Slashdot Mirror
Transmeta To Becomes Fabless Chip Supplier
Crazy Diamond writes "Transmeta has bought back its technology licenses from IBM and Toshiba in order to market its products on its own as a fabless chip supplier. This story comes from EETimes for more information."
I think Transmeta will make history.
..and I don't think it will be because of these chips. I think Transmeta will go down as one of the biggest flops that the Valley has ever seen. They'll probably do an IPO and their employees and founders will cash out and make millions. Once that is complete, they'll string it out for a few more months, drawing their fat paychecks, before the whole mess comes tumbling down. I'm guessing that they will release a product but it will be nowhere near as exciting as we've been anticipating. Performance will be mediocre and their competition will eat Transmeta for breakfast.
Maybe I'm wrong, maybe we'll all be using Crusoe-based laptops in two years. Somehow, though, I doubt it.
No, I think you have this backwards.I seem to recall the buyback being announced in articles serveral weeks ago.
8/17/00 Transmeta announces IPO . This article mentions the buy back so I don't know why this is new news. So does the VNUNET article that the spawned the slashdot discussion you linked to.
The IPO article predates Toshiba article. Perhaps Transmeta's board formally finalized the deal recently, but I suspect it is in their S1 filing.
So perhaps it sour grapes on Toshiba's part. IBM got cash and the fab business while Toshiba just got shares. They'll probably make money, but they could have made a lot more. This situation resembles when Apple cancelled the Mac cloning effort. That didn't exactly make them any new friends.
Quoth the poster:
Firstly, I think this does not bode well for Transmeta. The fact that Transmeta was able to buy back their licenses means that neither IBM nor Toshiba sees Transmeta as a good return on investment. If the technology was as impressive as the hype would have you believe, Transmeta would have only been able to get the licenses back from Toshiba and IBM by prying them out of their cold, dead hands. Sure, IBM and Toshiba will likely continue to supply devices using the Transmeta chip ... but considering they're some of the few firms that have actually seen the thing, it's not a good sign that they want to get rid of their licenses.
I would agree that the technology is no better or worse than when they first bought the licenses. However, I would argue that IBM and Toshiba bought into the hype somewhat too much. While Transmeta may be a great chip, IBM and Toshiba apparently don't see enough return in their investment to prevent them from selling their licenses. Ignoring this is choosing to close your eyes. Transmeta may still be the "killer-app" of chips. But I wouldn't hold your breath.
This isn't flamebait, it's just my opinion. I could easily be wrong. BUT, I think that on a fundamental level it indicates that the hype overtook the actual product.
On a different note, it's "Not much effect", not affect ... I can affect your decisions, but my actions will have an effect on you.
This is your life, and it's ending one minute at a time.
What does that means? It means that without a good business plan and the right people at the right place, even big names does not necesseraly means success.
You fundamentally misunderstand the goal of Transmeta. The product's goal is to compete against _Intel_ in mobile _PC_ products. In PDA products like Palm (i.e. where X86 compatibility is not needed) non-X86 products (particularly, Intel's StrongARM) are much better because they use less power, are faster, and are cheaper. Transmeta specifically said that they are NOT competing in this market. Transmeta will only compete where X86 compatibility is needed in lower power devices, and there has yet to be a market for that (aside from laptops, and Transmeta has yet to demonstrate superiority to Intel in that market).
Transmeta bought back the rights for 600,000 shares of stock not for Cash. If IBM and Toshiba think that Transmeta is worthless wouldn't the Stock be worthless too? They the should have held out for Cash.
Transmeta now appears to have a similar business model to ARM. They both design processors for the low-power, portable markets, but there are differences: Transmeta's processors are larger and faster, but use more power. ARM's processors are used in devices like mobile phones and organisers where low power consumption is vital, and pure speed is not important.
ARM have been doing extremely well so far with this business model...
Second I'm not a 'fanboy' - rather I'm somebody who tracks this market as part of my job. Torvaldis or not Transmeta has some interesting technology. Commercially successful is still TBD but it is interesting stuff that could have applications beyond their original market.
Third if you recall (or are aware of) Intel spent years dealing with the results of it's licensing production of it's 80386 chips. They ended up competing with their own licensees for prices and allocations. It was these licenses that allowed many of the clones to convert (against Intel's lawyers and many court battles) to 'licensees' and thus buy legitimacy when they were manufactured by license holders or when their designers bought or were bought by license holders.
If I was a hot new thing on the block possibly considering another round of investment I'd certainly want the license variable off of the sheets. It's not unlikely that the licenses contained a standard buy-back clause which Transmeta exercised. Furthermore it's telling that the deal was done with Transmeta stock and not their cash.
Whether Transmeta's chips are 'all-that' we don't know yet. Certainly they have promise for competing in the mobile-marketplace with Intel. Whether the big boys are just backing them to keep pressure on Intel in the market or they're seen as a truly viable isn't clear. It's possible that they're both a pressure tactic and a potential breakthrough - a two-fer in the eyes of IBM, Toshiba & others. Either way it's too soon to bury them.
Right now Transmeta probably has another year in which to make some success with their original plan. If they don't succeed look for them to retarget to some less obvious/less immediately profitable market. I'd bet they've already got a team working on developing some legacy-emulation for aging big iron (something that would certainly pique IBM's interest.)
Any which way - they've got some interesting Intellectual Property and make for a good story.
I don't read ACs: If a post isn't worth so much as a nom de plume to its author then I wont bother either.
Perhaps. Or it could mean that Transmeta included a buy-back opion in the original agreement, (perhaps with a fixed buyback price), and Transmeta has elected to use the option.
It is intirely possible (and likely) that all three companies are exlicitly not allowed to make any sort of public statement about the license agreements.
You are welcome to speculate, but I find fault with your logic, and I'd like to point out that your conclusion is not supported.
is going to do some very cool things. The simple fact is in the market that Transmeta is trying to get into people are *really* excited over a processer with less than a 386 in it (think palm). I was telling my boss today about some of the things this chip would allow. He was ready to order. It is not supposed to compete with the P-4 or whatever. It is competing with the Palm and it can win this. Why? Because it is the only solution really portable, long battery life, and can give me a real OS. This is important stuff.
Cypherpunks: Civil Liberty Through Complex Mathematics. Those who live by the sword die by the arrow.
Founder's Camp
Founder's Camp
News for non-Nerds. Stuff that matters.
heard in the Transmeta boardroom:
"today is a good day to die!"
:)Fudboy
:)Fudboy
I guess I'm only a Fudboy, looking for that real Transmeta
"But its competitors are two massively-resourced chip giants - Intel and AMD."
:-) :-) :-)
... and this is only a good sign, them putting AMD on equal grounds
-----
From the article:
Microprocessor startup Transmeta Inc. (Santa Clara, Calif.) has quietly changed its business plan. The company has bought back its two main technology-licensing agreements from IBM Corp. and Toshiba Corp., and now plans to market products on its own as a fabless chip supplier
First a major investor,Toshiba,claims that their chips are more hype than substance then they buy back their technology licenses from IBM and Toshiba?
Looks to me like Transmeta jumped the gun on their speculative announcements and have realized they cannot put their money where their mouth is. Unfortunately since this is Slashdot and Linus works there, its likely that this post will get moderated down and even though people here slaughter Intel when it comes to bad news, this will be glossed over. *sigh*
(-1 Troll)
After seeing Toshiba laying down for Rambus and the floppy drive lawsuit, I'd be wary of having my company's fortune rely on their guts. Much the same with IBM. It might not be a bad idea for Transmeta to strike out on their own. The company I work for has done both sides of the silicon foundry deals, and while it has major pains, the old business model had a lower upside potential. To build their own fab would take a bunch of capital, and there is a bit of stray capacity around the world.
I'm not downplaying the debt, but in a way, selling the old licences and buying them back was an interesting way to get early capital. Now, at least, they get to live or die on their own marketing/design skills. I wish them luck.
I don't think it bodes well for Transmeta, either. I don't see why you wouldn't want big partners like IBM and Toshiba if you are Transmeta. Unless, neither IBM nor Toshiba are not interested in bringing products forward using the Transmeta technology, which we have some reason to believe because of the recent Toshiba criticism of the Transmeta chip set.
Now, this is a leap. We don't know that there wasn't a favorable buy-back clause in the licensing agreements that Transmeta signed with those guys.
-Jordan Henderson
One of my favorite examples when this doesn't happen is Rambus RDRAM.. if it wasn't for Intel trying to shove the technology down our throats RDRAM probably won't have even made it off the drawing board. All this time Intel could have been pushing for better technology (ie. DDR-SDRAM) instead of covering their asses (not to mention loosing God knows how much money with their faulty RDRAM motherboards).
Toshiba decided to complain about how Crusoe was not "all 'dat" so TM buys the rights back from them... now if we see a Crusoe inside of a Toshiba notebook we'll know that Crusoe is not all hype.
I've read as well that some of you think that Crusoe is all hype and is total vaporware... i would really like to think that the talented engineers (yes there are many other talanted engineers working for TM besides Linus) that have been working on Crusoe for 5 years + actually have technology that works. Many of them left jobs at other large companies to work at TM because they were excited about what they had the potential to produce. I'm sure many OSS developers here know what it is like to be excited about their product... you would not produce shitty stuff and be excited about it. If even the other engineers were not excited about their work, i'm sure that Linus would not work for TM if he didn't think he had something cool to show. Of course he could just be in it for the money, or to pay the bills, or just to stay in the US. However, Linus doesn't exactely strike me as a money grubbing freeloader if you know what i mean. I'm sure there are more than a few companies that would pay him a rediculous amount of money for his skillz and his name.
Anyway, i'm willing to give TM the benefit of the doubt and not come to conclusions until there are more than a few reviews saying that Crusoe really does bite. I believe in "innocent until proven guilty" and i also believe in "good tech until proven shitty".
Ceres
Or simply that the original license agreement contained a buy back clause, saying something like:
Toshiba and IBM might not have had a choice...
"The invisible and the non-existent look very much alike." -- Delos B. McKown
guy one: "how are we gonna get funding and access to test and design facilities?"
guy two: "Let's make our technology look really sweet and we'll license it out to some major corporations, including one or two with chip design and testing facilities. Then, after we get what we need out of them, we'll just try to buy back our rights with some stock from when we IPO and a relatively small cash payment compared to what we'll make on the IPO."
guy one: "Sweet, let's get drunk."
Transmeta's has had everything look very well thought out up until now, I fail to see how they would have just now "changed their minds". They're doing alot of very risky things, but it seems like it's been well planned since day one.
Need Free Juniper/NetScreen Support? JuniperForum
sig:
sig:
See the "..for smart people" banners Wired runs here? Look elsewhere guys.
Being fabless certainly didn't hurt Nvidia or 3dfx (back when 3dfx still just made chips). I know that they are in the video market, but those companies made good and highly competitive products. If the final chip is as half as good as transmeta is leading on (at this point believing their hype is an exercise in futility) they have a shot. But going this way is risky as hell. If they are the best, then there will be plenty of money to buy fabrication and all will be well. If not, they are going to be gone so quick that it will make your head spin. That being said, yes right now its alot of "Linus works there, it must be good" hype going on. Call me when they have a product.
"My head hurts, My feet stink, and I dont love Jesus." -Jimmy Buffett
Frankly this seems more like "News for Finance Nerds" then anything directly technology related.
I don't read ACs: If a post isn't worth so much as a nom de plume to its author then I wont bother either.
-
Transmeta thinks its technology is so great that it wants exclusive rights to it.
-
Transmeta's customers think its technology sucks and weren't interested in manufacturing it.
Personally, I suspect #2. Especially since Transmeta, which has been big on PR, didn't announce this; they just put it in a mandatory filing with the SEC.I have the feeling that Transmeta's technology is going to go down as one of those things that can be done, but isn't worth the trouble. The AMD Athlon already converts instructions to a RISC-like uniform width at cache load time, so the most touted Transmeta feature is already shipping in volume from someone else.
Transmeta's biggest assets are Linus and Paul. Just on the name value alone Transmeta as a good chance of making it work.
Well, if you read further IBM will remain as the foundry. Which means IBM will still provide the fabrication facility. While Transmeta won't have its own fab, it still has fabrication capacity lined up