Crusoe and Intel does compete in the laptop arena, assuming Transmeta gets one or more of the big laptop makers to use their chips.
The problem they face have been discussed for a few days in news:comp.arch, basically that the cpu power usage isn't really a dominating part of the total for a laptop: Disk, DVD, LCD, 3D etc are all power hogs as well, so you'll need to improve those other parts as well.
What will be really interesting, is to do some micro-benchmarks to figure out how much of the speed they get that is directly related to the Code Morpher rewriting/optimizing the x86 code, and how much is handled directly by the initial interpreter/fast compiler.
Terje
-- "almost all programming can be viewed as an exercise in caching"
If you could have the equivalent to a 500 MHz Pentium II in your laptop, but have it run for 5x as long, why wouldn't you? The Crusoe is a viable alternative to anything that is using Intel and could possibly benefit from lower power usage,
-----------
"You can't shake the Devil's hand and say you're only kidding."
While we were listening to the Transmeta press conference here at the office the other day, a collegue of mine made a comment about this. Intel has the ability to cut prices and cut below transmeta prices. Sadly enough this will determine wether or not a new chip company can succeed. Can ANY new company succeed with this type of threat?
-- "Fighting the underpants gnomes since 1998!"
"Bruce Schneier knows the state of schroedinger's cat"
Intel has the ability to cut prices and cut below transmeta prices. Sadly enough this will determine wether or not a new chip company can succeed. Can ANY new company succeed with this type of threat?
Perhaps they can cut their prices easily, but it will be harder for them to make the chips run cooler. These price cuts probably have more to do with AMD - K6's are killing intel in the mobile market.
Note that Intel is now under attack in *all* their major market segments, so this limits their ability to undercut in one market while making up for it with extra-high prices in another market.
This won't be the case for Transmeta's Crusoe, if Intel can't field a mobile chip that will last 4-9 hours on a single battery charge, Transmeta will have their Lunch and Martini too. The price won't matter.
This is a clear case of a distruptive technology changing the marketplace in laptops and mobile devices. Intel's lost this market if Transmeta can ramp up manufacturing. IBM's going to make them, and they didn't say exactly who they were dealing with in Taiwan, did they? (Might be more fabs overseas cranking theses baby's out.)
I was going to get a Sony VAIO a while back, but I was waiting to see what Transmeta was going to do. Now I'm waiting for Crusoe based products instead of Intel based products because I really want that battery life. _______ computers://use.urls.People use Networds.
Yeah. Essentially if Intel tries to undercut Transmeta, all Transmeta has to do is market their chip to an even smaller application. The Transmeta chip really does have a competetive advantage: it's much cooler. All Transmeta has to do is find the market that Intel can't support in the first place.
Incidently, just because a large behemoth of a company _can_ do something monopolistic doesn't mean it will do it well. Look at the bumbling antics of microsoft against perceived future threats. And Transmeta is a very real threat against Intel in the present!
A few times each year, Intel and AMD announce price cuts. This is typical marketing for both companies. They announce price cuts on all but their highest speed/most expensive processors. This is not news, just the typical PR we see every few months.
This happened before Transmeta's release.
by
nevets
·
· Score: 2
In the article it states that the prices took effect on January 16th. Unless they were able to break the security and secrecy of Transmeta, I don't think they knew. I'm sure some of you will argue that this was Intel getting ready for Transmeta, but I don't think they are too worried yet. I'm sure that AMD gives them a bigger scare.
Simply put, the Crusoes are designed for Laptops and below. This means they don't have the best IO architecture in the world, especially in the x86 world. They can handle laptop IO, and the incorporated NorthBridge supports PCI, SDRAM, DDRSDRAM etc, but no AGP support, no I2O etc. It might be possible to add an external chip to support these though.
Transmeta have created the high-end of Code Morphing processors. Having proved they work quite well, they can now concentrate on putting more of the software in silicon - e.g., better IO handling, AGP, More functional units etc, and in a year or so you can expect fast desktop Transmeta CPUs (not Crusoes though!). The Transmeta is good at running a limited number of programs at once - any more and the translation cache will keep on filling up and you will be accessing main memory more and more, which is bad - so fine for PDAs and Laptops, but bad for desktops where having 50-100 different processes running is common.
Expect a desktop Transmeta to include at least 1Mb of on-board cache as well as even more powerful silicon and software.
~~
Cheap Cheap Cheap
by
Anonymous Coward
·
· Score: 2
Well...that is a good start. I keep waiting for laptops to start whoreing themselves out like desktops and this is a good start. Why can't I get a decent laptop for under a grand ????? It doesn't have to be a PIII screamer. I would be happy with a PII 300 - 400 with 64 megs of ram, 4 gig hd, cd..etc..
You can get an average PC for under $800...why not an average laptop for under a grand.
The thing that cranks me is that they stop selling the older chips when the new faster ones come out. Why not sell a laptop, cheap, with an older slower chip... I know it's all about making money.. but it still hacks me off.
You can get an average PC for under $800...why not an average laptop for under a grand.
Are you that dense?
TFT screens are EXPEN$IVE
tiny HDDs are EXPEN$IVE
cases are costly to create (big tooling charges in excess of $10k) so you have to sell a whole lot of laptops to bring this down
Heat considerations affect the design, making it costlier
custom motherboards
Li-ion batteries aren't cheap
In short, the sheer volume of standard PC components outnumbers the volume of laptop components so they are far cheaper to buy in consumer quantities. Until everyone buys laptops and not PCs, this won't change.
Perhaps Transmeta is an issue, but believe me, these price cuts have been in the works for a while and were planned to happen when Intel introduced it's SpeedStep technology. Intel always cuts prices on old stuff around the time it intros new stuff.
It does look like Transmeta found out the SpeedStep announcement date and scheduled their event the day before to preempt coverage of SpeedStep. That worked quite well, IMHO.
--LP
Something doesn't add up here...
by
theMAGE
·
· Score: 2
From the article:
In an unusually deep round of cuts, the chip vendor slashed the price of its 500-MHz Mobile Pentium III processor by 54 percent, from $530 to $245, according to information on Intel's Web site. Prices of 450-MHz and 400-MHz versions of the chip also fell sharply, by as much as 46 percent.
And later on...
The price cuts could translate into savings of as much as $100 on the price of a typical Pentium III notebook PC, said Dean McCarron, principal analyst at Mercury Research, in Scottsdale, Ariz. Celeron notebook prices may also come down slightly, though probably not by as much.
Can anybody explain me how by cutting the price on a component by $200 the price of the whole goes down by $100?
Hmmmm....
Re:Something doesn't add up here...
by
Anonymous Coward
·
· Score: 2
Can anybody explain me how by cutting the price on a component by $200 the price of the whole goes down by $100?
You're onto something here - they must only use half a processor per laptop! We're being ripped off!
I think you are thinking in 3 year old categories. Namely:
Typical desktop disk consumption at the moment is under 20 watt (for laptops there are some as low as 2.5 watt). For example desktop 3.5 inch IBM 4.3GB SCSI-II manufactured last year is 330 mA from 5V and 200 mA from 12V. This is 1.65 + 2.4 W = 4.05 W MAX.
CDROMs and DVDs are still a power hog but they hardly go over 20-30 watt. For a 40 speed SCSI-II Toshiba it is (unsure here;-) under 20.
LCD is also somewhere there (20-30 at most) as well.
Video is under 5 watt.
The biggest hogs at the moment are CPU. If intel it can go above 100watt + 3W fan and sound which can also be over 50 watt in some configurations. Thus, a transmeta chip will drop your average power consumption on a computer that does not have its speakers blasted 100% by 50%.
This will result in either weight decrease by as much as 30-40% or battery life increase by as much as 50%.
I would bet on the weight decrease. Because less batteries means not only less weight. It means as much as 20-50 dollars off the price (NiMH are bloody expensive).
Overall this price decrease barely compensates for the manufacturers price decrease due to less battery expenses. If intel wants to beat transmeta fairly they have to drop by further 20-30$. This drop will be enough to start FUD wars though because very few people remember to calculate the weight and the battery pricing in laptop comparisons.
-- Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
transmeta chips are alwredy in production at IBM fabs. This is scaring intel more then AMD. Amd has limited production capabilities and have a history of fabrication yield problems, that they are only now STARTING to come out from. IBM basicly has almost unlimited fabrication resources relative to AMD.
There are benchmarsk on Transmeta site. They are typical task benchmarks, not plain CPU blast and according to them the smaller of two chips is somewhere around 0.8 - 0.9 Pentium III 500 and the bigger one (due midyear) is equivalent or better.
Transmeta still has (technologically) the big advantage in mobile computing because of the extremely low power draw of its chips. No one would use a celeron in a web pad or palm-type device when they can use a Crusoe that draws a fraction of the power. (Yes, I know Transmeta claims not to be going after palmtops and cell phones right now, but who knows they may just have to take what they can get.)
The translation cache, along with the Code Morphing code, resides in a separate memory space that is inaccessible to x86 code.
It simply doesn't make sense to use the hard disk for the translation cache. Consider:
RAM access times are measured in nanoseconds and hard disks measured in milliseconds. It's magnitudes faster to simply re-morph code on the fly than use a disk based translation cache.
The Crusoe is designed to be embeddable in systems which don't even have hard disks (i.e. set-top boxes, Webpads, etc).
To maintain a hard-disk based code cache, the Crusoe would need to work with SCSI and IDE drives and understand partition tables and filesystems. Why build this complexity in at the CPU level, especially when you're aiming for a simple power-saving design?
Transmeta announced the unveiling date back at fall comdex( november). Linus announced it during his keynote.
One thing I have learned after 12+years in this business is that Intel never drops prices just to be friendly. They (intel) only drop prices when shifting products or when someone actually comes up with a competing product. That meant many fewer price reductions in years past. Many more cuts in '99. With AMD finally getting a (kickass) product launched with volume production, they scared the crap out of intel. Then consider the i820 fiasco and you can see intel executives peeing their expensive suits at their predicament. Now they are being seriously threatened by Transmeta on the low end where most of the significant processor growth is expected for the next 5-10 years. I think the SpeedStep introduction was a factor, but I also think the price drop was a bit steep to be just an internal response and making room in their product line.
--
Kindness is the language which the deaf can hear and the blind can see. - Mark Twain
That's because despite all the yipping and wishful thinking by the Linus Torvalds fanboys here, Intel's move has absoutely nothing whatsoever to do with Transmeta. Ever heard of AMD, which has been putting a hurting on Intel lately?
Cheers, ZicoKnows@hotmail.com
Re:More interesting questions
by
David+Greene
·
· Score: 2
A number of arguments against running on the bare Crusoe metal have been given in the past, so I won't bore anyone with the details. Suffice to say that running like this may very well decrease performance due to the nature of the translation layer. Furthermore, the flexibility code morphing gives Transmeta greatly outweighs, I think, any small performance improvement that might result from eliminating it.
As far as AMD's 64-bit chip is concerned, it's not clear to me which form it will take. But they may be too late. Unless it is 100% IA64 compatible, they're going to have a hard time convincing Microsoft, et. al. to port their software and operating systems. And yes, of course Linux will run on it. But will RedHat make a distribution?
I think Intel is sweating over two things: Transmeta's low power and Sony's Emotion Engine. I have nothing with which to back this up save a few rumors and scuttlebutt, however.
--
--
This Happens Every Quarter...
by
John+Murdoch
·
· Score: 2
The InfoWorld article refers to Intel's quarterly price cuts. You will note that computer manufacturers are also announcing lower prices, reflecting the drop in prices. This is nothing new--it happens in the second week of every quarter.
But since the question has been raised, how concerned is Intel about Transmeta? Probably not that much. Remember that Intel owns the microprocessor business--not just the chips, but also the technology required to make the chips. They have extensive R&D projects with companies like Applied Materials (and a major client of mine) that affect all sorts of things. And remember that Intel is extremely aggressive about protecting its patent rights. Much of Transmeta's press conference palavar about not worrying about Intel and patent issues is so much PR--if IBM is planning on using Intel-developed technology to produce the Transmeta chips you will see mention of a "cross-licensing agreement" in the trade press. (Remember that IBM will actually manufacture the chips for Transmeta.)
There is another factor to consider: producing cutting edge chips in volume requires massive capital investment. If Transmeta can't sign up a lot of business they may not be able to produce enough chips to bring the unit costs down low enough....
At the moment Intel is continuing to cruise right along, printing money. AMD? An annoyance. Transmeta? A set of spects and a press conference. Down the road Transmeta might be reason for Indel to be concerned--but a threat? The moo-ha-ha department? No. We're way too early for that.
Crusoe and Intel does compete in the laptop arena, assuming Transmeta gets one or more of the big laptop makers to use their chips.
The problem they face have been discussed for a few days in news:comp.arch, basically that the cpu power usage isn't really a dominating part of the total for a laptop: Disk, DVD, LCD, 3D etc are all power hogs as well, so you'll need to improve those other parts as well.
What will be really interesting, is to do some micro-benchmarks to figure out how much of the speed they get that is directly related to the Code Morpher rewriting/optimizing the x86 code, and how much is handled directly by the initial interpreter/fast compiler.
Terje
"almost all programming can be viewed as an exercise in caching"
-----------
"You can't shake the Devil's hand and say you're only kidding."
While we were listening to the Transmeta press conference here at the office the other day, a collegue of mine made a comment about this. Intel has the ability to cut prices and cut below transmeta prices. Sadly enough this will determine wether or not a new chip company can succeed. Can ANY new company succeed with this type of threat?
"Fighting the underpants gnomes since 1998!" "Bruce Schneier knows the state of schroedinger's cat"
A few times each year, Intel and AMD announce price cuts. This is typical marketing for both companies. They announce price cuts on all but their highest speed/most expensive processors. This is not news, just the typical PR we see every few months.
In the article it states that the prices took effect on January 16th. Unless they were able to break the security and secrecy of Transmeta, I don't think they knew. I'm sure some of you will argue that this was Intel getting ready for Transmeta, but I don't think they are too worried yet. I'm sure that AMD gives them a bigger scare.
Steven Rostedt
Steven Rostedt
-- Nevermind
Simply put, the Crusoes are designed for Laptops and below. This means they don't have the best IO architecture in the world, especially in the x86 world. They can handle laptop IO, and the incorporated NorthBridge supports PCI, SDRAM, DDRSDRAM etc, but no AGP support, no I2O etc. It might be possible to add an external chip to support these though.
Transmeta have created the high-end of Code Morphing processors. Having proved they work quite well, they can now concentrate on putting more of the software in silicon - e.g., better IO handling, AGP, More functional units etc, and in a year or so you can expect fast desktop Transmeta CPUs (not Crusoes though!). The Transmeta is good at running a limited number of programs at once - any more and the translation cache will keep on filling up and you will be accessing main memory more and more, which is bad - so fine for PDAs and Laptops, but bad for desktops where having 50-100 different processes running is common.
Expect a desktop Transmeta to include at least 1Mb of on-board cache as well as even more powerful silicon and software.
~~
Well...that is a good start. I keep waiting for laptops to start whoreing themselves out like desktops and this is a good start. Why can't I get a decent laptop for under a grand ????? It doesn't have to be a PIII screamer. I would be happy with a PII 300 - 400 with 64 megs of ram, 4 gig hd, cd..etc..
You can get an average PC for under $800...why not an average laptop for under a grand.
The thing that cranks me is that they stop selling the older chips when the new faster ones come out. Why not sell a laptop, cheap, with an older slower chip... I know it's all about making money.. but it still hacks me off.
Perhaps Transmeta is an issue, but believe me, these price cuts have been in the works for a while and were planned to happen when Intel introduced it's SpeedStep technology. Intel always cuts prices on old stuff around the time it intros new stuff.
It does look like Transmeta found out the SpeedStep announcement date and scheduled their event the day before to preempt coverage of SpeedStep. That worked quite well, IMHO.
--LP
From the article:
In an unusually deep round of cuts, the chip vendor slashed the price of its 500-MHz Mobile Pentium III processor by 54 percent, from $530 to $245, according to information on Intel's Web site. Prices of 450-MHz and 400-MHz versions of the chip also fell sharply, by as much as 46 percent.
And later on...
The price cuts could translate into savings of as much as $100 on the price of a typical Pentium III notebook PC, said Dean McCarron, principal analyst at Mercury Research, in Scottsdale, Ariz. Celeron notebook prices may also come down slightly, though probably not by as much.
Can anybody explain me how by cutting the price on a component by $200 the price of the whole goes down by $100?
Hmmmm....
I think you are thinking in 3 year old categories. Namely:
Typical desktop disk consumption at the moment is under 20 watt (for laptops there are some as low as 2.5 watt). For example desktop 3.5 inch IBM 4.3GB SCSI-II manufactured last year is 330 mA from 5V and 200 mA from 12V. This is 1.65 + 2.4 W = 4.05 W MAX.
CDROMs and DVDs are still a power hog but they hardly go over 20-30 watt. For a 40 speed SCSI-II Toshiba it is (unsure here ;-) under 20.
LCD is also somewhere there (20-30 at most) as well.
Video is under 5 watt.
The biggest hogs at the moment are CPU. If intel it can go above 100watt + 3W fan and sound which can also be over 50 watt in some configurations. Thus, a transmeta chip will drop your average power consumption on a computer that does not have its speakers blasted 100% by 50%.
This will result in either weight decrease by as much as 30-40% or battery life increase by as much as 50%.
I would bet on the weight decrease. Because less batteries means not only less weight. It means as much as 20-50 dollars off the price (NiMH are bloody expensive).
Overall this price decrease barely compensates for the manufacturers price decrease due to less battery expenses. If intel wants to beat transmeta fairly they have to drop by further 20-30$. This drop will be enough to start FUD wars though because very few people remember to calculate the weight and the battery pricing in laptop comparisons.
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
transmeta chips are alwredy in production at IBM fabs. This is scaring intel more then AMD. Amd has limited production capabilities and have a history of fabrication yield problems, that they are only now STARTING to come out from. IBM basicly has almost unlimited fabrication resources relative to AMD.
--------------------------------
There are benchmarsk on Transmeta site. They are typical task benchmarks, not plain CPU blast and according to them the smaller of two chips is somewhere around 0.8 - 0.9 Pentium III 500 and the bigger one (due midyear) is equivalent or better.
Have a look at them
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
Transmeta still has (technologically) the big advantage in mobile computing because of the extremely low power draw of its chips. No one would use a celeron in a web pad or palm-type device when they can use a Crusoe that draws a fraction of the power. (Yes, I know Transmeta claims not to be going after palmtops and cell phones right now, but who knows they may just have to take what they can get.)
i do
is Jesus your personal savior? click here
The Crusoe processor doesn't use the hard disk for its translated code cache.
From The Technology Behind the Cruesoe Processors: It simply doesn't make sense to use the hard disk for the translation cache. Consider:
Transmeta announced the unveiling date back at fall comdex( november). Linus announced it during his keynote.
One thing I have learned after 12+years in this business is that Intel never drops prices just to be friendly. They (intel) only drop prices when shifting products or when someone actually comes up with a competing product. That meant many fewer price reductions in years past. Many more cuts in '99. With AMD finally getting a (kickass) product launched with volume production, they scared the crap out of intel. Then consider the i820 fiasco and you can see intel executives peeing their expensive suits at their predicament. Now they are being seriously threatened by Transmeta on the low end where most of the significant processor growth is expected for the next 5-10 years. I think the SpeedStep introduction was a factor, but I also think the price drop was a bit steep to be just an internal response and making room in their product line.
Kindness is the language which the deaf can hear and the blind can see. - Mark Twain
That's because despite all the yipping and wishful thinking by the Linus Torvalds fanboys here, Intel's move has absoutely nothing whatsoever to do with Transmeta. Ever heard of AMD, which has been putting a hurting on Intel lately?
Cheers,
ZicoKnows@hotmail.com
As far as AMD's 64-bit chip is concerned, it's not clear to me which form it will take. But they may be too late. Unless it is 100% IA64 compatible, they're going to have a hard time convincing Microsoft, et. al. to port their software and operating systems. And yes, of course Linux will run on it. But will RedHat make a distribution?
I think Intel is sweating over two things: Transmeta's low power and Sony's Emotion Engine. I have nothing with which to back this up save a few rumors and scuttlebutt, however.
--
The InfoWorld article refers to Intel's quarterly price cuts. You will note that computer manufacturers are also announcing lower prices, reflecting the drop in prices. This is nothing new--it happens in the second week of every quarter.
But since the question has been raised, how concerned is Intel about Transmeta? Probably not that much. Remember that Intel owns the microprocessor business--not just the chips, but also the technology required to make the chips. They have extensive R&D projects with companies like Applied Materials (and a major client of mine) that affect all sorts of things. And remember that Intel is extremely aggressive about protecting its patent rights. Much of Transmeta's press conference palavar about not worrying about Intel and patent issues is so much PR--if IBM is planning on using Intel-developed technology to produce the Transmeta chips you will see mention of a "cross-licensing agreement" in the trade press. (Remember that IBM will actually manufacture the chips for Transmeta.)
There is another factor to consider: producing cutting edge chips in volume requires massive capital investment. If Transmeta can't sign up a lot of business they may not be able to produce enough chips to bring the unit costs down low enough....
At the moment Intel is continuing to cruise right along, printing money. AMD? An annoyance. Transmeta? A set of spects and a press conference. Down the road Transmeta might be reason for Indel to be concerned--but a threat? The moo-ha-ha department? No. We're way too early for that.