AMD splits in two...

silent_guy

Veteran
Supporter
http://www.nytimes.com/2008/10/07/technology/07chip.html?hp
In a dramatic effort to maintain its position as the only real rival to Intel, Advanced Micro Devices plans to announce Tuesday that it will split into two companies — one focused on designing microprocessors and the other on the costly business of manufacturing them.

There you go. Definitely a good thing for the design part of AMD, but it's amazing that they still find greater fools in far away countries who are willing to sink money in pure fab operations.

I guess I just don't get it... Anyway willing to clarify the financial math about how this is a sensible investment?
 
its sensible in that there is much less overhead, is one company going to be a parent company and another a daughter or two seperate identities?

That is suprising they are still getting outside investors....
 
Won't this get them in legal trouble with Intel for technically outsourcing more than 20% of x86 chip production to a third party, an item that was rumored to be part of the licensing deal between the two companies ?
 
That is suprising they are still getting outside investors....

Not if your country is filthy rich, like UAE and Saudi Arabia by proxy, but is doomed to extreme sub-saharan poverty after the cheap oil is gone.
 
The only way this gets sensible is "The Foundy Corp." being extremely large scale.


The good part for AMD however, seems to be Luther Forest.
 
http://www.nytimes.com/2008/10/07/technology/07chip.html?hpI guess I just don't get it... Anyway willing to clarify the financial math about how this is a sensible investment?

I'm no expert on merchant fab business models, but I think it boils down to having more customers, which has two benefits:

First, multiple customers with various products means demand is probably going to be more consistent. It costs a lot of money just to keep a fab operating and staffed, even if it's not running at full capacity. AMD has to have enough fab capacity so that if their products are as successful as they hope, they won't be supply constrained. But then if the products are somewhat less successful, they've got underutilized fabs. So the fixed cost of running the fab is amortized over less revenue. Changing fab capacity isn't something you can do one a dime.

Second, costs of developing new process and fab tech are going up. The cost is amortized over the chips you build with that process. So while increasing gross margins helps pay for it, at some point you really just need to have more volume. Since AMD isn't able to grow volume enough on its own, the fabs need to take on other customers.
 
Won't this get them in legal trouble with Intel for technically outsourcing more than 20% of x86 chip production to a third party, an item that was rumored to be part of the licensing deal between the two companies ?

The limit itself is no rumor. AMD execs were forced to admit it exists in a recent conference call.
The unknown is the percentage, which was last documented publically at 20%.

Splitting the company like this obviously puts it at 100%, so what kind of game AMD is playing here might depend mightily on the particulars of the split with regards and what the foundry contract agreement between the two entities contains.


I guess I just don't get it... Anyway willing to clarify the financial math about how this is a sensible investment?

For Abu Dabi, it's a chance to spend some billions they could have spent on credit default swaps to instead become in a year a partner in a multinational company with presence in high-end semiconductor design and manufacturing worldwide. In the long run, it might bring high-tech industry and an advanced engineering draw to a rather tiny country that couldn't support such an infrastructure with people or a market otherwise.
They had to spend that money somewhere, I guess.
 
Last edited by a moderator:
According to wiki, the sovereign wealth fund for Abu Dhabi is at 875 billion dollars in assets.

This AMD thing was probably funded by them checking the crevices of the sovereign wealth fund's couch.
 
Well if no one else was willing to pony up and buy AMD to keep it going, I guess we can't complain. No one wants to see an Intel-only processor world or a graphics market without ATI graphics.

AMD have reorganised and sold equity in order to get the cash infusion they needed. They needed to do something bold like this, as their previous business model wasn't really going anywhere by banging it's head against the Intel brick wall.
 
In the long run, it might bring high-tech industry and an advanced engineering draw to a rather tiny country that couldn't support such an infrastructure with people or a market otherwise.
They had to spend that money somewhere, I guess.

They are a business proxy for Saudi Arabia, so I guess it is good for both of them. Saudi Arabia has a moderately population, 28million, 5 million of immigrants to do regular jobs, so I think they have enough potential man power... But their education, even higher education is still based on religious studies, so, I am not very sure if this technological plan will work in the long run.

Anyway, after their cheap oil is over, or people stopping using it as fuel they will be in a worse situation than sub Saharan countries, because they are nothing more than an oil field, economicaly speaking.
 
Yeah - it seems to me that UAE are spending a lot of money to diversify their economy away from oil.

The Saudis, on the other hand, aren't.

This split looks like a pretty good deal for AMD to me. Time will tell whether or not this is the case but I feel we certainly need a strong competitor for Intel.
 
For Abu Dabi, it's a chance to spend some billions they could have spent on credit default swaps to instead become in a year a partner in a multinational company with presence in high-end semiconductor design and manufacturing worldwide. In the long run, it might bring high-tech industry and an advanced engineering draw to a rather tiny country that couldn't support such an infrastructure with people or a market otherwise.
Ok, but it means you need to have quite a bit of confidence that fabs are the way to go. I didn't look up how profitable UMC and Chartered are, but they're really eating the crumbs that are falling from TSMC's table. It's safe to assume that The Foundry Company has no plans to go head on against TSMC, so it'll be just one more party eating crumbs. Most companies use UMC as a way to get TSMC to lower its prices, they are rarely selected because of their own merits. I don't see why it would be different for TFC.

The problem with fabs is that there's the never ending need for tons of capital to make it work. I think the guys from Abu Dhabi are going to be throwing money in a bottomless pit.

They had to spend that money somewhere, I guess.
Hey, treasuries have been doing great lately!
 
Ok, but it means you need to have quite a bit of confidence that fabs are the way to go.
The UAE's modus operandi is a gigantic version of "if you build (acquire) it, they will come", looking at everything else they've been buying and building.
With the biggest towers, the biggest commercial districts, artificial islands, and attempting to pay for higher education branches in Dubai, they're hoping something will stick.

Perhaps Abu Dhabi will commission AMD to build a national vanity chip, the tallest microchip in the world. ;)

The fact that many of these ventures have not really gained any traction is a sign that things can flame out badly, though AMD might be able to snap itself back if they do. It's not like the UAE can seize much, since nothing so far is actually within the country.

I didn't look up how profitable UMC and Chartered are, but they're really eating the crumbs that are falling from TSMC's table. It's safe to assume that The Foundry Company has no plans to go head on against TSMC, so it'll be just one more party eating crumbs. Most companies use UMC as a way to get TSMC to lower its prices, they are rarely selected because of their own merits. I don't see why it would be different for TFC.
TFC does have a pretty good piggy bank backing it up, and a significant direct customer of TSMC's is a primary client for CPUs. It's a pure-play foundry that from the outset has financial friends who aren't affected by any ongoing credit crisis and unafraid to spend lavishly, something the other smaller fabs and maybe TSMC would kill for.
The UAE doesn't have to cut capex in line with the rest of the industry.

The problem with fabs is that there's the never ending need for tons of capital to make it work. I think the guys from Abu Dhabi are going to be throwing money in a bottomless pit.
I'd say it's likely. Given the cash they make on their oil thing, and the costs if they fail move off of it before it fails, the comparatively minor costs of a blunder into semiconductor manufacturing seems like a calculated risk worth taking.

Hey, treasuries have been doing great lately!
Remember the last country that modernized by mastering buying bonds? ;)
 
The Saudis, on the other hand, aren't.

They haven't been successful in that for lack of proper education initiatives, but they are starting now to focus properly. See this for example:

http://www.kaust.edu.sa/news-releases/king-abdullah-university-and-ibm-build-supercomputer.aspx

I say that UAE and Abu Dhaibe are proxies for Saudi Arabia, because they are a nice place to make business, without having to showcase the obvious poverty and extreme dictatorship. I don't have proof right now with me, but I guess it makes sense, that's why I said that.
 
Splitting the company like this obviously puts it at 100%, so what kind of game AMD is playing here might depend mightily on the particulars of the split with regards and what the foundry contract agreement between the two entities contains.

Here's the actual agreement. It's not complete, but I think it's sufficient:
http://contracts.corporate.findlaw.com/agreements/amd/intel.license.2001.01.01.html

Look at section 1.22, which define a "subsidary". AMD doesn't own 50% of The Foundry, but they DO have 50% voting power, so it must be considered a subsidary and therefore included in the agreement.
 
I'm with silent_guy here. How could The Foundry Company possibly compete with TSMC, which just has so many many competitive advantages? Scale, expertise, and relatively low labor costs being soem of these advantages.

Moreover, because of these advantages it will be seemingly very difficult to match TSMC on price. So does this actually hamstring AMD's ongoing businesses, especially in light of the likelihood of falling CPU prices.

Was this the only option available to AMD? Was this decision made so AMD executives could have a new company to run. Were there no ptential buyers for the Dresden fabs? If not what does that say about the potential competitiveness of these fabs.

I mean AMD does not have a great strategic track record. Had they done this years ago they would not be in the financial mess that they are today.
 
Why is it safe to assume this?

They're the de-facto standard for fabless companies. If you ever need a piece of silicon IP, standard cells, IO pads, Ethernet/USB/PCIe/... transceivers, fuse boxes, ram compilers, PLL's etc, you're going to find them first (and, in many cases, only) for TSMC processes.

Their only customers are external ones and they have years of experience in keeping them happy. Company dedicated fabs are often used to one particular tool flow which makes then less flexible.

They have massive economy of scale and should be able to get the best pricing wrt new equipment. They are able to partially fund new fabs with older fabs that are already written off (6 according to their website.) AMD has never had that luxury and TFC won't be able to do so any time soon. It won't be easy to compete on price with TSMC.

They're a safe bet. They're highly profitable and aren't going anywhere. It's to be expected that AMD will have priority when new state-of-the-art processes are coming on-line and that other customers will have to wait in line.

Location? There's a large amount of third party services around TSMC. Testing, packaging etc. Not sure how important that really is though.
 
Back
Top