Usefulness of the CPU and platform

Is a more powerful CPU useful ?

Computer performance has quickly improved. However, the computer capabilities haven’t followed the same rapid progression. With the arrival of multi-core processors, performance will increase once more. But is it really useful? More and more functions are processed by chips other than the CPU, like the GPU, for example. Three-dimensional and video processing (encoding and decoding) requires a lot of resources and this task is being less and less associated with the CPU.

Video games still require powerful CPUs, this is true, but it’s important to put this in perspective. For example, Doom III requires a lot of CPU resources but the game outside of the benchmark is restricted to 60 FPS. A Pentium 4 Northwood 2.4 GHz is powerful enough to reach that limit. Faster processors are better in tests, but in practice they don’t add much.

Of course, there are other counter examples, but it’s important to notice that it’s time for manufacturers to find new CPU capabilities. As the technological leader, it’s Intel’s job to find justification in buying a $500 CPU. Intel is constantly on the lookout for new processing capacities to provide the right product, but they are also anticipating demand by finding new uses. The Santa Clara giant’s dream is to place an X86 chip in every car, article of clothing, surfboard, cell phone, TV, watch, etc.

New uses

Initially Intel wanted to promote the Digital Home and Digital Office. Not only a house or office with a computer it’s an integrated central system to our office and home. They want to increase the technology’s transparency and usefulness. All this is a little vague and seems to be a good strategy to avoid discussion of the Pentium 4’s average performance in non-professional setting. There is a lot of work to do in integrating computers in every day life, which will require more performance, flexibility, new “intelligent” software and an increased security that is more efficient and less restraining. These new programs’ algorithm will use processors in a different way, and it’s important that Intel understands future performance requirements in order for it to provide the right processor.

If you read between the lines, it’s possible to see that Intel is trying to justify the current domination of the Athlon 64 in terms of overall performance. AMD currently provides highly efficient processors, but in discussion with AMD, they showed no position regarding future technological evolutions. AMD doesn’t have the same resources as Intel for research, but they should start to think of their position in the future. AMD is currently improving their processors’ performance and reducing its electrical consumption to have an advantage on the performance level. But is this the right move? Shouldn’t they care more about the platform?

Regarding this subject, AMD told us that the platform was not one of their priorities, though with the Opteron they put a very powerful platform on the market. Why didn’t they do the same thing for the desktop and laptop market? The PCI Express still isn’t available for the Athlon 64 and AMD can’t change this situation because they have stopped chipset production. It could have been beneficial for AMD to take advantage of the gaming market and offer a platform corresponding to this processor. NVIDIA is going to release a chipset able to manage two PCI Express graphic ports, for example. But how much will the nForce 4 cost and when will it be available. Some motherboard manufacturers thought that the $75 price of the nForce4 chipset was a joke. NVIDIA decided to double everything including the price. AMD could have benefited from this situation if they had been in control of the platform.

The gaming market isn’t Intel’s priority. Intel identified other current demands and upcoming requirements for processing capacity and they intend to find solutions.

Solution : dual-cores

Solution : dual-cores for everyone

The dual core should be 2005’s leitmotiv at Intel! During the IDF, Intel made demonstrations and intrigued us with the Montecito, Smithfield and Yonah. These names correspond to:

- Montecito : Itanium dual-cores
- Smithfield : Pentium 4 dual-cores
- Yonah : Pentium M dual-cores

Montecito

The Montecito is a very ambitious CPU. The Montecito will raise the Itanium to the level of a technological marvel. Does it really happen in practice? It’s hard to tell. But on paper the Montecito is really impressive! The number of transistors reaches 1.72 billion. It’s useless to mention the yield, multi Gigahertz, die surface or fabrication costs.

Why so many transistors? Is it because the CPU has two cores? This is not exactly accurate, but partly true. Each core has its own cache L1, L2 and L3. These cores are rather impressive; 1256 Kb for the cache L2 and 12 MB for the cache L3 multiplied by 2. The Montecito integrates a total of 26.5 Mb of cache. With the redundant areas it represents more than 1.6 billion transistors. Compared to the Itanium 2, the new core benefits from several optimizations that Intel didn’t explain. The cache 2 has been split into Data Cache (256 Kb) and Instruction Cache (1024 Kb) like the cache 1. The Hyper Threading has been integrated into a slightly different version from the Pentium 4 in order to take into account the Itanium’s architecture. The Montecito has an advanced thread management system for optimal or at least not random use. Sometimes two threads will be processed by two cores and sometimes they will be processed faster on a single core via HyperThreading.


We witnessed an impressive demonstration using a system made of 4 Montecitos, but no information was revealed on the CPU frequencies. Intel announced, however, the CPU electrical consumption at around 100 watts, this despite the chip’s complexity. A power management system reduces voltage to below 100 Watts and therefore CPU frequency if the system overheats.

Smithfield

Intel put on display an avant-garde style desktop model with a dual-core CPU. Without clearly mentioning it, Intel insinuated that it was equipped with a slightly modified Pentium 4 dual-core Smithfield CPU.

In the demonstration opposing a Pentium 4, Pentium 4 HT and Pentium 4 dual-core in several video flows our interest was limited. Intel didn’t specify the computers’ details, so it was useless to speculate on performance.

This processor should be released in 2005 and reach respectable frequencies, equivalent to the current Pentium 4’s. Initially, it won’t be HyperThreading compatible. There are two reasons for this. Programs optimised for 4 process are still far from being available and the OS are completely shifted with the future CPU. More than two threads managed corresponds to a server for Microsoft.

Yonah

For laptops this CPU belongs to the Napa platform (“Centrino 3”). It integrates 2 Dothan cores (corresponding to the last Pentium M) with 2 Mb of cache L2 and will be compatible with LaGrande Technology and Virtual Technology but not with the other T (no HyperThreading and no 64 bit support).


Electrical consumption will be higher than the current Pentium M but the power management system will be able to reduce it by deactivating one of the cores. This processor should be the first with a 65 nm fabrication process. We have attended a conference with Marc Bohr, in charge of the development of the manufacturing process. According to him the process is well advanced. The equipment used for the 90 nm has been designed to migrate easily to 65 nm. This new process should allow putting cache areas not used at a specific time in stand-by, a little like the current Pentium M.

AMD Dual-cores

During our meeting with AMD, we saw a demonstration of two servers equipped with 4 dual-core Opterons. It was impossible, however, to see them or know their frequencies. The frequency is still most likely below the Gigahertz level. With AMD yield problems, will it allow them to build high frequency dual-core CPUs? If the yield isn’t very good and the low number of CPUs able to work at high frequencies, the probability of having two high frequency dies is strongly reduced. This might be a problem for AMD if the 90 nm yield isn’t improved. AMD is planning on 1.6 GHz for this CPU release. This confirms that our ideas on the yield issue are probably true.


One of the two servers had a strange configuration. The screen displayed 8 CPUs, or 4 dual-core CPUs, but only 512 MB of memory were present and placed in the memory bank of a single processor. This configuration is suspicious and we doubt the real functioning of the computer with the indicated configuration… The other server didn’t have this anomaly.

Solution : the Ts

As for new CPU uses, Intel introduced five technologies:


- HT : Hyper Threading
- LT : LaGrande Technology
- EM64T : Extended Memory 64 Technology
- VT : Virtual Technology
- iAMT : Intel Active Management Technology

These uses are added to Intel’s usual Gigahertz and cache size. The cache size facilitates the marketing department’s job, helping to keep up with “Moore’s law“. Technological evolutions and horizons can be touted more easily. It is important to remind you that a few years ago Intel announced that according to the Moore’s law, processors would have more than a billion transistors in 2010. The Montecito already has more than 1.7 billion transistors. It’s not important, though, and is only a matter of interpretation of the law, which is vague enough to be true with any situation.

Hyper Threading

One technology that never gets the honours it deserves is Hyper Threading. With most of the benchmarks Hyper Threading doesn’t show any significant improvement, but this is normal because programmers only start to write their code (when they want) to separate it into several threads. Nevertheless systems equipped with this technology are more reactive and provide an increased ease of use. Intel is going to pursue their efforts in that direction and Hyper Threading is going to be more and more present in our tests.

On the top of that, Hyper Threading integration’s cost is relatively low and will force developers to optimize their programs for multi-cores before the CPU’s release. A dual-core CPU will be able to demonstrate its performance as soon as it’s released. As computers have more and more tasks, CPUs equipped for hyperthreading (via Hyper Threading or dual-cores) will have an advantage.

LaGrande Technology

This technology´s target is to help the OS to improve the computer security, like with the Palladium and Longhorn. Unfortunately we don’t know have much information, and the Microsoft Palladium technology possible misuses have already been the subject of many articles. Everyone would like to see improved security but no one wants to see the major music companies in their computers.

Extended Memory 64 Technology

Even if it has been discussed along with other technological problem solvers, the EM64T remains a taboo subject. Where are the Pentium 4Fs promised a few weeks ago that are compatible with this technology? We caught a glimpse of one in its box in the expo. They are on sale but unavailable for journalists.

Apparently Intel doesn’t want these CPUs tested, an abnormal situation whatever Intel’s strategic arguments on the subject are.

For Intel the 64 Bit isn’t as important as AMD affirms. The name chosen by Intel is quite clear: 64 bits: extended memory management. Both AMD and Intel exaggerate their arguments. The 64 bits process will be useful, but needs to be first mastered by developers. And contrary to AMD’s claims, the 64 bits process won’t be the much expected messiah. The first objective of the 64 bits is to efficiently increase memory size and then increase the speed of algorithm. We should see this improvement as several new available instructions rather than a radical change.

Virtual Technology

Also known as Virtualization Technology or Vanderpool/Silvervale Technology, VT should bring about a major improvement. VT permits dissociation of the OS from the hardware. In other words VT allows the operation of several OSs on a single computer.


The possible uses of this technology are numerous as is the likely future demand. This is an interesting potential for Intel and also for users if the software lives up to its promises. One possible use: if one OS is attacked by a virus, another OS could be used as a guardian and decontaminate the system, preventing the virus from contaminating other systems. Personal and professional systems could be completely separated. Using Windows and Linux simultaneously could also be possible. It will be a major technological breakthrough for the computer platform. CPUs and chipsets will have to be adapted for use with this technology, as they already partly are. But the most important part will be on the software level. Longhorn will probably be required so we will have to wait until 2006 to start using VT.

Intel Active Management Technology

The IAMT targets mainly professionals who manage servers or computers networks. The objective is to facilitate their tasks, give them more power from a distance and of course reduce management costs.

One of the strong points of the iAMT is its independence from the OS. With any OS and even if its not functioning, the iAMT should always be accessible. That will allow the IT network administrator to solve some problems at distance even if the OS is inaccessible because of this problem. The iAMT should also be able to warn the administrator before it starts damaging the system in order to correct it. The administrator should be able to do this without leaving his desk. The efficiency, rapidity and cost benefit should be of importance.

Mobile platform, conclusion

Mobile platform

The release of a whole platform instead of just a CPU has been a huge success for Intel. Centrino has quickly become the standard and now completely dominates the market. Laptops need to offer good performance but also connectivity and autonomy. All these elements are required at the same time.


Intel will of course continue to develop this platform. Early 2005, the Sonoma platform ("Centrino 2") will finally be released. Its release was postponed due to a problem with the base of this platform, the Alviso chipset. This platform will be compatible with the PCI Express and the DDR-II. The CPU will be a Pentium M Dothan with an improved FSB of 100 to 133 MHz. Sonoma will demand the use of a WiFi 54 Mbps solution instead of a current 11 Mbps.

Later, the Napa platform will replace it with a more powerful processor, the Yonah (Pentium M Dual-cores), the chipset Calistoga with a more advanced core graphic, and a new wireless solution (WiMAX?) called Golan.

All these improvements will require more power, and therefore more efficient batteries and better energy management. Intel mentioned new types of batteries, like the Lithium Polymer already used in cell phones. This technology increases battery life and reduces bulkiness and/or increases their capacity for the same size. Intel isn’t planning the release of fuel cells batteries before 2010, which could be the only one able to provide 8 hours of autonomy to current laptops. Intel also briefly introduced the ACPI 3.0 which improves the energy management of new technologies like the PCI Express and dual-cores.

Conclusion

The big surprise of this IDF was the position of Intel on dual-cores. Of course, we had some idea but we didn’t expect the dual core release on the server, desktop and laptop markets so early. 2005 will definitively be the dual-core year. More than their performances Intel spoke of their possible uses. As a technological leader they aren’t only working on new products but also on their possible uses. Virtualization is a very interesting and cutting edge example.

All these improvements lead to several questions and will probably also redefine « processors » and « system ». Should a processor with two CPUs be considered as one or two processors? Will we have to pay for one or two licences? Should a computer that is able to manage several OSs be considered as one or several computers?

As a technological leader Intel clearly indicates that they will solve these problems and new ones to push the market forward instead of waiting for it to evolve on its own (like some of their competitors that hope it evolves by itself). Amongst other issues is secured data transmission inside a home. With the DTCP-IP you will be able to share on all computers in a home secured data like a movie purchased on the Internet and without illegal distribution. Intel is also working on the "New net", an improved version of the current Internet which could resolve a few current issues. But at this point Intel’s vision of this doesn’t seem realistic enough for us.

You may have understood Intel’s strategy is to integrate technology in daily tasks, to facilitate our life, and be more and more present. It is the best way for the American giant to continue selling more processors. It is also the best way to attract the other billions of people which haven’t yet had access to this technology and will progressively start to use it.