It's A Wrap!!!

Well, thats it people. I have come to the end of my blogging duration. But fingers crossed, I will be maintaining it. I wont be posting as often as I do, but if i come across any information related to next generation processors, I will post it up.

Honestly I think that the main idea of this assignment is to build a blogging interest in students and at the same time, expose them to new technologies and skills. I myself have learnt many things from user interaction, to content development, embedding media elements, etc..etc. It really has been a worthwhile experience.

Special thanks to my HCI lecturer, Mr.Andrew for the support and ideas he contributed throughout the assignment period. Not to forget my classmates and friends for sharing the technical and interaction information.

Okay then, it is time for me to sign off. Catch up later on future posts. Ciao..!

Thermoelectric Cooling and Peltier Effect

A key component of the H2Ceramic technology is the thermoelectric cooling exchange. The core of the heat exchanger are two thermoelectric metals or semi-conductors that use the Peltier effect to pump heat from one compound to the other compound when current is applied.

As mentioned before, the concept is similar to a refrigerator that uses a refrigerant to move heat out into the environment, but the thermoelectric cooling uses electrons instead of refrigerant.

At the cold compound, energy(heat) is absorbed by electrons as they pass from one element to another while at the hot compound, electrons release the excess energy. Heat sinks and fans are placed at the hot compound to dissipate the heat away from the system.

Reversing the direction of electron flow in a thermoelectric mechanism causes heat to flow in the opposite direction. Plus varying the current allows for tight temperature controls.

Having said so much, it is obvious that H2Ceramic will become the future of processor cooling. It is able to target the cooling level to be just above ambient room temperature. Therefore, for extended CPU life and risk reduction in overclocking, H2Ceramic technology is the way to go.

Hybrid Cooling Mechanism...

CPU cooling has become more of an issue in recent years as the industry moved to higher frequency processors. CPU clock rates have risen sharply since, producing higher heat generation.

More advanced cooling is often required when a CPU is overclocked in a high-end gaming system. CPU heat output tends to rise exponentially during overclocking. With the rise of overclocking expectations, even liquid cooling solutions and heat sinks are trying hard to keep up. Therefore, a new cooling technology is required.

Let me introduce you to the new H2Ceramic cooling system recently unleashed by Dell. It uses a two-stage cooling process that combines high performance liquid to air heat exchanger, thermoelectric fluid chiller, and control circuitry to optimize CPU cooling with minimal power(Dell Inc, 2007). H2C uses sensors, controls, fan speed management and thermoelectric cooling to keep CPU temperature slightly above room temperature which prevents the formation of humidity condensation or frost.

Thermoelectric cooling modules at the center of the heat exchanger rely on the same concept that is used to counter the effect of direct sunlight on spacecrafts(Dell Inc, 2007). This concept is called the Peltier Effect which pronounces that heat is evolved or absorbed at the junction of two dissimilar metals carrying a small current, depending upon the direction of the current. This is especially useful in transferring or dissipating heat away from the processor.

Thats all for today. In the next post I will be describing more on thermoelectric cooling and the Peltier Effect. So, stay tuned if you are interested.

Qubits and Quantum Computation

Today's computer work by manipulating bits that exist in one of two states: 0 or 1. However, Quantum computers are not limited to only these two basic states. They encode information as quantum bits, or qubits which can also exists in superpositions: effectively this means that the qubit is both in state 0 and state 1.

Picture From http://www.quantiki.org/

For example, any classical register composed of three bits can store only one out of the eight different number combinations in a given time. A quantum register on the other hand is able to store all eight different numbers simultaneously in a quantum superposition.

Once a quantum register is prepared in a superposition of different numbers, operations can be performed on all of them. In brief, quantum processors can perform many different calculations in parallel. This has impact on the execution time and the memory required in the process of computation. According to physicist David Deutsch, this parallelism allows quantum computers to work on a million computations at once while your desktop PC works on one.

Because quantum computers contain these multiple states at once, it has the potential to be far more powerful than today's most powerful supercomputers.

(Qubits: Atoms that work together to act as computer memory and processor.)

(Superposition: Principle of quantum theory that describes the concept about the nature and behavior of matter and forces at the atomic level. It claims that any object is actually in all possible states simultaneously as long as we don't check. The measurement itself causes the object to be limited to a single possibility. Definition from Whatis.com)

Quantum Processing...

Today's post will be about the future of future processors. Its capability of quenching our thirst for speed and computing capacity spreads far from the horizon of silicon chips. Behold the introduction of Quantum Processors...

Moore's Law states that the number of transistors on a microprocessor doubles every 18-24 months, the year 2020 or 2030 will find the circuits on a microprocessor to be measured on an atomic scale. And the logical step would be to create quantum processors which will harness the power of atoms and molecules to perform memory and processing task (Bonsor & Strickland,n.d). Quantum computers have the potential to perform calculations way faster than any silicon-based computer.

On the atomic scale, matter obeys the rule of quantum mechanics which are quite different from classical rules that determine the properties of conventional logic gates. Quantum technology can offer much more than cramming more silicon and multiplying clock speed of processors. It can supply entirely a new kind of computation with new algorithms based on quantum principles.

Following on, I will be discussing on quantum bits called qubits and its difference from classical bits. Plus, I would also discuss briefly on how quantum processors work and its giant leap ahead of today's processors.

Weekend Wrap...

We have finally uncovered the topic about materials...It has been a long process especially with the importance of explaining the past, present and future of processor architecture. But I find all the effort put into gathering information and posting extremely worthwhile. I mean the knowledge of future processors is in your hands to be used!

In the upcoming weeks, I would be posting about different topics that are important especially in the semi-conductor industry. These topics include overclocking, cooling mechanisms and some basic functions. I may spend at least one post on explaining the functions of some processor parts such as FSB, clock speed, cache, etc..etc. I don't think I need to go deep into details as I'm sure most of you guys are quite familiar with those terms.

I will be including many relevant information about future processors in upcoming posts. So, you don't have to worry about the derailment from the original purpose of this blog.

Make sure you stay tuned and keep updated.

(Note: I recently changed the appearance of the blog. Plus, I have added a shoutbox and a media player. Feel free to make use)

High-K Gate Di-Electric Benefits...

It is probably one of the biggest advancement in fundamental transistor design. High-k gate dielectric together with metal gate electrodes deliver faster processors that consumer less power.

The entire semiconductor industry is striving to keep up with the heat produced by processors which increases as the number of transistors get more and more. Current leakage control through this new high-k material is a big step towards ensuring that transistors run cooler. It is proclaimed that high-k is able to reduce leakage of over 100 times.
At the same time, silicon gate electrode is replaced by new metal gate electrode to overcome compatibility issues between the gate dielectric and gate electrode.

If you are interested in finding out more about these new materials and its benefits, I strongly recommend you to visit the Intel research website ==> http://www.intel.com/technology/silicon/high-k.htm

On the other hand, if you are new to this topic and require step by step information you can visit http://www.intel.com/pressroom/kits/45nm/Intel%20High-K%20metal%20gate%20glossary_FINAL.pdf. This site explains in detail each term and component that is involved in the process of using new materials and 45 nanometer technology. Enjoy!!

Transistor Materials Part 3...

New materials show potential in substituting silicon dioxide gate dielectric where continued thinning makes it increasingly difficult to control current leakage. New class materials known as "high-k" will replace silicon technology that will provide support over several generations!

"High-k" stands for high dielectric constant, a measure of how much charge a material can hold (Intel, 2007). Different materials have different abilities to hold charge (sponge concept) and that directly relates to transistor performance. "High-k" materials, have a dielectric constant above 3.9 (the "k" of SiO2) and that means it is able to hold more charges than the silicon element. It is important to know that higher "k" increases transistor capacitance so that the transistor can switch properly between "on" and "off"without leaking any charges.

It is already known that the high-k material to be implemented in future processors will be hafnium based.

Because this high-k gate dielectric is not compatible with today's silicon gate electrode, a new METAL gate is introduced. Incompatibility between the gate dielectric and gate electrode causes undesirable effects and lower transistor performance. (Refer to part 1 if you need to see the connection between the gate di-electric and gate electrode).

The combination of new materials (metal gates and high-k gate dielectric) leads to transistors with very low current leakage and that brings forward energy-efficient and high performance processors.

Transistor Materials Part 2...

Before I start, I would like to clarify that the reason I have been focusing much attention on this topic is because the substitution of materials in transistors would be the foundation in processor architecture for the next 5-10 or even 15 years to come. Therefore, it is vital that we master this topic as it holds the key to the gates of next generation processors. (Note: New materials due to be implemented in processors coming second half of this year)

Anyway, lets not get carried away just yet and take a step back to reality by analyzing our current modern processors. (You need to understand the first part for this)

By 1980, technology improved exponentially according to Moore's Law and there was a need to upgrade the transistor design. What Intel (leading manufacturer at that time) did was, it added a low resistant capping layer over the same transistor design. This additional layer reduced current leakage back down to a low level (Smalley, 2007). However, when the gate dielectric gets too thin, current leaks from the main gate through the gate dielectric. When this happens, conducting carriers(electrons or holes) will be reduced at the bottom of the main gate electrode to form a depleting region. This causes an increase in thickness of the SiO2 gate dielectric which in turn causes the transistors to be less efficient hence reducing processor capabilities.


To overcome this problem, new materials were to take place of the Silicon gate dielectric and the main gate. I would explain this in the upcoming post which would be the last part about materials..

Transistor Materials Part 1...

To understand fully about the importance of the new materials, it is helpful to know certain important knowledge regarding transistors in processor architecture.

Transistors are basically switches that process the ones and zeroes in the computer environment. A gate is used to turn transistors on and off, and a gate dielectric (Sio2) is an insulator under the gate where its job is to separate the gate from the channel where the current flows.

The gate dielectric is also designed to prevent current leakage between the electrode, the source and the drain (see diagram above) Image from, http://www.bit-tech.net/hardware/2007/01/27/intel_45nm_technology _overview/1

If we put things to perspective and refer back to the previous post, we find that for the past few years, engineers were increasing transistor density in processor architecture by removing layers of the gate dielectric in terms of atomic layers. This can cause negative effects as current leakage through the SiO2 increases exponentially when the dielectric gate walls get thinner. As we know excessive leakage causes power release and heat output.

Besides that, manufacturers will run into other problems because as the dielectric gates get thinner, engineers run out of atoms to increase transistor density. Something is needed to be done because this problem imposes a limit on the extensibilty of Moore's Law and that is where the new materials come into play.

In the next post, I will describe more on how our modern processors are being built and following on, the concept and new materials on how future processors will be created.

45 nm Fun Facts..

I was thinking of taking a breather away from processor technology, therefore I plan to start on the transistor materials at the next post. I have posted much information on this blog and figured that it is best to take a short break. I believe that receiving too much knowledge at once can backfire as the chances of forgetting are higher. =)

Anyway, I came across some interesting facts and figures from the Intel website about the new 45 nanometer transistors, so I thought of sharing it with everyone. You might be surprised by some of the statements... Here it goes:

1. The original transistors built in 1947 could be held in your hand while hundreds of the 45 nm transistor can fit on the SURFACE of a single red blood cell. (That makes it micro micro small)
2. The price of a transistor in next generation processors are around 1 millionth the average price of a transistor in 1968. If car prices fell at the same rate, a car today would worth 1 cent!
3. 45 nm transistors switch on and off approx. 300 Billion times a second. Light travels less than a 10th of an inch during the time it takes a transistor to switch on and off (Extreme fast)
4. Bacteria is 45 times larger than a 45 nm transistor.

For more interesting facts you can download the pdf document from Intel at this hyperlink http://www.intel.com/pressroom/kits/45nm/Intel%2045nm%20Fun%20Facts_FINAL.pdf

Be prepared for the next post about transistor materials. It is slightly complicated, and might require you to have some Chemistry and Physics knowledge to fully understand the topic.

Core Architecture: 45 nanometer Transistor Technology

According to Moore's Law, the transistor density on processor circuit doubles every two years. Only a year ago, companies released their processors using 65 nanometer transistor technology but today, Intel has announced that it has made a significant breakthrough with technology using 45 nanometer transistors with different materials. This new technology increases transistor density in processors over twice the current value.

Over the last 10-15 years, manufacturers have used silicon dioxide (SiO2) insulators inside transistors, but the increase of transistors have caused these insulator walls to become much thinner. This is not good as the insulators are designed to prevent current leakage from the transistors which causes additional heat release, power consumption and even damage to the chip. Therefore, the 45 nanometer technology brings forth new materials into the production of the transistors. We will discuss more on the materials in future posts as for now the focus is more on the 45 nanometer transistors.

This new transistor breakthrough allows manufacturers to continue delivering high end computers, laptops, servers, etc. Besides more functionality, it also decreases cost for customers. Compared to the 65 nanometer, the new 45 nanometer provides the following benefits :

• Increased transistor count (Mentioned above)
• 20% faster transistor switching speed which will deliver higher core speeds and increased instructions per clock cycle
• Extend energy efficiency by reducing transistor switching power
• Higher cache memory and front side bus speed.

This new technology is due to release in the second half of this year. So those planning for an upgrade should wait to check out these new processors.

Multi-Core! Get More With One

Previously I spoke about 64 bit computing which allows the processor to handle more data, hence increasing the efficiency of computers. Today, the subject of discussion would be about Multi-Core Processing Units. The multi-core technology just surfaced a few months ago and it has already taken the market by a storm. It isn't just about a new generation of processors. It is an huge leap that will influence computing for years ahead.

What is it?
In simple terms, multi-core processors contains two or more "executional cores" within a single processor (Intel, 2007). It plugs into a single socket but the OS refers each core as a individual logical processor. The idea behind this implementation lies in the "divide & conquer" strategy. Meaning that, by distributing computational work done by one processor and spreading it over multiple cores, a multi core processor can perform more work within a clock cycle (Intel, 2007). This function is called "threading". Processors equipped with this function are able to execute completely different sets of codes simultaneously (Useful for programming and multimedia apps as their operation run parallel). This capability also enhances users experience in multi-tasking as a number of applications run concurrently with the background applications such as anti-virus, security and network management.

Future Multi-Core Processors
Beginning of the year, Intel recently showed their 80-core CPU!! Running at a staggering clock speed of 5.7Ghz. Each core is connected by a router to pass instructions around. For power saving, each core can be turned when its processing capabilities is not needed. For more information about this chip, visit http://news.softpedia.com/news/Intel-039-s-80-Core-CPU-Running-at-5-7-GHz-46881.shtml

Please do post your opinions..

Take Off with 64 Bit

Example Of A 64 bit Processor.

Courtesy of http://www.ispirian.com/wc/default.asp?Info=504311136_1-AMD--0

Okay..so here is where we start off our discussion about processors. The first subject that I will touch is 64 bit processing/computing.

I'm sure you must have heard stories /reports about 64 bit processing. But what exactly is 64 bit processing? 64 bit CPU architectures are those that are based on registers, address buses or data buses of that size. A 64 bit platform CPU generally supports 64 bit of integer data. This system holds more memory and performs larger number of calculations comparative to lower bit platforms.

Understanding Bits

A bit is a single 1 or 0 integer processed/stored by a transistor in the CPU. All processors are differentiated by the bit processing ability (data calculations). The bit rating of a processor (Ex. 32bit, 64 bit) concludes the largest number the processor can handle (Kyrnin, 2007). A number greater than that will require more than 1 clock cycle. A 32 bit processor handles a number up to 4.3 billion but a 64 bit processor can handle a number up to 18.4 quintillion!!! (Kyrnin, 2007)This concludes that a 64 bit processor reduces clock cycles and is more efficient in large number mathematics.

64 bit architecture makes working with huge data sets in application such as digital video, scientific computing, and large databases easier and faster. Not only speed but applications like multi-tasking and clustering are better suited for 64 bit architecture.

Memory

32 bit chips can address up to 4 GB of memory with the maximum allocation to any program is 2 GB. A 64 bit on the other hand can address up to 16 exabytes of memory (over 16 billion GB of memory!!). This isnt a big deal for remote computers but its a huge boost for super computers and servers. The biggest benefit is obviously the data transfer rate as pulling bits out of the memory is 60,000 times faster than from the hard disk.

If computers are to increase in processing speed, it is neccessary to take the 64 bit platform seriously. System that require more memory and large calculations will directly reap rewards from the 64 bit chip. Therefore it is clear that 64 bit computing holds the key to the next generation of processors.

For more information you can refer to the article from http://compreviews.about.com/cs/cpus/a/aapr64bit_2.htm

First Post

As you can see, this is my first post for the topic processors. I rather not go into much details at the moment as I need to spend more time on getting used to this blogging process. In brief, the discussion in my blog would include next generation processors(mostly technologies from Intel, AMD and IBM), more deeper information such as the clocking speed, FSB speed, cache memory size, architecture, etc, and also other relevant discussions such as overclocking and cooling mechanisms.

Everyone is welcomed to join in and bring up any interesting points they might have. Don't be shy to share your information ...