Friday, March 31, 2006


I have just read a very nice article:

Although the author makes the mistake of not understanding how investments are accounted, his point holds.

He compares all the hardships AMD has experienced along these years to the family that struggled to put through college all the children, and got to the graduation year in very good shape. The oldest of the family is already having a top-flight job, and this is just the beginning of the new situation of prosperity.

The miracle is that this family believed so much on the kids. Not everyone was born to go to college. But AMD's competitive advantages deserved full support. Since the days of the Athlon, in 1998, it has been a constant investment in a set of principles that didn't pay off perhaps until last year.

One of those principles is to respect the market.

I very well remember my days at the University when I was full of admiration for Intel. In those Pentium days, Intel was perhaps at its zenith, it was a company that innovated with vigor, it was an open company, it had an inviting attitude toward its products, and you could perceive a technical excellence not present in previous generations (which lasted only until the Pentium Pro, as was later proved). But of course, Intel has always been greedy. In any case, we reached the years of the incoming Pentium III with Intel being the undisputed leader of the computing necessities of the world, when my great dissapointment happened: Intel started to promote the Pentium III as "optimized for a fast internet experience". That is, it was not enough to supply everything that the market asked for, greed pushed them to *lie* [there is no connection between any of the features present in a Pentium III and a fast internet experience, if anything, it would be the MMX (the integer SIMD extension) already present in the Pentium II, SSE (floating point SIMD extensions) was something different aimed towards computing that didn't have to do with "Fast Internet Experience"].

Then, AMD put in the market the Athlon, and because it had superior architectural features, it took the performance lead from the Pentium III. Intel, who apparently succeeded in fooling the market, decided that they were not going to do actually superior processors to those AMD was doing, but that they were going to do inferior processors that they could sell as if they were superior by lying and deceiving the market. So, what I call the "gigahertz obsession", was born.

The gigahertz obsession was simply to do processors with ever increasing switching speeds, regardless of how much useful computation (real performance) the processor could deliver, and later on, regardless of how much power it would consume.

By this antics, deceiving the market, Intel substantially helped AMD to have a close encounter with bankruptcy.

Then it came the transition to 64 bits, when, because of greed, Intel wanted the whole world to go their propietary architecture route (Itanium) and didn't do an extension to x86.

But AMD got in their way again, it did the extension, with striking simplicity and elegance, so AMD64 was born.

What has done AMD about AMD64? has it fooled the market telling it that it is imprescindible? no, of course not.

What is Intel doing with V//V nowadays? V//V == snake oil

Live! is an entirely different approach.

That's how Intel's credibility decays exponentially under AMD's respectability pressure.


Intel got to where it is because it actually had very good products. But in an indeterminate moment, marketing took charge at Intel. Perhaps with the Pentium III they were convinced of the fallacy that you can fool the market, so, marketing at Intel began to think that the products sold not because they provided value to the customers, but because they, the salesmen, could find all these clever lies to misled the public to buy their products.

So, they lost the technical focus and were unable to do something even remotely worthy ever since.

The promotion of merit: Current CEO, and current President are people with undoubted merits to have arrived to their positions, and this culture is very evident at the different levels of management.

But Intel is a different story. The fact that a young Israeli team, operating so far from the corporate power center, was the one who received the challenge and responsibility to design and develop the flagship products for the next two years means that top management is not focused nor concerned about providing the absolute best products. It is evident that top management concerns itself with politics and financial manipulations.

AMD was nearly bankrupt, but stood on the principles. And demonstrated, beyond any doubt, that that is the way to succeed.

Right now, we are in the harvesting season, all of what AMD sawn has grown and is ready to be turned into profits. The neighbor doesn't have anything but weeds. There is still a lot of people who has faith in the gigantic neighbor, he says that in the remote corner it has the greatest grain ever grown in these fields, but that just seems to be him once again bragging about what he doesn't have. If his harvest were good, he would be very busy, as we are.

Intel is not even aware of when or why it went wrong. They are just awakening to the fact that they went wrong.

Wall Street is even more pathetic. Wall Street is just complaining that it doesn't have anymore the cash cow it used to milk

Thursday, March 09, 2006

15% Haircut, why?

Why is AMD down 15% in four trading days without any bad news whatsoever?

The answer is simple: To put the money in AMD is speculative, that makes it vulnerable to influences. In this situation, one has to remember that AMD is a member of many indexes where Intel is a major component, and Intel went down heavily. That means that even though the reason why Intel is down has been the competitive pressure AMD is putting on (meaning that AMD should be doing things extraordinarily well), the fact remains that beause of Intel the indexes where AMD and Intel are members are down. So, there is that bit of negative sentiment, that gets amplified and a sell-off happened. The following words are the explanation of this phenomenon.

1) AMD investment is speculative: I am convinced that AMD is going to take the absolute leadership of the processor market from Intel in about a year and a half. But the fact remains that AMD is NOT the leader today, that it hasn't been at the top for decades as Intel has. So it takes a leap of faith to believe that the semiconductor market is going to be changed so radically. Actually, there are solid rational reasons and for those in the know such as yours, truly, for it to be a simple prediction, but for the vast majority of investors and for those who handle most of the money in The Street, it is not a simple fact. While they don't understand the significance of advantages such as Direct Connect Architecture with Hyper Transport over the classical Front Side Bus, they at least can barely hint that AMD seems to be on a roll and put some money on it. But that conviction is not as solid as having seen Intel double and double over and over for decades, giving dividend yields in a reliable fashion. Other things equal, the money put into AMD is the money to "spare". That's why as soon as the Dow Jones or any index is down half of a percent or that a stock market recession looms in the horizon, due to bad economy, a political crisis, or the Federal Reserve raising the interest rate half of a percent, it always happen that the money put into AMD is seen as not safe and the urge to "secure the profits" arises.

2) Intel is being killed by AMD, the competitive situation of Intel is truly pathetic, and all negative options for its future are on top of the table, including bankruptcy. Thus Intel pulls down with some strength all the indexes it is a member of, carrying AMD down.

3) But is AMD's success what is taking down Intel!!: Yes, but Mr. Lemming Nilvestor and Joe Sixstock doesn't believe this despite all the evidence, because they are utterly unable to understand the evidence. And not even in a million years if Intel says that the market for computers or processors is weak. That fits very well into the formerly successful hard coded rule that if Intel is doing Ok., then the market for semiconductors is Ok., and viceversa.

4) But if AMD is still doing great and is Intel's main competitor, then the money that gets out of Intel has a natural place at AMD: No, not really. Intel and AMD may be the two players in their market, but they are companies as different as it gets: Intel appeals to the greedy investors looking for a predatory company that juices every possible penny from their customers, that ruthlessly squashes competition and has no morals. Technological Leadership? as long as it becomes earnings per share and dividends, otherwise, they couldn't care less. AMD is the opposite: A thrifty company that doesn't even aspire to milk the customers but to create an ecosystem where all can share the profits (look at Hector's speech). Note: Not being greedy is not an innefficient strategy, nor it is the aspiration of creating an ecosystem where all share the benefits, it is just a different strategy aiming at sustained growth through cooperation rather than in a zero-sum game (Intel's approach). More to the point, an investor looking for dividend yields may go to Intel but not AMD. Thus the money leaking out of Intel is *NOT* moving to AMD

5) There is also this issue of the historical transition from Intel to AMD. Whenever there is an underdog that manages to score the win against an intimidating opponent that looks as invincible as Intel, a lot of disrruptive processes have to happen. Disrruptive processes are difficult to predict, that's why those of us who are betting on AMD have the merit to gain 5X our investments, because through our Due Dilligence we understood some of those processes while others couldn't.

Look at That's the blog of a guy with multiple acclaimed university degrees. But he just can't accept that AMD could have a lead of perhaps 5 years above Intel... Aristotle: "It is the sign of an educated mind to entertain an idea without necessesarily accepting it", well, Dr. Ge couldn't entertain the thoughts that AMD is crapping Intel with both AMD64 and DCA/Hyper Transport at 90nm. He can't understand that Intel has lost successively in all of its attempts to diversify beyond their core products of processors: They were beaten to humilliation at the 3D graphics accelerator game some years ago, that despite appereances AMD has had the lead in processors since the Athlon 550 MHz in 1999, that the Pentium 4 approach of meaningless gigahertz was actually a doomed marketing response to a concrete technological challenge, that a company that is not able to understand Itanium's failure and keeps pumping billions more into that catastrophe without ever entertaining the idea of killing it for good can't possibly be well managed. The Dr. Ges are not able to understand that it takes years upon years to develop all the concrete performance advantages AMD has developed or is using: Strained Silicon On Insulator processes, AMD64, DCA/ccHT, multicores, Pacifica, etc. Meanwhile, Intel distracted and keeps distracting entire mid-size national budgets into dead ends such as the gigahertz obsession, the Itanium, 32 bits, Front Side Bus, etc; and the Dr. Ges keep thinking that all Intel has to do is to wave a wand and AMD will be dead. Dr. Ges don't understand that at Intel they haven't come up with nothing good in a very long while, that they are going back to the Pentium III (because the high gigahertz obsession in Pentium 4 derailed them from their evolution path) and keep being confident that Intel is going to keep succeeding in the platform business now that AMD broke their dominance in the processor market and they don't have nothing to leverage things such as the Centrino platform, V//V, or any buzzword they can come up with... because they don't see much more dynamic companies such as Broadcom, ATI and nVidia getting into platforms for the duration with far better offerings than what Intel can come up with...

6) Nevertheless, they have become good at Intel at selling "Blue Crystals", "Snake Oil", "Vaporware" and related products. They are good at promising, because they promise so much and deliver so little that that some people still believes them is remarkable. So, they have their IDF, pump Conroe, Merom, Woodcrest as the nukes that are going to destroy AMD (magically, because there is no indication about how that may happen) and panic ensues in some AMD investors. And to top it off, a good PR stunt of some rigged benchmarks and the panic waves add to the other phenomenons and induced a 15% haircut on AMD without that company having bad news at all and having concrete good news...

Now, this are the good news for us, the AMD investors: Once that the weakling money is out from AMD, the proportion of those of us who understand what is going on among the investors will increase and that will give the stock price stability with higher growth just as happened until mid Jnauary when the vapor downgrades began thanks to Wall Street gazing their greedy eyes in our direction. Learn the game: After a sell-off, stable growth for a while, until AMD becomes noticeable again, and the roller coaster will begin again, until another sell-off.

Another positive thing to mention of this episode is that management demonstrated calmness but just ignoring Intel's shows. They keep speaking not with words, but with facts, with products in the market that everyone can test and benchmark.

Wednesday, March 08, 2006

The Live Intel Lies Thread, AMD Truths

  1. Gigahertz is the only thing that counts
  2. Deep pipelines is the proper way to do processor architectures
  3. Nobody needs 64 bits computing
  4. Itanium will conquer the world
  5. 64 bit computing as done in AMD64 is no good, Itanium is the way
  6. We are not copying AMD64, and will not ever support that instruction set, we will focus on Itanium
  7. AMD's dual cores are as good as ours (Intel didn't even have proper single die dual core, but dual dice), ditto for quad cores (dual dice dual core)
  8. Core Duo will be 64bits
  9. V//V is 64 bits
  10. Paxville will whip Opterons
  11. Intel will regain the performance crown in 2005
  12. We lost just 1% market share to AMD in last quarter
  13. SOI (Silicon On Insulator) is unnecessary and expensive, plain Strained Silicon is better
  14. Skype can't do 10-way conference calls without an Intel Core Duo (this may be a lie from Skype)

Please, extend the list with a comment, and I will update the post. You may also comment, document the source, whatever.

Also, don't forget to leave a name to address you in the post and to give you proper attribution.

Saturday, March 04, 2006

According to Intel, AMD would be worth $189 in a price war

For those of you concerned about a posible price war among Intel and AMD, I have prepared an original analysis.

The fundamental assumption is that this would be a war for market share only. I subscribe to the opinion that AMD simply doesn't have to lower its prices in the case Intel chooses to do so, but assuming that AMD doesn't even want lose market share and just wants to keep the status quo, this is how I see the things unfolding:

1) Intel has about 75% market share. If they cut one dollar their average selling price in all segments, then for every $75 that Intel is not winning, AMD doesn't win $25, a 3 to 1 ratio.

2) But it is not quite like that. A gamer won't use a shitty Intel, a datacenter won't use shitty Xeons, no matter how cheap or free they are.

a) A gamer who pays $5000 for a computer won't use nothing but the best in the most important component. Look at what Dell is doing with the XPS line to have an understanding: they are so desperate to get rid of their Gaming computers that they have to give fabulous pricing/loan terms for customers to purchase them. This is so extreme that I wonder if it is not just a Dell book-cooking scheme to hide losses and weak sales. So, the multiplier may be another 2, that is, a 6:1 ratio in Gaming computers. But gamers are the makert tracers: The gamers are the people other people ask what to buy. So, if gamers keep buying AMD, recommendations will keep flowing to AMD.

b) Servers: It doesn't matter how cheap, Intel Xeons have actually **NEGATIVE** price (Intel is the one who has to pay to sell them, not the customer). Besides, Server Customers are buying today their upgrade path for 64 bits, multicores, virtualization, etc. Intel is not even claiming it is going to be competitive there anytime soon. I won't increase the multiplier here, but understand that I am being extremely conservative.

c) There remain Desktops and Laptops: Desktops: Tracers will keep biasing the market toward AMD, look at how deep AMD has penetrated the computer enthusiast spectre, since 2003, 2004, 2005, thus, the factor could be higher than just the base of 3. But that's not all, take Vista into account, and shitty EMT64s become inappropriate, Video editing. So, in desktops the ratio could be 1.5*3 = 4.5 : 1 or 9:2

Laptops: With a relationship of 80% to 20%, it has a base of 4:1, but what is Intel going to accomplish by dimishing AMD's market share? AMD itself is caring not all that much about this segment, that's perhaps why they don't have any dual core Turion, while they have dual cores in all other segments. But there is a problem: There is resilient demand for Turions. Some people such as me actually care about the 64bits, it doesn't matter what Intel does in this regard. So, the multiplier for Laptops is again 4.5:1, 9:2 just as desktops.

3) AMD's growth has a momentum, that is, current partners and key customers are increasing demand, so Intel's price war won't actually decrease it, just its growth; in other terms, demand is strong. Include another 1.3 multiplier.

4) Think again what will happen if Intel cuts a 25% their prices: It will automatically become in the minds of consumers, customers, partners and analysts a confessed inferior provider of chips, and will legitimize AMD as undisputed leader. Then, this effect will push forward AMD's marketing (to have the leviathan of the semiconductor business desperate slashing prices is excellent advertising), and will give AMD a "cool underdoggie" aura. So, the multiplier is higher, let's say another 1.2 to the numbers already mentioned (4.5*1.3*1.2 = 7).

5) Since AMD already has about the same profit margins as Intel, it would be Intel the one that goes into losses first than AMD, and being AMD so historically thrifty, with a not so distant encounter with the possibility of Bankruptcy, I would give it better chances to survive a price war than the spoiled kid of Intel.

A long etcetera, but this is the core of my argument:

Intel would be not earning (or virtually expending) 7 dollars to prevent AMD to earn 1 dollar, that demonstrates the confidence that Intel has in AMD as a threat; Said in another way, for Intel, AMD's earnings have an strategic value SEVEN TIMES GREATER than its face value. By the way, a price war would distinguish Intel as much more confident in AMD than most longs.

Expressed into the terminology of publicly traded companies, Intel's current share price to earnings per share ratio of 14.5:1 means that, for investors, every dollar Intel earns is worth $14.5 per share. Since every 7 dollars in Intel earnings are worth one dollar less in AMD, then according to Intel, the proper P/E ratio for AMD would be 101:1; so, using yahoo finance numbers for AMD, trailing P/E of 100.3 and forward P/E of 21.36, the valuation for AMD would be 101/100 (trailing) * $40 ~= $40; completely justifying current market valuation, or forward 101/21.36 * 40 = $189 (!!)