backtop

Print E-mail del.icio.us 29 comment(s) - last by Justin Case.. on Jul 14 at 8:03 PM
Transmeta and AMD put aside their differences to give each other a shoulder to lean on

AMD last week announced a strategic monetary investment in former CPU manufacturer Transmeta, despite quickly mounting debts. Transmeta will receive $7.5 million cash -- approximately 4% of Transmeta's market capitalization -- from AMD in exchange for preferred stock.

"AMD has long been a leader in the development and delivery of energy-efficient, high-performance computing technologies, standards and initiatives," stated Transmeta president and CEO Les Crudele. "Transmeta has been proud to endorse and contribute to those industry leading activities, and we look forward to continuing our collaboration with AMD on technology initiatives in the future."

Transmeta posted a net loss of near $24 million for fiscal 2006; AMD posted a net loss of $611 million for the first quarter of 2007.

Last year, AMD closed the doors of its Geode and Alchemy divisions. Both divisions produced low-power processors that directly competed with Transmeta's Crusoe and Efficeon family. AMD then spun off the remnants of its Alchemy division to Raza Microelectronics, leaving AMD with an x86-only product lineup.

AMD hopes to take advantage of energy-efficient Transmeta technology to benefit AMD products and customers. AMD President and COO Dirk Meyer claims, "Transmeta was a key ally in helping to bring our highly-successful AMD64 technology to market and has supported the widespread industry adoption of both AMD64 and AMD’s HyperTransport technology."

Last February, Transmeta cut its workforce by 39% to focus its interests on just intellectual property. For the majority of 2007, Transmeta's primary revenue prospect revolves around an open patent infringement lawsuit aimed at Intel over ten patents allegedly found in low-power Intel processors.  Intel immediately counter-sued over this claim.

Unlike the acquisition of ATI, this is only an investment in Transmeta by AMD. The companies have not announced any plans of collaboration.

Comments     Threshold


This article is over a month old, voting and posting comments is disabled
I don't get it
By phatboye on 7/8/2007 9:00:39 AM , Rating: 2
I don't get it either. The article didn't explain what did AMD gain from all of this or did I miss something?




RE: I don't get it
By Gul Westfale on 7/8/2007 9:22:24 AM , Rating: 2
well i guess that there might be some sort of collaboration/exchange of patents (transmeta has a lot of stuff related to energy efficiency and software emulation).


RE: I don't get it
By Obujuwami on 7/8/2007 1:09:38 PM , Rating: 2
The article states that AMD got premium stocks, which in the simplist of terms means that AMD get a voice on Transmeta's board of directors. Similiar situations can lead to hostile take overs (one company buying up a bunch of stock them forcing a sale/merge with another company) but I doubt that AMD is looking to do that.


RE: I don't get it
By Justin Case on 7/8/2007 3:58:13 PM , Rating: 5
Three things:

1. Patents, patents, patents. Code morphing, power management, you name it. Licensing those was currently Transmeta's core business. With added weight in Transmeta's board, AMD gets to license that technology for less and (more importantly) gets a say on who else can license it.

2. Lawsuit(s) against Intel. Transmeta is also suing Intel (for different reasons) and has a lot of documents that were seized before Intel "misplaced" a bunch of stuff, due to the ("bigger") AMD suit. Those could be very valuable to AMD. Also, if Transmeta wins, AMD's investment pays off directly.

3. Nvidia. With AMD buying ATI and Intel developing a "real" line of GPUs, a verly likely move for Nvidia would be to buy Transmeta (or VIA) so they could get their hands on a viable (and licensed) line of x86 CPUs. By buying a share of Transmeta, AMD can influence that deal, or at least profit from it.

So this move would have been a no-brainer if AMD had plenty of cash. But they don't. This money is probably coming from their latest loan, and they'll have to pay interest on it, and give up other things. Presumably they did the maths and figured out it was still a good deal.


RE: I don't get it
By omnicronx on 7/8/2007 4:12:36 PM , Rating: 2
quote:
So this move would have been a no-brainer if AMD had plenty of cash. But they don't


if amd has 600 million to lose in a quarter, 7.5 million for a deal that will keep intel away from soi and possibly nvidia from the cpu market is more than worth it for amd.
If amd wants to remain competitive, they need to keep close ties to transmeta for the time being


RE: I don't get it
By TrogdorJW on 7/8/2007 7:24:21 PM , Rating: 2
7.5 million out of 611 million is a drop in the bucket. It's not like people are going to be more upset because net losses went up 1% - and that's assuming they do end up losing money on the Transmeta investment, which isn't guaranteed.


RE: I don't get it
By vignyan on 7/10/2007 6:39:57 AM , Rating: 2
FYI,

Intel Opted out of SOI not that they did not have the technology to develop it! AMD spent $2Bn in retooling its fabs to 65nm and 300mm. You can get a brand new Intel Fab for half a billion more! ;)


RE: I don't get it
By Mitch101 on 7/9/2007 10:28:58 AM , Rating: 2
Fully agree and to add I think AMD expected Barcelona to have already launched increasing AMD's buying power allowing them to fulfull a rumor that was going around.

There was a rumor going around about 6 months ago that AMD was also looking to buy Transmeta after they purchase ATI. Most probably for the patents and their low power line.


RE: I don't get it
By TomZ on 7/8/2007 9:48:40 PM , Rating: 3
quote:
I don't get it either. The article didn't explain what did AMD gain from all of this or did I miss something

The enemy of my enemy is my friend.


Its nice
By Acid Rain on 7/8/2007 6:42:33 AM , Rating: 3
It's nice that small innovative companies like transmeta can be supported by large profitable... oh wait.




RE: Its nice
By MaK2000 on 7/8/2007 4:23:08 PM , Rating: 2
Intel is only thinking next Gen? They have how many teams working at one time to bring out the next three or four gens of CPUs. The have their CPU lineup all the way to 2012. 45nm Penryn this year. 8 core 45nm Nehalem next year. 16 core 32nm Westmere in 2010. I don't remember what comes out after that but if they stay on their roadmap AMD will be hard pressed to keep up. I hope they do or it is bad for everyone but you can't just speculate that AMD has this surefire plan to destroy Intel. They are doing some very risky business and it could pay off for them or they could be asking the government for money just to stay afloat. The next 2 years will be very interesting for the CPU market. Intel has top notch onboard graphics and who knows what they will do in the discrete market. You just can't say that this is going to work out for them. They are in a very vulnerable position and competing with a company that knows how to push the right buttons to make another one faulter.


RE: Its nice
By MaK2000 on 7/8/2007 4:24:21 PM , Rating: 2
Guess I clicked the wrong reply. This is to the "chess game"


RE: Its nice
By PrinceGaz on 7/8/2007 5:14:45 PM , Rating: 3
Intel will only have two teams working on next-gen CPUs: one on the next enhancement of the chip itself, and the other on a die-shrink of current designs.

Intel do not have top-notch onboard graphics. They're fine for Vista but no match for AMD (ATI) or nVidia's integrated graphics chipsets, either in raw power or driver quality.

AMD will no doubt survive one way or the other. We need them to otherwise processor prices will rise steeply. They'd probably stay at current levels initially (or gradually increase) with new faster Intel only processors being introduced at steadily higher prices until $1000 was common-place. Before you know it, you'd be looking at $250 for an entry-level CPU, $500 mainstream, and anything from $800 upwards for high-end consumer processors (the top-end could easily rise from $1000, up to $1500 or even $2000 for the very fastest chips). That would be great for Intel but not so great for us, the customer.

AMD are in trouble at the moment and it all depends on how well the K10 performs in real-world situations, and at the clock-speeds they can produce it at. Things do look bad with figures like 2.4GHz being thrown around as a high-end part, as it can't be so much more efficient at that speed to outperform a roughly 3GHz Core 2 Duo, but hopefully the K10 design will at least match the Core 2 clock-for-clock and be able to make AMD some much needed money.

If K10 cannot at least match Core 2 clock-for-clock, then AMD may as well admit defeat themselves and look for someone to take them over and provide a big investment to make them a match for Intel.


RE: Its nice
By Justin Case on 7/9/2007 11:11:52 PM , Rating: 2
quote:
If K10 cannot at least match Core 2 clock-for-clock, then AMD may as well admit defeat themselves and look for someone to take them over


Huh? By that logic, Intel should have "looked for someone to take them over" in the Pentium III and Pentium 4 days, which were slower (both clock-for-clock and in absolute terms) than the K7 and K8. The Pentium 4 was slower clock-for-clock than the Pentium III (and the first models were slower than the Pentium III, period). And yet both were profitable.

AMD and Intel are not in the business of having the fastest CPU in the world (they'd probably end up losing to IBM anyway); they are in the business of making money. What determines a CPU's success is how profitable it is. Most people don't buy the absolute fastest model anyway; they buy mid- and low-end CPUs. If AMD can produce those in enough volume, and at low enough manufacturing costs, they can make a lot more money than if they had a super-duper CPU that was so expensive and so hard to manufacture that no-one could buy it.

Today AMD has deals with a lot of big PC OEMs and HPC companies - something they didn't have in the K7 days. News of its death are greatly exaggerated. ;-)


RE: Its nice
By vignyan on 7/10/2007 6:46:34 AM , Rating: 2
quote:
AMD and Intel are not in the business of having the fastest CPU in the world (they'd probably end up losing to IBM anyway);


IBM has the fastest CPU in the world! NOTTTTTT!!! :P


RE: Its nice
By Justin Case on 7/14/2007 8:03:04 PM , Rating: 2
For highly specialized situations, no. In terms of general-purpose chips, yes, by a pretty comfortable margin.

http://arstechnica.com/news.ars/post/20070521-ibms...


RE: Its nice
By bfonnes on 7/9/2007 3:29:55 PM , Rating: 2
huh? Since when has Intel had a decent onboard graphics solution...


so aggressive
By skyyspam on 7/8/2007 5:26:20 AM , Rating: 1
Is it just me, or does everything about AMD in the recent past just seem risky? Borrowing billions of dollars, buying ATI, cutting the Geode and Alchemy divisions, buying Transmeta...

I wonder how much sleep the execs manage to pull at night?

Personally, I know it's a long shot, but I hope AMD can pull it off...




RE: so aggressive
By rbuszka on 7/8/2007 8:42:36 AM , Rating: 3
Read the article - they didn't 'buy' Transmeta.


RE: so aggressive
By codeThug on 7/8/2007 2:32:33 PM , Rating: 2
For the 17+ years I've known AMD and it's employee's, everyday is a risk in the shadow of ChipZilla. Fear is a mighty motivator, and AMD folks live with it constantly.

This deal with Transmeta and the purchase of ATI may well allow them to pull something off that no one expects.


RE: so aggressive
By TomZ on 7/9/2007 9:37:32 AM , Rating: 2
You mean an accelerated bankruptcy, or sale to a private venture capital firm that will slice 'n dice AMD and sell it off as pieces and parts?

In my opinion, the ATI acquisition is a potential long-term investment that AMD needs to more effectively compete with Intel. But it is also an acquisition that it could not afford, in terms of money and focus.