to the differences in Transmeta's chips vs Intel they are insignificant at this time).
The only way to get people over this confusion hump would be if the performance were FAR AND AWAY better or the power savings really made any kind of notable difference in the real world vs Intel's solutions (the certainly less geektastic power-stepping stuff they have in their new chips). I haven't seen a Transmeta powered system in real life that I could mess with, but from all reports of those who have, the chips offer neither of these benefits to any noticable degree. And they don't have the enormous price/power advantage that AMD started out with when they first started producing the Athlon.
We need to get these crusoe chips into some hardcore techies who know what they're doing. I don't think any of the synthetic benchmarks are going to help any because the more tiems you run them, the better it'll perform. Instead, maybe they should try real world benchs and watch as it suposedly gets faster (at least isn't that what the code morphing does?).
I've also seen some tests done by (i think it was HP) a few months ago with something called "Dynamo" and they showed that after the 3rd or 4th time through, the code was arithmetically faster.
Just a thought.
tap 2 blue, I counter that
Why doesn't someone just loop the frickin' benchmarks and draw the graph? It should be flat on AMD/Intel and show an increase for the Crusoe.
This seems awfully unfair to transmeta...to test their cpu against the same standard as the big guns, when their main feature in dynamicism.
Anyone who thinks dynamic code/data analysis is BS...read up on SGI/Nvidias video driver strategies.
It seems painfully obvious that a computing system should optimize itself to the task at hand.
Treatment, not tyranny. End the drug war and free our American POWs.
See my user info for links.
Intel is about to release new low power chips. If these things are anywhere in the same ballpark as the Crusoe in power use, it isn't worth the change.
With the new Intel chips Compaq doesn't have to reengineer much. Customers already know what a Pentium is. How many know what a Crusoe is? And, there are no chances of an odd incompatibility with the Pentium like there may be for th Crusoe. Just seems to make sense to me.
There's a story on Yahoo about the IPO, which is now at $48 and climbing, less than one hour from the IPO.
/. has rejected three stories submitted by me over the past week about the IPO, with probably about four story links per submission.
...
I picked up 75 shares at IPO price myself.
Amusingly,
Conflict of interest? Inquiring minds want to know
Anyone catch the little stab in the back by Compaq today? Et tu, Brute? Mr. Grove should be ashamed of twisting their arms over such a silly thing as a little bit of competition.
--- Will in Seattle - What are you doing to fight the War?
Transmeta will live or die by dockable WebPads. Low power consumption chips are most useful on devices with non-heavy graphics screens - this form allows one to use an HDTV screen for web browsing, but browse news articles while sitting at the couch or working in the kitchen. Here the difference between a 2-3 hour battery life (Intel/AMD products) and a 4-6 hour battery life (Tranmeta) will become critical.
This may also prove useful in PDAs, where battery life with reasonable non-hard-disk usage can have a major impact. I would love to have a PDA that plugs into my fridge webpad and synchs, plugs into my car display and activates voice tech modules, and can be stuck in a slim pack or a purse.
The rest is GIGO.
[note - I'm biased, since I have IPO shares in TMTA]
--- Will in Seattle - What are you doing to fight the War?
Imagine a Windows CE or Palm device with as much processing power as an Intel PII-300.
Imagine changing the AAA batteries in my Pilot every 20 minutes. Look up the power consumption of a Dragonball sometime: Crusoe is a power-sucking hog in comparision
Transmeta's between a rock and a hard place. They're trying to say "Look at the low, low 1 watt power consumption". But that doesn't help much in laptops, since the LCD and disk take a lot of the power, more than the CPU. At the same time, that 1 watt is something like an order of magnitude higher than a Dragonball[1]: installing a Crusoe in a Pilot just isn't going to work
Eric [1] A Dragonball uses 20 mA at 3.3V running flat out, most of the time much less than that.
"Seven Deadly Sins? I thought it was a to-do list!"
It does sound kind of similar. In fact, here's what seems to be a common theme:
Nobody wants to be first, but everybody wants to be second.
I haven't been watching this as closely (I'm happy with my PowerBook G3), but it did seem to go a bit like this: for a long time everybody was interested but nobody was actually announcing anything; then somebody announced a product, and everybody else rushed to follow suit; now everybody is backing off. I guess the optimistic view would be to suppose that everybody wants to do it but they are nervous about the risk of being the first to test the waters, and that once somebody finally does it, the rest will again follow.
David Gould
David Gould
main(i){putchar(340056100>>(i-1)*5&31|!!(i<6)<< 6)&&main(++i);}
"Accroding to this article on ZDNet,
Typos, the greek god of slippery fingers strikes again!
If you're really interested, I've read the size of the "translation cache" somewhere, probably on Transmeta's site. I can't think of it off the top of my head, though :(
I don't know a lot about the Crusoe, but I do know enough to say that it wouldn't take a whole heckuva lot of cache to make it worthwhile - "trouble spots" in code are what slows the computer down - just speeding those up will make a significant impact on performance. In a good way.
Dave
'Round the firewall,
Out the modem,
Through the router,
Down the wire,
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
Intel just finished (within 2 months) a 3 year investigation and came out clean as a whistle, its not Intel's fault that the instruction emulation is slower and that intel can create a lower power chip. Transmeta has nothing to offer
From recent reports on slashdot and other sites that occasionally cover chip news, Transmeta has been lacking from the beginning in many things. Primarily the chips speed. These "reports" haven't been indepth and haven't really showed the public in general what Transmeta is about as of yet. We can only speculate as of now. So far we know that the chips perform slower than its competitors (I don't understand why its such a big issue as from the tech specs and the way the chip was designed you HAVE to expect it to perform slower), they cost more money, YET they use less power and will make things like laptops and pda's battery lifetime increase.
Its a brand new company, brand new chip, alot of publicity yet they seem to peek from behind a curtain. Any major corporation is going to DELAY or NOT use a product until its generally proven. Transmeta hasn't been in the game that long at all; Hell! they are brand new, but as PDA's become more and more an everyday part of peoples lives I expect to see Transmeta take off (imho).
However I think that the best thing to do instead of speculating against major corporations which we don't have any major foundation or proven claims of which to base our speculation on. Is to wait, let the chips fall where they may (no pun intended) and let the cream rise to the top. Eventually if Transmeta has a good product which has been proven to work well and a major "put-in-a-box" computer corp backs them; then they'll be the first to reap the rewards. Everyone else will fall into place.
Lots of hype, promises of changing the market with a wonder product that isn't quite there yet. Unlike MS however they haven't got the muscle to enforce inferiour products on the rest of us.
I'm sure some people will state this is Intel pressure, but just possibly Compaq (who ship AMD chips) didn't see the cost/performance/power benifits that many on Slashdot have hyped up like a conference of MCSEs.
An Eye for an Eye will make the whole world blind - Gandhi
Companies I worked for generally require to jump to a radically new product from a new company a price/performance increase of five times,
sometimes ten. 2-3X for an established vendor. Because in this business if you just WAIT 6-12 months without chnaging vendors, you are going to see that kind of increase.
The Transmeta chip should not be targeted towards conventional laptops, but instead should be used to improve the performance of PDA's. Imagine a Windows CE or Palm device with as much processing power as an Intel PII-300.
I really don't want to see Transmeta die, it has a novel concept. It would really shine if they developed a native Java-VM for it that would translate the byte code just as it does 80x86 instructions.
BTW for all the anti-Intel people out there, don't forget unlike M$ they do a real competitor or two (AMD, for a while Cyrix, for a while IBM, TI on occasion, and now possibly Transmeta). Also don't forget when a problem is discovered in their product they have to fix it (unlike the 65,000 problems another company hasn't fixed). The actual product that reaches 99% of the public is a good and reliable. Yes its sad that now their marketing department runs the company, and that they haven't modified the core integer pipeline in 5 years, but the engineers working for them are good.
Its at $40, 13:30 EST. It was planned at $14,
and opened at $21. Fly high, penguins, fly.
No. It would be a "precedent" if (a) it were the first time and (b) it serves as an example or rule for subsequent. The Microsoft case is not, by any stretch of the imagination, the first example of "involving the government in market concerns", and plenty of examples already exist. The government in the US, as in most places, is involved in market concerns in all kinds of ways, often to the benefit of businesses (especially large corporations).
(When a corporate spokesdroid whines about "government interference" and speaks glowingly of the "free market", what they mean is "we only want the government to interfere when it benefits us". Heck, the very act of creating a corporation involves the government in market concerns. I suspect that in an real free market, corporations would be just so much tissue paper in hell.)
Performance isn't the only issue with Apple..Application compatibility plays a big role there, as well. (For Transmeta's sake, hopefully that's not true in their case).
Since no one mentioned it, the Transmeta (TMTA) IPO was today.
But does the Chip have performance issues? If so, I would suspect that Compaq would want to hold off until these issues are resolved. In the laptop wars, speed and battery life is everything...the Chip just may be unproven and Compaq may wait until someone else tests the waters first. Just a thought. Digger.
-- Don't you love a world where they give paranoid sociopaths guns, and tell them to shoot traitors and subversives?
CNN runs a story about Transmeta's IPO. In the story thet say how Transmeta is the first IPO since September to double it's IPO proce on it's first day of trade.
But I don't think anyone should believe that the folks at Transmeta are seeing this money. The folks at the company won't be able to cash out for some time. Wall Street has always been good at hyping stocks. The shares are allocated to friends who then turn around and sell them in the aftermarket for a quick buck. It's a very sleezy process and it's not surprising that IPO shares are used as grease in many political situations.
This kind of wild ride can often hurt a company. I'm very bullish on Red Hat and VA Linux, but all I hear about is their descent from the stratosphere. Somehow the companies have a cloud hanging over them. The money from the peak didn't end up in the company treasury. It didn't end up in the pockets of the programmers or the workers. It only went into the hands of the speculators. But the company and the programmers must pick up the pieces underneith this cloud of failure.
The Compaq decision is surprising and probably the result of some brassknuckle dealing between the companies. Transmeta probably pushed too hard in negoatiations and didn't have the benchmarks to justify their demands. I can't say I blame Compaq or IBM. They can always come back after the IPO when real chips are shipping and negoatiate a more stable price.
The saddest part about this event is how it shows how the nerds are being used by the Wall Street guys.
those nice young lawyers from the Department of Justice!
No boom today. Boom tomorrow. There's always a boom tomorrow. - Cmdr. Susan Ivanova
I wonder how this will affect Transmeta's IPO this morning. Typically the noon hours are light on trading, but if Compaq is pulling it, and IBM is delaying, will they reach the predicted $55/share that analysts are forecasting, or will they drop below the $18/$21 intro price?
Information is the catalyst for revolution
Apparently not. However, there are a couple of other markets that the Crusoe could target fairly easily, that may be more viable:
Wouldn't it be nice to have a low-power chip that could run standard PDA application and OS code "natively" but with far more heft? The Crusoe could fit this niche easily. Add appropriate support to a pared down Linux or *BSD, and you could even run, say, PalmOS and Windows CE as two processes on the same PDA. This has potential.
Code morphing is a wonderful idea. I've taken my share of compilers courses, so I know how sub-optimal most code is going to be. If you have OS support for profiling and incremental optimization on the fly without too much overhead, you could blow the competition out of the water. Right now people try to do that in silicon with complicated scheduling units, with limited success (I've had my share of VLSI courses, too).
In summary, I feel that Transmeta has a few useful markets still open to it.
Compaq sells Athlons by the truckload in their desktops. Just go to any Best Buy and look.
With IBM and Compaq both now not doing Crusoe notebooks, I think there is some problem that we don't know. Either price, or a real performance issue. Maybe the final power savings in a full size notebook just aren't worth the effort required to make a new model like this.
So, the Dept. of Justice is the catch-all solution for issues such as this? Jeezis...
Did I cheer when M$ got their (initial) slap from the govt? Kinda. You want to cheer for the big bully getting smacked around.. but from a bigger bully?
Don't you think involving the government in market concerns sets a rather bad precedent? I wanted to cheer like mad seeing Gates stomped.. but for what reason was I cheering? Is it any better that the government is trying to tell these companies how to conduct busisness? Look at the state of some of this contry's agencies & their desire to micromanage every single little detail. Case in point: the 'gas crisis' this summer combined with the DOE (Dept. of Energy) & EPA. See what happens to an industry sector when you let government have any sort of control, regulatory or otherwise?
It's said a bunch, but take it to heart. Vote your Intel dissention yourself; with your wallet. I'll happily investigate Crusoe when/if the need arises. I'm more than happy to avoid Intel (I'm avoiding them right now; this is on an hp pa8600). But leave it to be my choice, not some damned 'market equality' mandate from some government agency. That is hardly better than the situation as it exists.
-'fester
Once again, the artificial benchmarking system seems to doom yet another competitor. Due to the need to translate the x86 intruction set to their own instruction set (VLIW), TransMeta's Crusoe seems to be not the choice for Compaq.
Obviously, there is going to be some slowdown when testing the entire instruction set (or a reasonable subset thereof), but Crusoe should beat out the P3 on repetitive tasks.
Just my opinion here, we need new benchmarking standards.
Eric Gearman
--
Atomic batteries to power! Turbines to speed!
I've been thinking, as I hear about companies 'declining' to produce notebooks with Crusoe processors. Could it simply be that the numbers don't look good?
Companies go nuts trying to please consumers (or so the story goes). Surely they have pre-knowledge the the best seller is the processor with the biggest number on it. Mhz sell. And let's face it, Crusoe is not a big Mhz getter. And quite honestly I don't think any company wants to take a risk on something new if it isn't 'NEW AND IMPROVED'. In this case, that would mean something with far greated Mhz. It doesn't matter to comsumers that there is no way in hell your computer is using the Mhz that it already has. All that matters is that they get the biggest number possible. Could this be a valid theory?
Oh wait, never mind. Conspiracies are the only accepted valid theories nowadays. Right folks?
Slow moving marsupials and the women that love them
Slow moving marsupials and the women that love them
Next time, on Geraldo...
The /. crew editted my original title of "Compaq Dumps Crusoe"
Thanks for making me look like a dope, guys.
Wait at least 30 days and see where it is. Anything under a month after the debut is all hype and frenzy.
Don't you think it's going to go Berzerk first?(ok bad pun)
Could it be that well established companies don't want to take a risk and use a new product from a company nobody knows anything about? Could it be that without Linus Torvalds, nobody would even bother to look at Transmeta? Could it be that the power savings from the CPU are hardly noticable in the large scheme of things when you have LCD backlights and hard drives to power?
It's been nothing but bad news for Transmeta lately, two product cancellations in two weeks, poor performance numbers and just a general lack of direction in getting to market.
Be careful here. Getting your news via secondhand web surfing tends to over-expose you to the negative side of things. There's a dangerous anti-Transmeta spin that Slashdot seems to enjoy promoting. Are there other non-Transmeta project cancelations that are not reported? How common is it for a computer maker to announce a project and decided to kill it before shipping? The over analysis of Transmeta is leading to warped views.
Tuesday November 7, 1:16 pm Eastern TimeTransmeta stock soars in Wall Street debut
NEW YORK, Nov. 7 (Reuters) - The stock of Transmeta Corp. (NasdaqNM:TMTA - news) more than doubled on Tuesday in the highly anticipated Wall Street debut of the maker of low-power-consuming microprocesors that are making a big splash in the laptop computer market.
Transmeta stock was up $21-1/2 at $42-1/2 in heavy trading on Nasdaq, well above its initial public offering price of $21, which already had been raised to $16-$18 on Friday from $11-$13 a share. Transmeta raised $273 million by selling 13 million shares at the $21 per share offering price.
In what appears to have been a false start to trading, a Nasdaq member firm began trading the stock before its official opening, which had been set for 1 p.m. EST, a Nasdaq spokesman said. In those premature trades, which were subsequently cancelled, the stock appeared to have traded up $19 at $40.
The offering was a bright spot after some dreary months for the IPO market. In October, 35 companies held off from going public, the highest number in one month since CommScan LLC started tracking the figures in October 1990.
Investors and analysts have been intrigued by Santa Clara, Calif.-based Transmeta's Crusoe microprocessors that consume far less power than those made by Intel Corp. (NasdaqNM:INTC - news) and Advanced Micro Devices Inc. (NYSE:AMD - news). That means longer battery life, a crucial goal for laptop computer makers.
Work for Change & GET PAID!
The Crusoe chips in their current incarnations are simply x86 wannabees. They emphasise power thriftiness over performance. Therefore if Intel and AMD come out with equally power efficient chips, their advantage would be gone. Besides, as has been pointed out numerous times, the CPU is only one part in the power budget of a portable, and not even the largest one.
.NET architecture. Or develop a hybrid chip that can switch between x86 and PPC instruction sets dynamically, allowing you to run Windows and MacOS for example on the same machine at the same time. There are lots of issues that would make the latter pretty hard beside the CPU, but my point is to take advantage of your strengths to do something your competitors can't.
Transmeta should instead apply the code morphing technology in fields without ready competitors. Maybe something like a Java chip, or something fitting into Microsoft's
The problem with today's benchmarks are that they don't simulate real honest-to-goodness usage patterns.
;)
Before the Crusoe, they came close enough. Running the same program on a processor multiple times yielded the same results, so long as the program wasn't competing for resources. Those processors were in a limited way "stateless," which is to say that they don't remember anything from the last time the program ran.
Since benchmarkers didn't need to worry to much about it, the benchmarks generally run a program once, then run a different one, then a different one, etc., etc., etc..
In real life, however, a person is likely to open Word, then maybe open Internet Explorer, close Word, open Excel(which shares a lot of code with Word), close Excel, open Internet Explorer, and so on. That is *real life* usage, and Crusoe performs fairly well at it - certainly not as bad as traditional benchmarks would suggest.
When was the last time you played Quake III for two minutes? Yeah, I thought so. Fact is, when you're running on a Crusoe, those first two minutes will be much slower than the subsequent hour
So, as you can understand, Crusoe doesn't need "rigged" benchmarks - it just needs *real* benchmarks. The problem isn't that Crusoe will only perform well if the benchmarks are rigged, the problem is that Crusoe *won't* perform well unless the benchmarks put a computer under the same stresses and loads(or lack thereof) that a computer will go through on a daily basis.
Dave
'Round the firewall,
Out the modem,
Through the router,
Down the wire,
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
Athlon on the other hand is a superior design and AMD is digging deeply in Intel's market share at the high end.
It took some time of course but Athlon's martket share have been growing since the introduction.
The problem was that the manufacturers initially (after K6) did not trust AMD's ability to produce enough of their processors.
The problem with Transmeta, however, is that it does not have a clearly superiour technology (i.e. the saving in battery life are offset by the fact the the performance is relatively poor). I was surprised that so many manufacturers jumped on the bandwagon so fast. The defections are not surprising at all. If (a big if) Transmeta can keep up and deliver, they will be come back eventually.
It doesn't matter how advanced the technology is, how cool it sounds, nor who helped design it. If there isn't a market need for the product, it's going to be hard to find people to integrate it. Crusoe offers somewhat higher battery performance at the cost of performance. Does the market want lower performance? I think you'll find that pricing is based on processor performance and features like built in ethernet, firewire, etc. right now, and battery life is something laptop users have come to accept as being less than 4 hours. If people are going to buy a laptop with lower performance, they expect lower cost. If Transmeta wants to fit in to the market, they need to adjust their pricing and help integrators market battery life better. And saying a 600MHz processor has performance like a PIII at 500MHZ when it is really more comparable to a 350MHz PIII doesn't help your marketting campaign one bit. It isn't like the users aren't going to verify that.
Just like Linux is having/has had a hard time overcoming the Windows mind-share, so will Transmeta have a similar fate with regard to Intel/AMD.
You can't really expect such a big splash right away with relatively unproven technology (market wise). Just be glad someone (Sony) has given Transmeta/Crusoe a chance and let the performance of their product dictate their success. Look how long AMD took to really get a foothold in the CPU market. After all their K6's, it wasn't really until the Thunderbird Athlon that they earned their place in the market with a fantastic product.
Praying for the end of your wide-awake nightmare.
I don't know if you're neccessarily wrong.
From what it seems to me is that IBM is waiting for someone else to test the waters and then reap the benifits of a tested product. While Compaq is just giving it up for now (no decision is buisness is final).
I suspect they found something wrong. IBM's attitude recently fits very well into what the Crusoe offers. It is new, popular with the geeks, Linux-Able, cheap(?), etc... And if they turned it down...
Devil Ducky
Devil Ducky
MY peers would get out of jury duty.
It looks like the Crusoe processor is really going to take off and fly. Like an Iridium satalite.
134340: I am not a number. I am a free planet!
Umm so on the basis that it is slower on the overall benchmarks and that these benchmarks aren't designed to make Transmeta win... you want new benchmarks. The benchmarks referenced are application benchmarks, and what better test of a CPU than how quickly it runs an application.
Sounds so like MS "we're quicker if you let us rig the test" comparisons with Linux.
An Eye for an Eye will make the whole world blind - Gandhi
Well "Compaq Holds Off on Crusoe" makes it sound more like Crusoe is going to be delayed instead of being shelved altogether.
Now that Intel has publically flogged Rambus, Intel's partners are finding Intel a bit less unattractive. That's probably mainly it. I'm sure there are technical reasons, but those are always squishy enough to work over, especially if Transmeta learns to be more flexible on price. If it's cut or die, of course they'll cut, so I doubt that's the issue. It's gotta be strategic.
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Remember:
megalomaniacal software developer + hardware manufacturer = DISASTER!!!
Just as:
Bill Gates + Pocket PC = debt, it might turn out that:
Linus Torvalds + Transmeta = bust.
"Ancillary does not mean you get to rule the world." --U.S. Circuit Judge Harry Edwards, speaking to the FCC's lawyer
Now I may be mixing up something but my question is - is the Word code small/uniformic enough to get translated once, or it will be translated again and again as something bigger than translated instructions cache? In that case performance will be just as low as benchmarks (because every time something happens in Word's tummy it will be re-translated again)...
-----
Hyperom.com
Okay... they only talk about IBM's not putting out a laptop but its an interesting piece (and in the print version it had Linus's picture to go along with it :)
e ws/Media_and_Business/a-87490.asp
http://www.nydailynews.com/2000-11-07/News_and_Vi
This space for rent. All reasonable inquiries will be entertained at proprietors discretion.