TABLE OF CONTENTS:
  1. Introduction
  2. Latest Brands and Families
  3. Latest Architecture and Technology
  4. General Operation and Specifications
  5. Performance
  6. Conclusion
  7. References

 

1. Introduction
The use of Central Processing Units (CPUs) first became popular in the seventies when Intel developed the first microprocessor for a calculator. Its complexity has continued to advance over the past decades and has not stopped growing at this present day. It is important to identify the currently renowned brands as well as the design, functions, and benchmarks if one were to truly identify the power of CPUs today.

 

2. Latest Brands and Families
According to the webpage, CPU History: A Timeline of Microprocessors (2008), there exists numerous processor makers though Intel and AMD are presently the major two. Nevertheless, Intel’s advancement in recent times has made their processors superior with the latest “Nehalem microarchitecture” based processors (Intel Microarchitecture, Codenamed Nehalem, 2010). The Nehalem family of processors include the latest Intel Core i3, Core i5, and Core i7 processors which have recently succeeded the “Core microarchitecture” based processors. The Core family includes the popular Core 2 Duo, Core 2 Quad, and Core 2 Extreme. In addition, Mac computers also manufacture with the current Intel Nehalem family of processors (Apple Mac Pro Features, 2010). Despite Intel’s superiority, the webpage, AMD Processors for Desktop PCs (2010) shows us many computers are still sold with AMD processors since they are bundled with graphic cards manufactured by ATI (a company owned by AMD).

 

3. Latest Architecture and Technology
Mentioned previously, the Nehalem microarchitecture based processors have today’s become the dominant processors. As stated on the Intel website in Intel QuickPath Technology (2010), this new family of processors introduces major changes such as a new technology called QuickPath Interconnect (QPI). The QPI replaces the Front Side Bus (FSB) which has been used for carrying data between the CPU and northbridge for many generations. Furthermore, processors such as the Core i7 now offer up to 6 cores in addition to its new level 3 (L3) cache and embedded DDR-3 memory (Intel Core i7 Processor Extreme Edition, 2010). Unlike previous models, the L3 cache also known as the common or smart cache is up to 8MB, and can be shared by all cores of the processor. The DDR-3 integrated memory controller may also be unfamiliar. Unlike the dual in-line memory modules (DIMMs) inserted into the motherboard, the Nehalem processors have a relatively small amount of memory integrated with them. As suggested by the diagram (Intel Nehalem Microarchitecture, 2010) below, the DDR-3 memory in the processor provides channelling between the QPI and L3 cache:

 

4. General Operation and Specifications
The basic functions of a CPU include five different requirements (Lan, 2007). These requirements include a “fetch instruction” which allows the processor to read instructions from the memory and an “interpret instruction” for the processor to determine what must be done. A “fetch data execution” command might also be called if an instruction requires further data from the memory. Next, Lan (2007) mentions a “process data execution” for any logical or arithmetic operation required. The control unit (CU) calls the arithmetic logic unit (ALU) to perform this step in the CPU. The final stage is the “write data execution” allowing data to be written into the memory if needed. Furthermore, there exists a level of memory called registers. Registers are divided into two types, user-visible registers and control/status registers. The user-visible registers are generally associated with the machine language that the processor runs. These registers can be further broken down into register types such as data registers to hold data, address registers for indexing or addressing, and condition codes registers set by the results of operations (Lan, 2007). On the other hand, control/status registers are not visible to users. Some examples of control/status registers given by Lan (2007) include instruction registers (IRs) containing the most recent “fetch instruction”, or memory buffer registers (MBRs) containing data yet to be written to the memory.

 

5. Performance
When referring to the graph from High End CPU’s – Intel vs AMD (2010), it is clear Intel has superiority over AMD by quite a large margin. The decimal digits such as 3.33GHz are the clock rates in hertz measuring how many cycles the processor can do per second. However, it is important to know that the clock rate alone does not determine a processors capability. There are other specifications one should take into account such as the bus speed, the cache memory, the number of cores, and the thermal design power (TDP) representing the amount of cooling power required. Notwithstanding there is a risk of malfunctioning, processors can be overclocked. This means their intended clock rate can be pushed to a higher rate than the original. The webpage Overclocked Processors and System Bus (2010) teaches users on the damage, instability, and unpredictability of overclocking a CPU. It also states that Intel does not warrant any overclocked processors. Thus they should be returned immediately to the manufacturer if a purchaser has initially received an overclocked processor.


This picture shows a Celeron 2.13GHz overclocked to 4000MHz(4.0GHz).

 

6. Conclusion
The progress of central processing units in the past decades has certainly been one of the biggest technological advancements made. This report nevertheless lacks the precise details of the mechanisms behind a processor. If exact architectural details are needed, readers must seek for it on a report based solely on how processors work. By observing the latest models and architecture, as well as the processor operations and capabilities, one can however identify and have a general understanding of the most advanced CPUs in the 21st century.

Download Full Essay in .pdf

 

7. References

CPU History: A Timeline of Microprocessors. (2008). Retrieved March 2, 2010, from
http://everything2.com/title/CPU+history%253A+A+timeline+of+microprocessors

Intel Microarchitecture, Codenamed Nehalem. (2010). Retrieved March 2, 2010, from
http://www.intel.com/technology/architecture-silicon/microarchitecture.htm

Apple Mac Pro Features. (2010). Retrieved March 2, 2010, from
http://www.apple.com/macpro/features/processor.html

AMD Processors for Desktop PCs. (2010). Retrieved March 2, 2010, from
http://www.amd.com/uk/products/desktop/processors/Pages/desktop-processors.aspx

Intel Core i7 Processor Extreme Edition. (2010). Retrieved March 2, 2010, from
http://www.intel.com/products/processor/corei7EE/index.htm

Intel QuickPath Technology. (2010). Retrieved March 3, 2010, from
http://www.intel.com/technology/quickpath/index.htm?iid=support

Intel Nehalem Microarchitecture. (2010). Retrieved March 3, 2010, from
http://upload.wikimedia.org/wikipedia/commons/6/64/Intel_Nehalem_arch.svg

Lan, N.T.H. CPU Structure and Functions. (2007). Retrieved March 3, 2010, from
http://www.lc.unsw.edu.au/onlib/ref_apa2.html

High End CPU's - Intel vs AMD. (2010). Retrieved March 3, 2010, from
http://www.cpubenchmark.net/high_end_cpus.html

Overclocked Processors and System Bus. (2010). Retrieved March 3, 2010, from
http://www.intel.com/support/processors/tools/frequencyid/sb/cs-007627.htm