The Evangelist

For many consumers, it can seem like it is you against the technology. Indeed, the very technology that is supposed to make our lives easier can seem to complicate things. This is especially true when new versions of software are released that, while they may have more features, typically introduce interface changes that may confuse users and can also have bugs in all of those new features.

This is especially relevant at the moment, with the recent release of Apple's Mac OS X Snow Leopard and Microsoft's impending release of Windows 7.

Despite the fact that Apple's Mac operating systems are typically known for their reliability and easy of use, Snow Leopard has not been without its problems. Most notable is a recently discovered bug that could cause the administrator account's data to be deleted. This, however, requires some unique circumstances and isn't exactly easily reproduced unless you are really trying to cause it. On the other hand, there have been compatibility problems, which has far greater potential to affect the experience of users. In addition, Snow Leopard does not come with Rosetta installed by default. While this results in a smaller default installation (as Apple has touted), it also means that the user has to install Rosetta before they can use any of their old PowerPC applications.

Also affected by Snow Leopard are those users who can't upgrade to Snow Leopard: those still using PowerPC Macs. Some of these users feel as if they are being "left behind" by Apple. While Windows users could theoretically install Windows 7 even on a relatively old PC (it has been demonstrated running on Pentium II PCs, although I wouldn't recommend using such a system for day-to-day work), even Mac users who bought PowerPC computers more recently cannot upgrade. It should be noted, however, that more and more applications are Intel-only, so Apple is not alone in this move.


On the other hand, Windows 7 presents its own unique problems. Although Windows 7 seems to at least be faster and more responsive than the much-derided Windows Vista, the difference doesn't appear to be by that great. Worse yet, Windows XP users will still face some of the incompatibility problems they experienced (or would have experienced) with Vista.

Users coming straight from Windows XP may still have problems finding some things due to the differences in the way things like the control panel are set up. While I feel that Windows 7's way of doing things is actually better, those who are used to doing things the Windows XP way will undoubtedly have problems.

The Windows 7 taskbar definitely has the potential to confuse users. While it is possible to switch to the classic taskbar, quite a few inexperienced users may not realize this is even possible, and may be thoroughly annoyed by the new feature.

Another problem is that while Windows XP Mode can help on the software compatibility side, the processor must have hardware support for virtualization in order for it to work. This has the potential to confuse the consumer even more considering the fact that Intel's budget processors typically do not have support for hardware virtualization, and support for virtualizating even among distinct product families can be inconsistent. AMD has been far more consistent, which could prove the be beneficial to the company. If retailers fail to make it clear to users which PCs are XP Mode compatible and which ones aren't, many users with old software could be especially angered when they find out that their new (or existing) computer with a copy of Windows 7 that supposedly supports the feature can't run it.


Of course, what many users are failing to realize is that Windows XP had the same kinds of problems when it was first released. It was criticized for its instability as well as its incompatibility with many old DOS applications as well as some older Windows apps. Of course, Windows XP improved greatly with service pack releases, and XP is currently the most widely used version of Windows.

Of course, there is no real solution to these problems. If new operating systems didn't include any new features, the companies that produce them wouldn't be able to sell them using their current business models. And, when you think about it, the thought of not having moved on from Windows 9x to Windows XP is absurd now. Somewhat ironically, the time between the release of Windows 95 and the release of Windows XP was a bit over 6 years, while the time between the release of XP and the upcoming release of Windows 7 will have been 8 years. While this is a testament to the quality of Windows XP, only so much life can be had out of an operating system in the consumer market. On the other hand, the lifecycles of some products are ridiculous. It's not Microsoft's fault that hardware vendors don't bother to create drivers for newer operating systems. It's also not Microsoft's fault that many of these drivers are subpar, and can be the cause of "Windows" instability; consider that if Microsoft were to enforce stricter standards, even fewer device makers might create updated drivers for new operating systems. Apple only avoids these problems because of the limited set of hardware on which OS X runs.

There are, of course, some ways to alleviate some of these problems.

One interesting possibility would be a subscription-based operating system. This could offer vendors a way to not force people to buy a completely new operating system every couple of years while still having a valid business plan. Newer elements could be modularized and added to compatible systems in the future, while underlying modules (such as those relating to drivers) could remain the same if compatibility would be broken by an update. This, of course, would introduce its own problems. Not everybody has an internet connection, and those who don't would likely be forced to call in. At any rate, it would likely require some time to renew the subscription. It could also anger consumers and make them feel as if they don't really have control over their computers.

As far as processor virtualization support goes, the key would be more consistency across product lines and at least being consistent when using such a feature (or a lack thereof) in order to segment the market. Intel has announced that hardware virtualization will be added to some additional processors.

In the end, consumers can do a few things to keep their experience is positive as possible. Before buying a product, check for reviews online. For example, go to Google and enter the name of the product you are thinking about buying and "review". It is even better if you can enter the specific model of the product that you are looking at buying, to see if it might be an older version and/or might have problems not present in other iterations of the same product. Also, proprietary formats, particularly connectors, can be a headache if there is not enough industry investment into them. Look into these technologies carefully, and be wary that they might not work with every device, and if they break, you might not be able to replace them easily.

Finally, don't forget one important fact: there are people out there that would be happy to help you, whether you are a beginner or seasoned veteran in the tech world. Chances are that there is a forum for your product (or type of product) where you can get help from others. Doing so can make what would otherwise be a major problem into a learning experience, and you just might find that the technology isn't so unfriendly after all.

You might remember that years ago AMD held the performance crown with their Athlon 64 processors. Intel struggled to catch up, adding 64-bit capability to their processors. The Netburst microarchitecture (Pentium 4) proved to be unable to catch up with AMD's innovation no matter how many gigahertz Intel put into their high-end processors; therefore, Intel turned to their older P6 microarchitecture (Pentium 3) to create the Core microarchitecture, which has since allowed Intel to reclaim its performance crown.

Even with the performance crown, AMD faced the fact that Intel was a marketing giant. While many enthusiasts, gamers, and other demanding users turned to AMD, Intel still held the mainstream market due to a number of factors. Most manufacturers sold mostly Intel computers. Some did not sell AMD computers at all. The public perception of Intel was also a large factor; many recognized Intel as the processor manufacturer, and looked to buy an Intel computer. I would like to note that one time, I talked to a user that was unsure of whether or not all of their software would be compatible with an AMD processor because they had always bought Intel.

Now that Intel has reclaimed the performance crown, AMD has had to compete in other ways. One way they have successfully done this is by offering exceptional value through lower prices. Of course, they have not been sitting around; AMD has forged ahead with their Phenom and, more recently, Phenom II processors. At the same time, Intel has come out with their Nehalem microarchitecture. Based on this microarchitecture are Intel's new 45nm Xeon server processors as well as the Core i7 and Core i7 Extreme processors for the high-end desktop market as well as the first of Intel's Core i5 processors, which will aimed more at mainstream computing. Eventually Intel will switch over to a 32nm process. he high-end processor for servers and desktops will then be Core i9, a six-core processor. At the lower end will be the Core i3 and a Pentium-branded processor.


So, the question becomes this: how will AMD continue to compete? It is fairly clear that Intel is also trying to get into the lower-end market with the Core i3 and their Pentium branded processors. If Intel is able to keep prices low enough and performance good enough on these lower-end chips, will AMD be able to compete? This is a complicated question, and there is no easy answer.

AMD's Phenom II processors are based on 45nm technology, and have improved performance from the original Phenoms. Benchmarks, however, show that high-end Phenom IIs still lag behind Intel's high-end Core i7 chips. While the Phenom IIs might cost less, those who demand performance at the high-end are bound to turn to Intel. AMD does, however, have quite a bit of potential at the mid-range to lower end. The Athlon II X2 and Phenom II X2 chips appear to be very good for the price according to benchmarks.

The other big question is this: where does AMD go next? It would appear that the next step is still AMD Fusion; these processors would function as both a CPU and GPU, eliminating the need for a separate graphics card. Intel will be exploring similar technology with some of their upcoming processors with the codenames "Clarkdale" and "Arrandale." However, given Intel's history with GPUs and the fact that these are intended to be mainstream-market processors, these processors won't be ideal for gaming or more intense graphics. One interesting thought is that, just as in the current Macbook Pro laptops with graphics that can be switched from dedicated to integrated to save power, perhaps computers with this technology could also an include a dedicated graphics card and be able to switch back and forth depending on performance needs or lack thereof at a given moment.

Since AMD now has ATI's resources, their Fusion processors could possibly deliver better graphics than Intel's will, but at this point this is just pure conjecture. According to Wikipedia's article on AMD Fusion, these processors may skip the 32nm process and instead be based on a 22nm process. This could give AMD a chance to get ahead once again, although there is no guarantee that AMD's processors will have superior performance simply because of the smaller process. Only the future will tell the fate of the two major x86 processor manufacturers, both of which face the challenges of the economic recession.

-

Opinions and comments are welcome. To post, simply click on the title of this article and scroll down.

Intel has put out a press release on the future of their Atom processor platform.

Two new Atom processors were announced at the Intel Developer Forum, which was held in Beijing. One of these processors, the Z515, will be made for smaller mobile internet devices (which will likely be very popular in markets such as China). It will be able to run at up to 1.2GHz thanks to Intel's Burst Technology Performance Technology, which is a definite improvement over the previous generation of processors for this form-factor. The other new processor, the Z550, will break the 2GHz mark for the Atom processor, and do it within the less than 3-watt envelope. No doubt this will present AMD's Neo with some competition in the higher-end market.

Along with this, Intel will release their new Moorestown platform, which should reduce idle power consumption by ten times, a great improvement. It would seem that Intel will beat AMD to their own motto of "The Future is Fusion." Moorestown will feature System on Chip technology (which was codenamed Lincroft). This will mean that the processor, graphics, and other controllers will be integrated into a single package, which should be great for the next generation of small form-factor mobile internet devices.

The Evangelist To Go - Download Article Audio
To comment on this post please click on the name of the article at the top of this post and scroll down.

If you have been following this blog or technology news in general, you probably remember reading that AMD was not planning on competing with Intel's Atom because they perceived the market to be very small. Even whilst this statement was being made, Intel was making a killing off of all of the netbooks being sold with Intel Atom processors.

It came as no surprise, then, when AMD decided that they wanted a slice of the pie (as mentioned in Meet The Atom's Competition). AMD had said that their new processor, now known as the Athlon Neo (formerly known by its codename Huron), would target the market for portables primarily between netbooks and laptops. This seemed a tad odd, considering the fact that most people who are in the market for a netbook only want a small secondary machine that does the basics and surfs the internet, and all for a fairly low price. They can't to do as much as notebooks, but they aren't meant to; it seems as if AMD is missing this. The Neo is meant to have more power, but this will only result in a larger end product, which basically contradicts the original idea behind the netbook in the first place. Sure, they probably have a niche market in those who want a decent amount of power in a smaller package at a reasonable price. But this market is small; for those who really need power in an ultraportable package there are products like the Macbook Air, although the Air has a far more considerable price tag.


The first 'notebook' to contain the Neo will be the HP Pavilion dv2, which will feature a 12.1" screen, hard drive capacity of up to 500GB, up to 4GB of RAM, the option of ATI Mobility Radeon HD 3410 discrete graphics (which makes it seem similar to the Atom-equipped Asus N10 with the GeForce 9300M Graphics Processing Unit), an expresscard slot, and an optional external Blu-Ray drive with an Athlon Neo MV-40 processor at the heart of the operation. The projected base price is $699, and it is due to be released this April. The Neo MV-40 is, unlike the Atom (which is currently based on a 45nm process), based on a 65nm process, which is no different than AMD's current processors, so these processors will run hotter than the Atom and draw more power, resulting in lower battery life (due to the 34W thermal envelope). It is single-core, has a clock of 1.60GHz, 512KB of L2 cache, and 64-bit support (which the Atom does not have). The laptop will also use the AMD Yukon platform. As far as size, the notebook will weigh just 3.8 pounds for a standard configuration and be only .93 inches thick. While this means that the Pavilion dv2 will be very small for a standard laptop, it is by no means as portable as the typical netbook.


AMD has revealed that they have plans to release dual-core Neo processors as part of their future Congo platform. It is clear from benchmarks that AMD's Athlon Neo is more powerful than the single-core Intel Atom in most cases and the dual-core Atom in some cases. When a dual-core Neo chip comes out it will probably beat out the Atom as well.


What will be really interesting to see is whether or not Intel takes this threat seriously. An article over at Cnet suggests that they may be cooking up a response to the Neo for the same target market based on the Core architecture. If they are, these processors would have less power than ULV Core 2 Duo processors which are used in expensive ultraportables and the Intel Atom, which is exactly the same market that the Neo is targeting. At any rate, I am sure that Intel will not stand still.

There is one final thing to consider when it comes to the Athlon Neo. Current economic conditions may actually help with Neo become sucessful. When it comes time for people to but a new notebook, current economic conditions will likely dictate that they purchase on a budget. At the same time, these people will be looking not for a netbook but for a regular laptop that can run all of their programs (and perhaps some games). This is where the Neo can possibly thrive. A number of factors will have to be met in order for this to happen, however. First, the manufacturers will have to be willing to take a risk in manufacturing these notebooks. Second, they stores will have to be willing to take a risk in carrying them. Then, finally, you have all of the factors that will contribute to people actually buying them, such as marketing materials and the attitude of those who work at the stores that sell the notebooks toward notebooks based on Neo processors.

At any rate, even if AMD and Intel aren't competing in the exact same market, they are competing in two close degrees of a larger market. If AMD hits the right spot in the market and the conditions are right, they could be quite sucessful with the Neo. On the other hand, if they are unlucky, their ranges of different products aimed at this small market will only serve to confuse and frustrate consumers. I really do want to see AMD succeed. They just have to be very careful, because every time they do come up with a sucessful product, Intel is a step behind with its massive Research and Development machine. And if one thing is for certain, it's that AMD has far more to loose than Intel.

You have probably already heard of Intel's Atom microprocessor, which is designed specifically for use in small devices (if you haven't, check out my article here). Intel has now expanded on the Atom with a dual-core version that sips more power but offers more processing power. While the Atom isn't as powerful as the old Pentium D dual-core desktop processors of the past, it is certainly getting close, and Atom processors are clearly more powerful than the old Pentium 4M processors that used to be found in upper-end laptops. Of course, they sip far less power and are far smaller than the Pentium D, which was made for high-end desktops; a truly impressive feat, but that's the way technology evolves.

The Atom isn't without competition, however; VIA, another company involved in the x86 (Intel-compatible) processor market has a competing product. VIA has been producing low-power chips for applications such as Tablet PCs for years, and so their producing a competitor to the Atom just makes sense.

The VIA Nano simply hasn't gotten the attention it deserves. While the Atom does not support x64 and therefore 64-bit applications, the Nano does. While this may seem minor, Microsoft has stated that Windows 7, the next release of Windows, will primarily come in 64-bit editions, and 32-bit editions, while still available, will probably take a backseat and be mostly for those who want to upgrade their older machines (many OEMs like Hewlett Packard have already started installing 64-bit editions of Vista as standard on their new PCs despite compatability worries). Also, Windows 7 may be the last version of Windows to support processors without x64 extensions. Many believe that Microsoft will take advantage of the quickly-growing netbook market with Windows 7. This may make the company change their plans and continue active support for processors without 64-bit extensions. Microsoft has not (as of yet) been very active in getting a product out in the netbook market; they have only released the stripped-down version of Windows XP Home for netbooks and Linux has done well in the market. Windows 7 with its MinWin kernel promises to at least be better for netbook applications by taking up less memory and using memory more effectively on lower-end hardware. It will certainly be interesting to see how this plays out. It seems likely that Intel will add 64-bit extensions to their Atom processors. The VIA Nano, however, offers that already.

The Nano was launched this year on May 29. It is a single-core processor based on either 65nm or 45nm technology and with a front-side bus (FSB) speed from 800 MT/s to 1333 MT/s. They have support for Error-Correcting Code (ECC) as well as hardware virtualization technology (based on Intel's implementation), which Intel usually supports only on their higher-end Core 2 chips. They also have a 128KB L1 cache and a 1MB L2 cache. If you are wondering about multimedia program comaptibility, the Nano supports the SSE, SSE2, SSE3, SSSE3, and MMX instruction sets.

A dual-core version is on track to be released in 2010. This presents an interesting parallel between current netbook chips and laptop chips of the past. Intel had released the original 32-bit Core Duo laptop processors, while AMD had released single-core Turion processors with 64-bit support. Here, we see that Intel has a 32-bit netbook processor while VIA has a 64-bit chip. The Core Duo processors outperformed the Turion processors, but the Turions were future-proofed. Of course, Intel is a marketing machine. AMD stood little chance, and VIA doesn't even have as many marketing resources at their disposal as AMD. This will likely result in the Nano being used in a smaller number of devices, which might mean that Intel will face no real competition. It's really up to the manufacturers now. Either they choose to put the Nano in their products, recognizing its merits, or they don't. To some degree, it is also up to the customer. Some will likely see "VIA" and read "not Intel so probably won't run my programs." Sure, there may be somebody there to explain, but old habits are hard to break. Even AMD has a difficult time because people don't always understand that their products are more like Intel's offerings then not.

I am afraid that the Nano will not attain much marketshare. The chip may have other effects on the market, however; in order to stay ahead, Intel may add 64-bit support and other features to their Atom processors. Competition can encourage innovation.

The Nano won't be the Atom's only competition; AMD will be entering the market, or so they say. While Intel was producing Atom chips and selling them like hotcakes, AMD was telling us that they didn't think that there was a market in netbooks and nettops and that they weren't willing to risk the investment. AMD is finally coming around, however, and their latest roadmaps indicate that they are now working on a chip that is meant for higher-end applications than the Atom but will be in the netbook and nettop market. The chip also promises to be cheaper than the Atom, but I have a feeling that Intel will try their best to ensure their dominance of the netbook market, and Intel has plenty of time to respond now that AMD has announced the product. That's the price they have to pay for reassuring their shareholders. I really do hope that AMD succeeds with their prouduct.

Here's to hoping that innovation triumphs.

If you are into tech or browse news sites such as Slashdot, you have probably heard of Intel's new Atom Processor. And you probably know that the processor is meant for use in ultra-mobile PCs and other mobile devices, even smartphones.



You might be thinking, "these have existed in the past without the Atom processor, how are things going to be any different?" If so, you are right that both AMD and Intel have had mobile offerings available for a while. However, in these cases the lowest-power chips available from both were used for an application they really weren't originally made for. Also, because they consumed little power and were so small, they were generally underpowered, which made these UMPCs slow. And then there are "Netbooks" and other small laptops that relied on processors like low-powered Intel Pentium-M and, more recently, Core 2 chips to keep their small form factor. And finally there are even smaller devices, such as Smartphones, that use processors not compatible with the x86 instruction set that is standard in Intel and AMD processors which run Windows and standard Linux distributions. One example of this is Nvidia's Tegra, which will supposedly have superior performance to Atom, but will not support the x86 instruction set, and instead be based on the ARM RISC architecture.

This is where Atom comes in. Intel's Atom has a maximum CPU clock of a decent 1.87 GHz (with the lowest clock being 800 MHz), and supports the x86 instruction set. While AMD's Geode processors have supported Windows operating systems and standard software, they are underpowered compared to Intel's Atom.

Now, with Atom, even a smartphone could run the standard applications end-users have gotten used to, not just special versions adapted for a mobile platform, often with less functionality. And there is software that is not available at all for current mobile platforms. Two good examples would be Flash and Java, which can be important parts of the web browsing experience. With Intel Atom, both Flash and Java could run on Windows or Linux on a UMPC or MID (Mobile Internet Device).

Hopefully manufacturers will realize the potential here, and that the companies selling these devices properly market these devices, making sure that consumers know that these devices can do what other mobile devices cannot.

At the same time, I hope that innovation marches on, and that AMD comes up with their own mobile solution to compete with Intel.

In the meanwhile, Intel has unveiled their Centrino 2 Platform. Highlights include an 802.11n-compatible Wireless card.