AMD soars high

[Urban Ranger]

Cray preps new Opteron-based product line

Cray has given AMD's Opteron chip a big vote of confidence, saying it will roll out a new line of supercomputers based around the product.

In 2004, Cray will begin selling systems with anywhere from 64 to more than 10,000 Opteron processors. All of the supercomputer-style systems will have a custom interconnect developed by Cray to deliver the kind of internal bandwidth needed for complex technical computing tasks. The decision to roll out the Opteron-based product line follows a deal Cray won last year to create a 10,000-plus processor system for Sandia National Labs.

AMD's new 64-bit chips should keep them in the race. This is a Good Thing.

[Asher]

Cray is still around?

[NeOmega]

I bought some AMD stock about 7 months ago...... I am very happy with my purchase.... My avatar for CFC has the AMD logo....

I love AMD honestly... I am fanatical about them, I love their never say die attitude, the way they actually came out with a faster, cheaper chip than Intel in the late 90's. Eventually I hope to own a whole lot more AMD stock, and never let go.

Everyone has their price, and I've marked my sell point, but it is very, very high.

I bought AMD for these reasons (My first stock ever):
1) they produce a product: (I can hold an AMD chip, I will never buy Insurance, banking, or any other "financial" stocks, whose only product is paper with numbers typed on it and memos.... ie Enron.

2) They are in second place: Fighters are in second place. I believe second place is where the most streamlined companies exist. First place = complacency.

3) No defense or Oil stocks..... no way.

4) An AMD fab plant is in Germany. An Intel Fab plant is in Israel. One is much more likely to get bombed than the other.

5) If I ever sell my shares, or some of my shares, it will be to buy a brand new computer, with an AMD chip as the CPU, of course.

[Asher]

Originally posted by NeOmega
4) An AMD fab plant is in Germany. An Intel Fab plant is in Israel. One is much more likely to get bombed than the other.

I don't understand this logic at all.
AMD has one plant capable of producing its chips, and it's in Germany. Intel has many plants capable of producing its chips, most of them in the USA, but one is in Israel.

5) If I ever sell my shares, or some of my shares, it will be to buy a brand new computer, with an AMD chip as the CPU, of course.

AMD, as a company, totally blows. Always has, always will. Every couple years they release a great product and people rally behind them, then they find some way to f*ck it up. It's just part of their history and who they are.

AMD still posted a loss this quarter, and they're still having trouble meeting demand, and they still lack any big OEM support. Not to mention that the market for desktops is stale and the growing market is in laptops, where AMD's marketshare is laughable and not growing at any rate, while Intel's Centrinos are flying off the shelves...

Notice that the Y-axis is logarithmically scaled, too.

[NeOmega]

Originally posted by Asher

I don't understand this logic at all.
AMD has one plant capable of producing its chips, and it's in Germany. Intel has many plants capable of producing its chips, most of them in the USA, but one is in Israel.

AMD has a plants the US, Japan, and Asia as well. Israel... well secondly, I wouldn't do anything to support Israel. I like Germany alot better.

AMD, as a company, totally blows. Always has, always will. Every couple years they release a great product and people rally behind them, then they find some way to f*ck it up. It's just part of their history and who they are.

AMD did not F up the Athlon, the dot-com bubble burst. Intel went down too.

AMD still posted a loss this quarter, and they're still having trouble meeting demand, and they still lack any big OEM support. Not to mention that the market for desktops is stale and the growing market is in laptops, where AMD's marketshare is laughable and not growing at any rate, while Intel's Centrinos are flying off the shelves...

The market for desktops will rebound. And AMD has a cult following that cannot be shaked... and go into a PC parts store.... I bet 3 out of 5 times, if you want a custom made desktop, the guy will want to use AMD. Just my impression of real computer fanatics.

Intel's "market share" isn't growing either... considering the only other contender is the only other chip maker to see any growth... Via.

[Mordoch]

Originally posted by Asher
AMD, as a company, totally blows. Always has, always will. Every couple years they release a great product and people rally behind them, then they find some way to f*ck it up. It's just part of their history and who they are.

AMD still posted a loss this quarter, and they're still having trouble meeting demand, and they still lack any big OEM support. Not to mention that the market for desktops is stale and the growing market is in laptops, where AMD's marketshare is laughable and not growing at any rate, while Intel's Centrinos are flying off the shelves...

Ok first just to make biases clear, I'm an AMD stockholder as well, I bought the company because I am confident of their future prospects.

Amd has historically done poorly, but I feel after their aquisition of many former DEC researchers, the people who designed the Alpha microprocessors, AMD stopped being a technological follower.

AMD posted a much smaller loss than analysts expected, I think its highly likely they will post a profit this quarter, and I'll be stunned if they don't do so by the first quarter of next year.

On big OEM support, just what do you think IBM selling AMD Opteron chips are part of their server lineup counts as?! I'd expect general OEM support for the AMD 64 chips to become more visable once they come out in larger volumes, right now many of them are being sold in high end gaming machines from companies such as Falcon Northwest.

On AMD's marketshare, I belive they are esentially doing just fine. What is happening is that the AMD 64 chips are bigger than the Athlon ones, meaning their total production numbers go down, but they command much higher prices. AMD is trouble having meeting demand for the AMD 64 chips since they are flying off the shelves! In the second half of 2004, the situation is going to drasticly change as AMD switches to 90mm technology. Besides allowing higher speeds and the like, this technology it will effectively decrease the size of individual microprocessors by half. This should lead to a doubling of AMD's microprocessor production, and if AMD still has trouble meeting demand, that means Intel is in deep trouble. AMD also potentially has the option of outsources production of some of its microprocessors to IBM fabs to meet demand if necessary.

On Intel Centrino chips, I actually recent read an article saying sales were disapointing and Intel was having trouble persuading potential buyers to purchase them. AMD already has a mobile version of their AMD 64 microprocessor for sale with laptop systems. As volume increases on the AMD 64 line, I expect the processors to start taking up more laptop space.

Finally, Intel has been stumbling recently with their technology. We'll see when the P4EE which was paper launched to try to steal the thunder from the AMD 64 line actually becomes available, but right now Prescott keeps on getting more and more delayed. Right now I'm hearing the Intel may paper launch the Prescott in December, but it won't be until January of February that it is actually released. I don't expect AMD to stand still technologically in the meantime.

[Asher]

Originally posted by NeOmega
AMD has a plants the US, Japan, and Asia as well. Israel... well secondly, I wouldn't do anything to support Israel. I like Germany alot better.

The only plant AMD has capable of producing Athlons and Opterons is in Dresden, Germany. Period. The others are for things like PCI-X controllers, Flash memory, other trivial things...

AMD did not F up the Athlon, the dot-com bubble burst. Intel went down too.

Well, I would disagree with you. Palomino and Thoroughbred were all thoroughly disappointing. AMD lost considerable marketshare in the past couple years because of this. Blaming it on the dot-com bubble doesn't cut it -- Intel's chips are more expensive and they managed to gain marketshare when the dot-com bubble burst while the cheaper one lost marketshare? Figure that one out.

The market for desktops will rebound.

It's not the point of economics for why it's slumping, it's because they've simply gotten fast enough that only gamers or media-type people need fast computers. The vast majority of computers out there won't be upgraded for a long time, a Pentium III 600 is more than adequate to do what most people want -- and they realize that.

The next step is to give them power on the go, and hence the laptop market's growth...

And AMD has a cult following that cannot be shaked...

AMD's cult following is incredibly tiny in the grand scheme of things.

Intel's "market share" isn't growing either...

Yes, it is. It went up 5% over last year.

[Asher]

Originally posted by Mordoch
Amd has historically done poorly, but I feel after their aquisition of many former DEC researchers, the people who designed the Alpha microprocessors, AMD stopped being a technological follower.

True. But the major concern now is that they're leading the industry in the wrong direction. Everyone else wants to ditch the 20-year old x86 which, itself, was a a "temporary" ISA from Intel rushed into production as per IBM's request. It wasn't meant to last 2 years, let alone 20+.

Everyone but AMD has agreed to lay to rest x86 and transition to far more modern ISAs -- IBM and Motorola chose PowerPC, Intel chose IA-64, etc. AMD doesn't have the R&D or influence to come up with their own so they modify x86 and tell everyone we need x86 chips in everything.

There is a real concern now that AMD is stagnating technology and performance by clinging on to x86 when everyone else wants to ditch it.

AMD posted a much smaller loss than analysts expected, I think its highly likely they will post a profit this quarter, and I'll be stunned if they don't do so by the first quarter of next year.

I recall Sanders promising a profit for Q203, and wasn't that a $500M some-odd loss?

On big OEM support, just what do you think IBM selling AMD Opteron chips are part of their server lineup counts as?!

That's true, IBM does sell Opterons.
But that market isn't what I had in mind, indeed it's quite small compared to the consumer market.

How many consumer OEMs do you see actively producing and advertising Athlon-based PCs these days?

I'd expect general OEM support for the AMD 64 chips to become more visable once they come out in larger volumes, right now many of them are being sold in high end gaming machines from companies such as Falcon Northwest.

Well Dell has given AMD64 the cold shoulder, Opteron and Athlon in all.

AMD is trouble having meeting demand for the AMD 64 chips since they are flying off the shelves!

This is another important point -- AMD lacks manufacturing capacity. So much so that it's actually quite a problem. In times like this where they can make a fortune off their rather-excellent Athlon 64 family, they can't make enough to satisfy demand.

It's another reason why Intel has such an advantage, it always has excess capacity even.

In the second half of 2003, the situation is going to drasticly change as AMD switches to 90mm technology. Besides allowing higher speeds and the like, this technology it will effectively decrease the size of individual microprocessors by half. This should lead to a doubling of AMD's microprocessor production, and if AMD still has trouble meeting demand, that means Intel is in deep trouble.

There are major flaws in this logic.
First of all, the 130nm->90nm shrink will only reduce the core by 33% or thereabouts. Second of all, the 130nm process they're on now is quite mature (very mature, actually). When the initial switch to 90nm chips is made, the process is anything but mature -- yields are far lower. At the end of the day they can probably get slightly more chips out at 90nm than 130nm initially, but at higher frequencies.

You also forget that Intel is going to beat AMD to the 90nm punch, along with the new Prescott core. Not only this, but Intel has the advantage of 300mm wafers while AMD is still on 200mm and won't upgrade to 300mm until 2005. 300mm wafers have the benefits of about 30% cost savings compared to 200mm wafers.

On Intel Centrino chips, I actually recent read an article saying sales were disapointing and Intel was having trouble persuading potential buyers to purchase them.

I think you're thinking of that TheRegister article.

I've no idea where they're getting their info -- I had to wait a month to get my ThinkPad T40 due to a dearth of Centrino 1.5GHz's.

AMD already has a mobile version of their AMD 64 microprocessor for sale with laptop systems. As volume increases on the AMD 64 line, I expect the processors to start taking up more laptop space.

The mobile Athlon 64 competes with the Pentium 4-M, not Centrino.

Finally, Intel has been stumbling recently with their technology. We'll see when the P4EE which was paper launched to try to steal the thunder from the AMD 64 line actually becomes available, but right now Prescott keeps on getting more and more delayed. Right now I'm hearing the Intel may paper launch the Prescott in December, but it won't be until January of February that it is actually released.

True, Prescott was delayed til January. The reason for this is this is the first time Intel used a state of the art computer program to lay out components of the CPU to maximize efficiency. Unlike all other CPUs we see, you can't take a die photo of Prescott and draw little boxes over where the FPU is -- it's actually spread out all over a wider area.

The problem with this is that the chip was too efficient at spreading heat, so now everything is more-or-less uniformly hot. Combine this with exponentially worse gate-leakage when moving from 130nm to 90nm, and it's a hot chip. They delayed it so they could get a C-stepping out to bring the TDP down to 90W or so from 100W.

AMD will have to deal with the same thing when they move to 90nm next year.

I don't expect AMD to stand still technologically in the meantime.

I do -- the only new products they're launching are rebadged Opteron 1xxs (AthlonFX 43, etc) and an Athlon 64 3400+ -- and Intel will launch a 3.4GHz Northwood to compete with the Athlon 64 3400+.

[NeOmega]

Originally posted by Asher

The only plant AMD has capable of producing Athlons and Opterons is in Dresden, Germany. Period. The others are for things like PCI-X controllers, Flash memory, other trivial things...

I don't think having any kind of plant... especially a fab plant, in Israel is a good idea. "losing "just one" could be devestating, and Israel seems to like the prospects of WW III, I would just hope all goes well for Israel if an armed conflaguration were to break out, and I were an INTC holder.

It's not the point of economics for why it's slumping, it's because they've simply gotten fast enough that only gamers or media-type people need fast computers. The vast majority of computers out there won't be upgraded for a long time, a Pentium III 600 is more than adequate to do what most people want -- and they realize that.

And that is why I love AMD. I have heard "computers are fast enough" before. This argument has no legs. A computer will never be fast enough for gamers, which is still a growing industry. And, I am truly convinced, it is gamers who demand better and fasater chips, who continue to press for more. Then business catches up once they realize what a faster computer can do.

AMD's cult following is incredibly tiny in the grand scheme of things.

Merril Lynch started his company by making basic observation of what people were buying, not looking at numbers. I have made my observations, and nearly doubled my money in 7 months. When I bought the stock no-one liked it, but I did. I observed. I shopped around for a chip, and allthough I was an AMD fan, the small shopkeepers were friendly, and AMD lovers. It was like the passion was with them.

Yes, it is. It went up 5% over last year.

Hmm... I must be seeing different numbers than you

Intel's market share 2002 Q1 80.8; 2003 Q1 81.7.

Off the top of my head, AMD's market share at it's highest was around 22%.

[Asher]

Originally posted by NeOmega
I don't think having any kind of plant... especially a fab plant, in Israel is a good idea. "losing "just one" could be devestating, and Israel seems to like the prospects of WW III, I would just hope all goes well for Israel if an armed conflaguration were to break out, and I were an INTC holder.

Intel has lots of excess capacity, it's not a real concern. If AMD's one CPU producing plant was there, then that'd be a concern. But since Intel has many of them, with the most high-tech of which being state-side, I don't see any cause for concern.

And that is why I love AMD. I have heard "computers are fast enough" before. This argument has no legs. A computer will never be fast enough for gamers, which is still a growing industry. And, I am truly convinced, it is gamers who demand better and fasater chips, who continue to press for more. Then business catches up once they realize what a faster computer can do.

I've already said gamers are an exception. But I'm not sure why you think this argument has no legs -- look at the sales statistics for last year.

Desktop sales are flat, Laptop sales are booming, Handhelds are falling.

Merril Lynch started his company by making basic observation of what people were buying, not looking at numbers. I have made my observations, and nearly doubled my money in 7 months. When I bought the stock no-one liked it, but I did. I observed. I shopped around for a chip, and allthough I was an AMD fan, the small shopkeepers were friendly, and AMD lovers. It was like the passion was with them.

Shop keepers love AMD because of their slack control on the grey market. Shop keepers have been in love with AMD for years, over that time AMD's stock has devalued.

I think that's a pretty foolhardy way to look at how one buys stock, but perhaps that's just me. Having a great product means absolutely nothing when: