|
|
|
Enter the world of processors
THE brain of any computer is a tiny silicon wafer called the microprocessor popularly known as the CPU. It determines, at least in part, which operating systems could be used, which software packages could run, how much energy is used, and how stable the system will be, amongst other things. The microprocessor is the most complex mass-produced product ever made, with more than 5.5 million transistors performing hundreds of millions of calculations each second.
According to Moore’s Law, formulated in 1965 by Gordon Moore, co-founder of Intel, the number of transistors per integrated circuit would double every 18 months. Moore predicted that this trend would hold for the next ten years. In fact, as the graph illustrates, Intel has managed to doggedly follow this law for far longer. In 1978, the 8086 ran at 4.77MHz and had less than 30,000 transistors. By the end of the millennium, the Pentium 4 had a staggering 42 million transistors and ran at 1.5GHz. Therefore, as far as processor performance is concerned, there is apparently no end in sight.
Architecture condensed
When the Hungarian born John von Neumann, first suggested storing a sequence of instructions, a program per say, in the same memory space as the data, the idea was heralded as truly innovative. This suggestion was in his “First draft of a report on the “EDVAC,” written in 1945. The report organized the computer system into four main parts: central arithmetical unit, central control unit, memory, and input/output devices. Today, more than half a century later, nearly all processors are based on the “von Neumann” architecture.
The underlying principles of all computer processors are the same. Fundamentally, they all take signals in the form of zeroes and ones, manipulate them according to a set of instructions, and produce an output again in the form of zeroes and ones. Modern day microprocessors contain tens of millions of microscopic transistors, which are fabricated directly on the chip, combined to form integrated circuits that, in turn, make up the processor.
The performance of any processor depends on a wide range of factors, but thanks to global marketing, most people use the clock speed of the processor as the yardstick of choice. Clock speed refers to the number of times a processor can go through its cycle per second. This is measured in Hz (Hertz). A cycle consists of fetching instructions and data from memory, processing them and returning the results.
Majority of the data resides on secondary storage mediums such as hard disks. When the CPU needs information, it is read from the storage on to the primary memory, usually called the RAM. From there on it is sent to the processor. The data bus determines the amount of information a processor can accept for processing in a unit time. Current generation of processors boasts 32bit and 64bit buses; these are referred to as 32bit and 64bit architectures, respectively and are discussed in more detail later.
Cache is a like a storage warehouse in between the primary memory and the CPU. It stores recently accessed data and attempts to predict what information the CPU will request next. Logically, the CPU checks the cache before requesting the information from elsewhere. As the cache resides on the processor itself, it can furnish the required information extremely fast, producing significant performance gains. Modern cache design divides caches into three categories, from Level 1 till 3; the higher the number, the further away the cache is from the processor. This architecture can be looked at as a warehouse system. L1 is the local, L2 is the city, and L3 is the territory warehouse. As expected, these caches are searched in an ascending order, from L1 to L3.
Microprocessor types
For some years, two families of microprocessors have dominated the PC industry - Intel’s Pentium and Motorola’s PowerPC. These CPUs are also prime examples of the two competing CPU architectures of the last two decades - the former being classed as a CISC chip and the latter as a RISC chip.
The majority of today’s processors cannot rightfully be called completely RISC or completely CISC. The two textbook architectures have evolved towards each other to such an extent that there’s no longer a clear distinction between their respective approaches to increasing performance and efficiency.
The 4004 was the forerunner of all of today’s Intel offerings and, to date, all PC processors have been based on the original Intel designs. The first chip used in an IBM PC was Intel’s 8088.
Architectural advances
The Pentium is the defining processor of the fifth generation and provides greatly increased performance over the 486 chips that preceded it, due to several architectural changes. Intel’s seventh-generation Pentium 4 represented the biggest change to the company’s 32bit architecture since the launch of the sixth generation Pentium II. In an attempt to better address the low-cost PC sector, hitherto the province of the cloners, AMD and Cyrix- Intel launched its Celeron range of processors in April 1998.
In the spring of 2001, the first Pentium 4 based Xeon was released, at clock speeds of 1.4, 1.5 and 1.7GHz. For several months after the Pentium 4 began shipping in late 2000, the leadership in the battle to have the fastest processor on the market alternated between Intel and rival AMD, with no clear winner emerging. However, towards the end of 2001 AMD had managed to gain a clear advantage with its Athlon XP family of processors. Intel’s response came at the beginning of 2002, in the shape of the Pentium 4 Northwood, boasting speeds in excess of 3GHz.
The new breed
A shining example of innovation is Transmeta’s Crusoe processor. It is a revolutionary family of solutions, specially designed for the handheld and lightweight mobile computing market. It consumes 60 to 70 per cent less power and runs cooler than competing chips, by transferring the most complex part of a processor’s job, determining what instructions to execute and when , to software in a process called Code Morphing. Because it enables a battery charge to last twice as long, this technology is ideal for road-warriors.
Crusoe’s hardware architecture requires far fewer transistors than conventional x86 microprocessors thereby minimizing power consumption. Modern day processors, including Crusoe, execute several instructions at once to improve performance. However, a large fraction of the transistor count in traditional processors is devoted to rearranging instructions for optimal parallel execution.
Crusoe’s unique design avoids this power and transistor penalty by implementing such complexities in the Code Morphing software. These missing transistors then allow Crusoe to operate at lower power, in addition to making it easier to design and cheaper to manufacture.
Future watch
No discussion regarding processors would be complete without peeking into the future plans of the two main processor manufacturers: Intel and AMD.
The era of 64bit computing: 8-bit, 16bit, 32bit and now 64bit CPUs. This count seems to be growing in multiples of eight. But what does this bit numbering stand for? This numbering specifies the number of bits the CPU is able to process at one time. Intel’s first microprocessor, the 4004, was a 4bit CPU, 8085 was 8bit, 8086 was 16bit and then finally, 80386 marked the beginning of 32bit microprocessors. After this, Intel’s Pentium, Pentium pro, Pentium II, Pentium III, Pentium IV, Xeon, Celeron, AMD’s K6, Athlon, Duron all have been 32-bit microprocessors. All said and done, the world is seeing a shift from these 32bit CPUs to 64bit CPUs, with the leading chip manufacturers like Intel and AMD offering new 64bit chips.
But the big question is whether we need these 64bit CPUs or are the present day 32bit processors sufficient for meeting our requirements. Also, we cannot say for sure that 64bit chips will run applications twice as fast as 32bit chips. The main benefit of 64bit computing isn’t that it helps in running existing applications faster; it’s that it allows totally different types of applications.
A major benefit of a 64bit computer architecture is the amount of memory that can be addressed. In the mid-1980s, the 4GB addressable memory of 32bit platforms was more than sufficient. However, by the end of the millennium large databases exceeded this limit. The time taken to access storage devices and load data into virtual memory has a significant impact on performance. 64bit platforms are capable of addressing an enormous 16 TB of memory, four billion times more than 32-bit platforms are capable of handling.
In real terms this means that whilst a 32bit platform can handle a database large enough to contain the name of every inhabitant of the USA since 1977, a 64bit one is sufficiently powerful to store the name of every person who’s lived since the beginning of time. Itanium from Intel and Claw-Hammer from AMD compete in the 64bit computing arena.
HTT: maximizing powers: For the last decade, designers have relied on instruction-level parallelism to improve processor performance. Today, as transistor density continues to grow at the pace predicted by Moore’s Law, older design models are yielding diminishing returns. Intel’s Hyper-threading (HT) Technology, first unveiled in 2001, is an essential component of Intel’s strategy to successfully ramp performance, control power consumption and manage heat dissipation into the coming decades.
HT Technology makes a single processor act like multiple processors to the operating system. In its current implementation, HT Technology delivers two logical processors that can execute different tasks simultaneously using shared hardware resources. While an HT Technology-enabled processor won’t equal the computing power of two physical processors, the newly announced Intel Pentium 4 processor with Hyper-Threading Technology is delivering performance increases of up to 25 per cent on desktop platforms.
HyperTransport: This technology is a high-speed, low latency, point-to-point link designed to increase the communication speed between integrated circuits in computers, servers, embedded systems, and networking and telecommunications equipment up to 48 times faster than some existing technologies.
HyperTransport technology is designed to:
— Provide significantly more bandwidth than current technologies
— Use low-latency responses and low pin counts
— Maintain compatibility with legacy PC buses while being extensible to new SNA (Systems Network Architecture) buses.
— Appear transparent to operating systems and offer little impact on peripheral drivers.
HyperTransport technology was invented at AMD with contributions from industry partners and is managed and licensed by the HyperTransport Technology Consortium, a Texas non-profit corporation. The full specification and more information about HyperTransport technology can be found at HyperTransport.org.
The final frontier
In the computer industry, hardware and software are the two basic links in the computing chain. They drive each other, exchanging leadership off and on. For instance, WindowsXP uses all those fancy aesthetic features not because Microsoft just thought of them. Rather, the hardware was never powerful enough to support such a rich interface till now.
Still there will always be those who require extreme processing power, such as in university research labs. No single processor can cater to these needs. The only conceivable method therefore is to link a number of processors together to act as one ultra powerful processor. Welcome to the era of Distributed/Parallel Computing. SETI@HOME is a prime example of harnessing the idle processing power of numerous machines distributed across the globe on the internet.
For more details, visit
The writer
