Storage Rising

The need for storage is an ever-increasing one. As soon as we move from 100GB hard drives to 500GB drives there arises new or improve applications and media formats to fill them up. Think of the move from people hearing MP3 on their PCs to them watching DVD ripped movies. And now the move to High Definition movies will keep the need for storage going for some time, not counting the possible applications that will emerge that now are unconceivable because the lack of capacity.

Of the many storage types Magnetic Hard Drives are specially interesting. As many people prognosticated the demise of this medium at the hands of NAND flash based storage, what we have actually seen is magnetic storage entering new markets besides the PC, mainly the consumer electronics market, expanding rapidly. (for a recent example look at Seagate, the biggest hard drive maker, has just raised its forecast of revenue)

The consumer electronics market include devices like iPods, Cable set top boxes, and video cameras. I am now buying a digital camcorder that has a bigger hard drive than a notebook I bought 5 years ago and I can still remember fondly my first PC's hard drive with a whopping 40MB of capacity.

It is true that flash based storage like pure SSD (solid state disks) or hybrids (part flash, part traditional magnetic drive) are increasing their capacity and falling in price (checkout Samsung's forecast that NAND memory will surpass DRAM memory quite soon). But still spinning disk magnetic drives hold (and probably will hold) a price per gigabyte advantage of at least an order of magnitude for some time to come.

Seagate and Western Digital are well positioned on the Hard Drive front and are pureplays fore this sector. Though Samsung has a competitive advantage in the form of being the biggest NAND memory maker as well as being a magnetic Hard Drive producer.

Who will win, magnetic storage or flash based memory or maybe something entirely new?
In my next post I'll explore emerging technologies that could replace both incumbents.

Google purchases Peakstream (Multiprocessor software company)

Google has bought Peakstream as reported by the register.
Peakstream is a company that focuses on making software to exploit to the maximum capability the resources provided by multicore CPUs and principally GPUs.

I talked about stream processing before in this article.
Peakstream solutions where aimed toward the HPC (High Performance Computing) market, like the pharmaceutical, financial and oil business but it was actually acquired by probably the only company which has just as deep pockets and maybe a greater need for computing power, Google.

This graph illustrates peakstream concept, to bridge diverse parallel computing resources to offer solutions to high demand spaces.

If Google prepares to use GPUs along its servers there could be two winners, mainly Nvidia and AMD, tha latter already sells its GPU as stream processors.
Peakstream software can also further enable the Cell chip to be used in a more diverse capacity.

Though Google undoubtedly will use the technology to augment its server and internal processing capacity as its needs for this are great (Google which makes its own servers prefers to have this capabilities in-house). It may just as well provide solutions based on Peakstream to the sectors I mentioned above. This possible solutions not only include complete products but services as well, like processing on demand.

The only major player left in the space Peakstream played is Rapidmind.
Which could become more valuable in the future if Google pushes de adoption of GPGPU.

I think this developments will have a tremendous impact in the near future.

Foundries, (When you can't do it yourself, delegate)

First of all, I'm sorry for the lack of updates i will try to keep up better in the future.

I want to talk a little about one of the major moves in the semiconductor industry. Mainly the move from the in-house production of chips to a 3rd party manufacturer, called foundries.
Foundries are basically fabrication facilities for hire. Say a company designs a chip but it doesn't have sufficient means to bring it to silicon. Be it because of a simple lack of a fabrication facility (called fab) or because of the need for a more sophisticated method of fabrication (65nm vs 90nm, etc) or maybe even because demand exceeds the in-house manufactured supply. The solution?

Enter foundries.
The foundry-designer relationship is as important in the chips world as division of labor is to traditional manufacturing. Not only it increases efficiency of production by having specialized teams working in their specific areas of expertize allowing to allocate most of the designer company's resources on chip R&D. It also splits risks in case of failure of the venture.

There is also another reason for moving to a foundry for production ,that is arguably more important than the advantages of production, and that is the lack of an alternative.
With each advancement in manufacturing technology (65nm to 45nm nodes; 200mm to 300mm wafers) costs for both R&D of the processes and the cost to actually build a fab spiral out of reach for many (read most) companies.
Fabs for 45nm costs in the vicinity of 3 billion dollars.

This leaves two, not necessarily exclusive, alternatives for companies, to partner up a and split R&D costs or use foundries.

As an example of this, take AMD, it has two fabs which use R&D developed jointly with IBM while it also uses Chartered as a foundry for CPU and it is also the biggest customer of TSMC (the biggest foundry around).
In contrast with Intel, that with it's huge resources is among the few companies that can go at it alone.

This pictures illustrates the relationships

The benefits are increased manufacturing flexibility, the trade off is decreased margins.

Many companies which traditionally manufactured in-house are moving to foundries.
This is a quote from CFO of Texas Instruments, Michael Masdea,

"...we have been working with the foundries now for 5-plus years, and really developed some good relationships and gotten pretty comfortable with how they operate.

During that period, we continued to develop our own advanced digital manufacturing processes, as did the foundries develop theirs. Watching their development capability over the last few years, we have gained a lot of confidence in their ability to develop those processes in a manner that fully meets our needs. So as a result of that, we no longer believe that it is necessary to maintain this redundant capability, if you will, by developing it internally and having them develop also.

Consequently, we are going to go ahead and discontinue the development effort just on the advanced digital processes."

As you can see, it seems that foundries have reached a maturity level that allows a complete delegation of manufacturing to them.

It seems that Sony in an attempt to decrease expenses will move to manufacture the Cell chip at foundries when it moves to 45nm on the 2008-09 timeframe.

Now for my take on this. I think Foundries will grow continually in the future as they become largely the only option for new companies the make chips, and in many cases the prefered manufactures of large already established companies. The mayor pure play foundries are

TSMC (Taiwan Semiconductor Manufacturing Company, Limited) (Stock:TSM)
UMC (United Microelectronics Corporation) (Stock:UMC)
Chartered Semiconductor Manufacturing (Stock:CHRT)

Of these stock TSMC also carries a decent 3,3% dividend yield.

More to come later on this topic.
Cheers.

AMD will announce new notebook Chipset and CPU, Puma and Griffin on 05/18/07

I`ve come across some interesting information. On may 18 of this year (this friday) AMD will announce a new Chipset know as Puma and a new CPU know as Griffin aimed at the notebook market.
The new products are aimed toward the higher-end notebook market.

The new CPU have improved power management which includes split power plane for the CPU's cores (each core can adjust its own voltage independently of the other).

What will be more interesting the Puma chipset.
Courtesy of the ATI acquisition this new chipset will support Microsoft's DirectX10 as well as H.264 hardware accelerated decode. It supports both HDMI and DisplayPort outputs. I also sports 5GHz PCI Express port.

Intel's roadmap don't include many of this features until 2008, specially 5GHz PCI Express.

I believe AMD stock will probably will go up another notch on the news following the recent run of last week.

News development

Every once in a while I will post some links about news developments, like company announcements or articles from different websites, that I use as sources to base my analysis upon.
These news in part substantiate my theories (or at least I hope they do) but the consequences are open to interpretation so if anyone differs from my take on them I welcome criticism.

• Microsoft Releases free VirtualPC 2007. Related to my previous article on virtualization. I think that the fact that it is a free release demonstrates in part the importance for Microsoft to make the Windows platform the default choice for virtualization.

• TSMC, a chip foundry (company that manufactures chips based for other companies that design them) may be tapped in the future as a source for AMD Fusion chips.I will talk about the importance of foundries in my next post.
  

• Intel is building a fourth 45nm capable fab. A while ago I talked about the concerns of investors about the overcapacity that AMD and Intel are building. While not proof tahat there won't be an overcapacity issue in the future I think that this shows that the companies that are actually betting the farm on the issue are optimistic that the capacity won't go to waste.
  

• AMD releases motherboard chipset with ATI integrated graphics and with HDCP (content protection for digital high def content) and HDMI (joint audio and video output) This is a coup for AMD into the enterprise segment as it is a low cost business solution, ,but the HDMI output it's specially important as it allows PCs built upon it to be use as media center PCs. Furthering the convergence. The chip also handles video processing on hardware.

• Also AMD demonstrates stream computing based on their next generation R600 GPUs, And Nvidia released CUDA a C compiler that enables code to run on their G80 GPUs as well as future ones. I talked about related things on my previous post.Here are two articles from Beyond3d that go into detail of what CUDA is about as well as a little overview of TCM (a similar counterpart from ATI). Quick overview, and a little more detail.
  

Another Digital Convergence Post

The fact that digital devices such as cellphones, TVs, and such have steadily grown in features and in interoperability is nothing new, but the fact that this devices have found their way into the mainstream market is a lot more recent.

The multimedia capability of the PC has been touted for many years now, but cellphones and digital cameras along with Ipods are the true culprits of bringing the PC to a more massive and diverse public for which this capability could be shown.
The Apple Ipod alone as the flagship of all digital music players has brought a large influx of people who now use the PC (or Mac) as the proverbial entertainment hub to store and listen to music. This has made people more adept and comfortable to use PCs and mainly the internet. The same goes for Digital cameras and for cellphones as the interoperability between this devices and the PC has not only made the PC ever more popular but also now more people are taking photographs with digital cameras than ever before with conventional photography.

This brings me to pronounce the gadgetization of the PC. Many people now buy Macs or PCs with only this capabilities in mind, much more like they would by an stereo system or a DVD player. Granted they are buying far more capacity than these, but the initial focus is on the multimedia aspect. These creates a new kind of user, one that isn't as technically proficient as the ones who enjoy the technological side of the PC but one that isn't totally computer illiterate either.

Just as the PC has assimilated audio, in the next step it will assimilate video. And just like in audio the bricks for this road has already been laid down by the peer to peer scene, now all it rests is the further advancement of the Video Ipod which just took advantage of the technology which was already tested and in place. The other pawns in this battle are the media streaming devices. Also know as PCs on disguise.

This media streaming devices which are no less than PCs with Tivo like capacity with more aesthetic enclosure than the traditional beige box are already on the market, but Apple with its own ways may provide the definitive one with the Apple TV. This things are no more than PC trojan horses, indistinguishable from PCs on nothing more than software and aesthetics.

To see how this interoperability is so important one must look at Microsoft and its Xbox 360. With every sale of an Xbox Microsoft takes a loss as the hardware is more expensive to manufacture than the sales price. The hope of this business practice is that eventually game licenses (Microsoft get paid for every third party title released on the Xbox 360) will offset this lost. Also the manufacturing costs will get cheaper, the rumor is they are at break even point now.
But the true focus of the Xbox360 is not on the games only. It´s on the online capabilities, the social aspect, the future interoperability with Windows and of course the Microsoft Marketplace, a place in which you can buy Movies, Tv shows, Music (read Zune).
Now for true PC dominance you not only need to provide for the core PC components but also for the many accompanied gadgets that orbit the PC. If you don´t do it, a competitor will.
Microsoft and Apple are in a unique position to do this, and I think they will both will be able to grow immensely.

Sony is another company that has the resources but they don´t have the necessary PC presence. They are doing a more traditional sort of media invasion with the PS3 as a trojan horse for Blu Ray discs but not true functionality between platforms, at least not yet.

As always I think Apple and Microsoft are the standard bearing companies in this battle but lets not forget the companies that will supply the hardware, Intel, AMD, IBM, Nvidia. The expansion of the PC architecture into so many fields will give the economies of scale of the technology provided by this companies a huge boost. I will discuss more of this subject in the future.

Economically this will lead to acquisitions like Nvidia´s purchase of Portal Player (a supplier of chips for Apple´s Ipod), creation of new markets like Microsoft marketplace for online delivery of media as well as new functionality in otherwise traditional electric devices like TVs an of course the transformation of the PC as it becomes gadgetized.

Cheers.

Of streaming and GPGPUs

Just as in the last post I wrote about virtualization the concept of streaming (or stream processing) isn't new. On the other hand its approach to maturity and mainstream adoption is a lot more recent.
When I say "streaming" I talk about a specific area of computer science, one that is mainly based on applying a same set of functions and apply the same work to a large array of data. It is a type of computing that doesn't fit well every kind of data, it has basically to be data with very little interdependency so all the work can be done in parallel.
When the workload does fit well for a stream processor the gains in performance against a more general purpose processor are huge.

Examples of streaming processors are today´s GPUs specially the X1000 series by ATI and the 8000 series by Nvidia as well the upcoming X2000 series also from ATI. This processors began their life from a place were parallelization is paramount, computer graphics.
Computer graphics are a problem that is embarrassingly parallelizable, this processors began their life as more fixed function processor but as they evolved they became more programmable (the developers could program the functions they needed and not exclusively depend of the hardware predetermined functions) thus they became more like modern CPUs but without loosing their focus on how to be best effective on multiple workloads.

While this was happening on the GPU front, the CPU was gaining some capabilities in parallelization, remember MMX? or SSE instructions? which mean Streaming single instruction multiple data extension. Now that CPU and GPU have gained similar functionality they are well on their way to converge, a processor has already been announced for 2009.

But i´m getting ahead of myself. My main point is that GPU are already useful for streaming and process non-graphics related data and thus named GPGPUs or General programming on Graphics Processing Unit. An example is the Folding a Home project that uses in part the X1000 series of GPU to make calculations on molecular dynamics. If you check this link you can see how it takes to 500 GPUs to produce 20 TFLOS while it takes more than 150000 CPUs to produce 145 TFLOPS, a 40x favor to the GPU.
Data for simulations that Gas and Oil companies need to run are particularly suitable for GPUs. As well as sorting, a problem which Google would much appreciate the help of GPUs.
The margins for profit in this segments are huge and the development of hardware as well as software in this areas are needed at a premium.

Streaming processors will merge and coexist with current general purpose CPUs as each serves its niche well. Like in AMD´s torrenza initiative there is already progress in the form of coexistence of processing modules on the same platform. The flagship of all examples in this regard is the planned IBM supercomputer, which is to be the world fastest, it utilizes Opteron (general purpose processors) as well as Cell processors.
Cell is a processor that is already a mesh of streaming and general purpose as it has one PowerPC derived core as well as 8 other vector (or stream) processors. It was developed by IBM, Toshiba and Sony and it already powers Sony´s Playstation 3.
Here you can see the "power processor element"(general purpose) and the eight stream "SPE" processors


While GPU adoption is already on the rise and will continue to be, some may accredit Vista for lending a hand but it really is just part of the wave, the benefits for players like AMD, Nvidia and Intel will be huge, invest accordingly.

Virtualization is all too real

Virtualization in a nutshell is a technique which allows to run many operating systems or applications simultaneously in a way that fools them into thinking that they got all the hardware of the host machine to themselves. Why is this important? Because many computers, mainly servers, aren't run at full capacity all of the time. This means if you have a lot of servers running at half capacity not only you spent double the money you needed to buy the equipment but each month you are also paying extra in electricity and managing needs.

Virtualization helps to alleviate this problems in a clever way: you run a computer with one operating system (Linux, Windows, etc.) you now create many virtual machines with a special software and each of this machines can have one operating system, this operating system doesn't have to be the same one as the host (check out the picture at the top - An Windows XP PC running an instance of Windows98 in a virtual machine).
One virtual machine can run Linux, while another one runs WindowsXP and another one runs BSD. You don´t have compatibility problems because each OS thinks it´s in its own PC.
The advantages are many. By running multiple machines in one PC you utilize the PC´s hardware closer to its maximum potential.
You can have programs that aren't available for an OS running in another virtual machine with native compatibility.
The most important feature of all nonetheless is manageability. A network manager can have remote access to many PC, managing them without having to come in physical contact with them. And if a virtual server (that is a virtual machine running as a server) crashes you can simply restart the virtual machine, not having to restart the host PC. You can even move one virtual machine from a physical PC to another without stopping the virtual machine.
This saves money increasing the uptime of a server.

Why is this important in an investing sense?
While virtualization isn't a new technology its adoption is on the rise. Now and for the last year (2006) all x86 server CPUs from Intel and AMD have come with virtualization features in hardware, allowing the virtualization software to have closer access to said hardware.
The growth of virtualization software has been astonishing and now one company, EMC that owns VMware, the software that has the biggest marketshare in virtualization, is planning and IPO for approximately 10% of VMware.
I think that with the increased adoption of x86 servers and continued growth in the use of virtualization, this IPO, planned for June 2007, has a lot of potential.
There are competitors, like Microsoft´s VirtualPC, and an opensource project called Xen. But VMware is the current dominant player and the market is exploding. With revenues that grew 83% to 703 million in the last year I think it will be wise to keep an eye on EMC in the future.

PS: Here are a couple of links with a lists some pros and cons of virtualization.

"The most powerful force in the universe is compound interest" - Albert Einstein

The power of compound interest is not to be underestimated as expressed in the famous quote by Albert Einstein. One place in which this force clearly operates is in dividend paying stocks. Companies whose stock pay a periodic, often on a quarterly basis, amount of cash to its stock holders. The beauty of this payments is that many companies increase the payment each year (though there are companies that pay irregular dividends). In this yearly increase is where compound interest comes in.

Let me give you and example, oversimplified as it may be I think it gets the point I´m trying to make across.
Lets assume you start out with a thousand dollars of a stock that pays 50 dollars annually in dividends, this corresponds to a 5% dividend yield. Now lets say this company has increased the amount it pays on its dividend each year for the past 100 years (as 3M has done). It is thus not far fetched to assume it will continue to do so in the future. If the dividend growth at 10% a year this means the first year you collect 50$ dollars, the next 55$, the next 60,5$ ans so on. Stay like this for ten years and you collected 796,87$. This is the equivalent as if you had collected 79,68 $ per year for 10 years, or a constant dividend which is 60% higher than the original one (that is, instead of a 5% yield you now have an almost 8% yield) View the appendix at the end for a more general look at this.
As you see a growing dividend equates to higher effective constant dividend, and the longer you hang on, the higher is the effective dividend. If you had remained receiving dividends for 20 years, you´re effective yield would be 14,3%.

As you can see, there is a catch, time. For compound interest to be significant you need patience, as it is not a get rich quick scheme. This is important as there is no better time to start investing than now.And this is why I think that a company with constant but conservative growth that pays a dividend is a very important component for any portfolio.
I personally recommend 3M (MMM) stock which is a diversified company that pays a dividend which it has grown for the las 100 years. Also Bank of America (BAC) which pays and already high yield and I think it´ll grow more.

You can also increase the rate in which compound interest works if the company you´re investing in has a Dividend Reinvestment Program (or DRIP). In this program all the money that comes from dividends it´s used to buy more shares of the company automatically. This furthers feeds the amount of dividend you receive in each payment.


Appendix:
This formula represent what we just did on our example, but in general terms.
If you replace in this formula Dividend with the amount you receive the first year (50 dollars in our first example), Growth with the rate that the dividend grows each year (1,1 in our example as it grows 10% a year) and number_of_ years with the total number of year which you retain the money, the results give you the entire amount that you receive in all years due to dividend payments.
If you divide this amount by the number of years and the original amount of money yo get the effective constant yield.

Cheers.

Capacity and Manufacturing, it's not over.

On the last few earnings calls Q&As AMD has been questioned, and fairly so, about building overcapacity, this means if AMD is growing the capacity to manufacture chips at over 30% a year and Intel is also building more capacity, what stops this from creating an oversupply problem in which supply outgrows demand?

I see two solutions to this:
1. the x86 architecture is taking over more and more markets; markets that were once dominated by other architectures, this includes servers, embedded and even non-PC electronics. While the x86 architecture has been criticized for not been the most efficient architecture, it has one thing going for it, and this is economy of scale.
Just because there is so much production and consumption of this architecture it becomes a commodity, being cheaper to manufacture than any other architecture just because of the sheer quantity of production its able to muscle out any other kind of chip and definitely outgun it on R&D.
For examples look at RISC chips or Itanium. The x86 architecture has mutated integrating various concepts from this chips and it has remained the survivor.
Thus, if the market for PCs doesn't grow according to supply growth, there is still room to expand in other markets.

2. Intel has managed manufacturing in a very smart way. It keeps the most modern facilities producing the cutting edge chips and the older ones manufacturing chipsets and other kind of products that doesn't benefit as much from the more advance technology. Then they upgrade the facilities and keep going. Now AMD has the same possibility, with a new set of products to manufacture (from ATI: chipsets, graphics chips, etc) it comes much more flexibility. Now there are more things to build than fabs to build them with. This is why AMD also leverages independent fabs like Chartered for CPUs and TSM and UMC for GPUs. So even if AMD doesn't grow marketshare I don´t see overcapacity as a problem. Besides the capacity not only comes from more fabs but from a more efficient manufacturing line, using 65nm and then 45nm process (smaller chips) and 300mm instead of 200mm wafers (bigger silicon plates onto which the chips are edged) this is definitively a plus in cost reduction, and in does provide higher margins, not to mention it allows to produce more advanced chips that sell at higher prices .

PS: For what it´s worth Hector Ruiz CEO of AMD says that all capacity building is based on what AMD´s partners demand, current and projected.

3M the great diversificator

I was going to write a post regarding 3M and its prospects but I just recently read one by another blogger that more or less summarizes what I meant to say.

The only thing I would add is that I think 3M its a great way to diversify a portfolio as it is a company that participates in many fields. Also as much as 60% of 3M´s revenue comes from outside the US. this helps in a market with a weak dollar.

3M recently sold its pharmaceutics business and its on a bench acquiring companies close to its core competency. I think considering the current dividend yield and recent price drop on the stock its a great opportunity for buying it for the long term.

The Consumer Segment

The consumer Segment which includes electronic devices for entertainment such as consoles, digital TVs, cellphones, is growing at a very fast pace, much more so than the PC industry. When AMD acquired ATI, they also acquired a foot in this market, and a very strong position in the products I mentioned above. ATI had, and now AMD has the Imageon line of chips for handheld devices along with the Xilleon line for Digital TV. Not to mention they supply graphics chips for both the Xbox360 and the Nintendo Wii. Last year ATI had acquired Bitboys Oy, a developer of mobile graphics technology, that was partly owned by Nokia, the biggest cellphone manufacturer, this provided an opportunity for the cracking of Nokia as a mayor customer, though a joint development between ATI and Nokia has been announced the fruits of this relationship are yet to bear. This is a side of the equation that many analysts ignored in the AMD evaluation and I think it could be a mayor factor in the future.
I grant you that Nvidia it´s a great company and are also moving in the cellphone market. I think they will grow a lot here but I also think the market is big enough for both companies to grow, at this point in time at least.
The conversion to digital broadcast of TV signals will help with the adoption of Digital TVs and the penetration of the Xilleon chips.

My take on AMD

One of the main reasons i started investing on technology was because i saw great entry points to the stocks of the companies that i followed. This happens not as infrequently as you may think considering the volatility of the sector.

Two years ago i bought AMD stock, it went to great heights soon after, though i sold them on the way up (rookie mistake). Then they came down to the same level they were two years ago. This is amazing considering how much the company has grown and how much room for growth is still has.

The acquisition of ATI came as a great surprise for me, at the time i owned both ATI and AMD.
And this is where i think analysts are wrong, the acquisition is seen as a liability, having hit AMD with debt and consuming its cash. Neglecting all the possibilities that the deal opens is just asinine. AMD is seen as just as a competitor to Intel. Now it has its foot in a faster growing market than the PC market, namely the consumer segment. Providing chips for cellphones, digital TV and more.

As for the PC market there is a line of thought that goes like this, "AMD screw it, Core 2 Duo mops the floor with athlon and opteron and now AMD is paying the price for its complacency " i disagree with this and my reasons are:

• Processor split. Look at this graphic provided by intelwe can clearly see that Core2Duo processors accounted for little more that 20 percent of processor mix for Intel in the fourth quarter of 2006, and it will account for little less than forty percent for the first quarter of 2007, a fabulous ramp i admit, but... For the following affirmations i´ll assume that Core2Duo wins over Athlon X2 who wins over Pentium 4 and Pentium D (Netburst) though you can argue all you want if it is all that clear cut. Assuming this you get that finishing 2006 AMD´s processor mix was better that more than 75 percent (more close to 80) of Intel´s mix. And ending 2007 is still better than 35 percent of Intel´s mix (this is not counting the introduction of any new architecture by AMD, at least not in volume which we just have to wait and see). I know this is an oversimplification as it does not take price and the sectors in which the processors will end up (notebook, desktop, server). But i think it does debunk the fact that Intel has an absolute technical advantage over AMD.

• The good enough factor and the image cleanup. Only the so called "enthusiast" cares what microprocessor its under the hood. I care, maybe you care but there are far more people that don´t. This is where the big twist comes in. Most modern processors are good enough for most people, i have anecdotal evidence of this from my circle of friends and known people and i think its generally so. Many years ago (not that many now that i think about it) hardware vendors and OEM folks considered AMD a second tier supplier at best, that has changed since Opteron and many vendors now tap AMD as a second source. The funny thing is that on the desktop front the change in the perception of AMD is far more valuable than the change in the architecture of the microprocessor, as i think the Athlon XP processors also were "good enough" if not better than their Pentium counterparts. But now all mayor vendors (HP, IBM, and DELL) sell PCs with AMD and Intel processors in the same place. And i think most people will be hard pressed to find the difference between this machines when they see them advertised on the sales site, that is, other than the price tag. This "the poor man´s Intel" image has been broken and there is no coming back to it.
• The way i think of it, on the desktop and notebook side, the hardware vendors needed to be convinced as the majority of the costumers just didn't care. On the server side i think both vendors and costumers must be convinced, as all network managers are tech savvy. But tech savvyness doesn't preclude prejudice as i know more than a couple of network managers who still label AMD as unreliable. This image has changed drastically but not 100 percent but the heavy lifting has been done by opteron and now only time is needed for the rest.
  

Well this has been part of my little analysis, i know i haven´t said anything essentially new, i just recompiled some information that has been making its rounds on the internet. I will continue my rants through the week and beyond. So thanks for reading.

Oh, and before anyone labels me a fanboy. I have been building my own computers for several years now and most of them have been AMD, but i am building a new PC now and its a Core 2 Duo PC. Consumers and Investors have no loyalties. But as i have AMD stocks now i´m just putting my mouth where my money already is.

Introduction

Welcome to my blog. This is my first entry to the world of blogs.
The reason i started this blog is to provide my personal view, to the people who may bother to read it, on a few topics that interest me, mainly investing, the technology scene (computers and else) and the space in which they live together, ie. tech companies.

I am in no way a professional in either of this fields, but i follow both of them closely and invest a little money in a couple of companies (3M, AMD, and ATI before it was acquired by AMD).
All my knowledge is derived by independent research, the downside of this i think is a the propensity to make amateurish mistakes, but on the other hand it may give me some fresh perspective, who knows, you say.
One of the reasons i started this blog is to learn from the people who may read it and post a response, so i encourage posting on my posts.

Thanks for reading.