This latest entrant in the market— the Pentium 4 is based on a 32 bit micro architecture. Intel has launched his processor based on a new and revised architecture after a long time. In fact, it is the first change after 1995 in the processor architecture by Intel. The launch of Pentium 4 processors has started a new chapter in the history of processors. Some important incorporated in Pentium 4 processors include Net Burst Micro-Architecture that has a size of .18 micron. It features 42 million transistors on board. One of the outstanding features of the Pentium 4 is its bus speed, known as "Quad Pumped" 400MHz System Bus. It is worth mentioning that till now, AMD’s Athlon has had the maximum bus speed among all processors Athlon works on a 200MHz bus speed. The Pentium 4 speed breakthrough is a major landmark in terms of speed of data transfer. The Quad Pumped 400MHz system bus delivers three times the bandwidth of the Intel Pentium III processor system bus. It has 128-byte lines with 64-byte accesses as compared to the 32-byte lines in the previous generation. This provides a 3.2 gigabyte transfer speed between the Pentium 4 processor and the memory controller, which is the highest bandwidth desktop system bus available till now. Moreover, the new Pentium 4 processors use "Hyper Pipelined" Technology, which incorporates 20-stage pipeline depth and significantly increases processor performance, besides frequency capability. The introduction of a 20-stage pipeline in the Pentium-4 is the longest to be implemented in an x86 CPU. Thus by incorporating the 20-stage pipeline the Pentium 4 processors have a capacity to run at very high clock speeds—in excess of 1.5 GHz. Pentium 4 supports "Rapid Execution Engine", which allows the processor’s Arithmetic Logic Units (ALUs) to run at twice the speed of the core frequency of the processor that makes it possible to execute certain instructions in 1/2 core clock tick, resulting in higher execution throughput and reduced latency of execution. The cache speed of the processor, which has always been the point of discussion in the Pentium processor has been well addressed in Pentium 4 processors as it has 256K L2 Advanced Transfer Cache running at core processor speed. It also has 8K L1 Data Cache. The execution Trace Cache also improves the performance of the processor, as it is basically an advanced Level 1 instruction cache that caches decoded instructions (~12K micro-ops), thus removing the decoder latency from main execution loop. This revolutionary technology delivers a higher performance instruction cache besides making a more efficient use of cache memory storage. In addition, the Pentium 4 processor’s Level 2-Advanced 256 KB Transfer Cache delivers a 48GB/sec interface that scales with core frequency increase. The Advanced Dynamic Execution used by Pentium 4 is one of the great features of this processor, resulting in an extremely efficient out-of-order speculative execution engine that keeps the execution units busy. Another new feature is the enhanced branch prediction capability that keeps the processor executing to the correct program flow and reduces the mis-prediction penalty associated with deeper pipelines. Furthermore, the Enhanced Branch Prediction Capability compensates for the deeper pipeline’s likelihood of mis-predicted branches. The list of advanced features of Pentium 4 is very long, which also includes Streaming SIMD Extension 2 (SSE2). As compared to the Streaming SIMD Extension in Pentium-III, the Pentium 4 processors have Streaming SIMD Extension 2 (SSE2), which has 144 New instructions including 128bit SIMD Integer, Arithmetic and 128bit double precision floating point instructions in addition to SSE and MMX instructions. Thus the SSE2 extends MMX and SSE technology that is capable of delivering performance increase across a broad range of applications. Pentium 4 processors are not just a higher clock speed processor, but offer features that would change the architecture of future CPUs. It’s features like SSE2 instructions and the 20-stage pipeline depth are what make this processor really superior. The trace cache and the improved branch prediction algorithms can create wonders in the computing if implemented well. — VV |