SPARC (derived from Scalable Processor ARChitecture) is a RISC (Reduced Instruction Set Computing) ISA (Instruction Set Architecture) developed by Sun Microsystems. These SPARC microprocessors can be found in notebooks to supercomputers such as enterprise servers. They run operating systems like Solaris, OpenBSD and NetBSD. SuperSPARC is the version of SPARC developed in 1992. SuperSPARC microprocessor uses the SPARC V8 architecture version. UltraSPARC is the SPARC microprocessor, which replaced SuperSPARC. UltraSPARC was developed in 1995 by Sun Microsystems. UltraSPARC used the V9 SPARC ISA and it was the first SPARC microprocessor to use the V9 ISA.
SuperSPARC
SuperSPARC is the version of SPARC microprocessor which was released in 1992 by Sun Microsystems. It was codenamed Viking. SuperSPARC microprocessor uses the SPARC V8 ISA. Sun introduced 33MHz and 40MHz SuperSPARC microprocessor versions. 3.1 million transistors were contained in SuperSPARC. Texas Instruments (TI) fabricated this microprocessor in Japan. SuperSPARC+ and SuperSPARC-II were two derivatives of SuperSPARC. The intention behind releasing SuperSPARC+ microprocessor was to fix few bugs present in the original version. However the SuperSPARC-II microprocessor, which was released in 1994, was an improved version compared to original SuperSAPRC microprocessor with speeds up to 80-90MHz. SuperSAPRC microprocessor had a L1 cache of 16KB. Its L2 cache had a capacity of 2MB. L3 cache was not present in SuperSPARC microprocessor. SuperSPARC-II was codenamed Voyager.
UltraSPARC
UltraSPARC is the version of SPARC microprocessor released by Sun Microsystems in 1995 replacing SuperSPARC-II. It used V9 ISA of SPARC architecture. In fact, it was the first SPARC microprocessor based on 64 bit SPARC V9 ISA. Texas Instruments carried out the fabrication of 64 bit UltraSPARC. 32 64-bit entries were in the integer register file. It is a superscalar processor, which executes instructions in-order in a pipeline with nine stages. There were two ALU units but only one could carry out multiply and division operations. UltraSPARC microprocessor has a special type of floating point unit called FGU (floating-point/graphics unit), which provides multimedia support as well. There are two levels of cache as primary and secondary. Primary cache is 16KB and secondary cache is 512KB to 4MB. It had six input and output ports in the form of three reads and three writes. It contained 3.8 million transistors.
What is the difference between SuperSPARC and UltraSPARC?
SuperSPARC and UltraSPARC microprocessors have many differences, especially since UltraSPARC microprocessor replaced SuperSPARC in 1995. SuperSPARC microprocessor used V8 SPARC ISA, while UltraSPARC microprocessor was the first SPARC microprocessor to use V9 SPARC ISA. In fact, UltraSPARC microprocessor was a 64-bit microprocessor. Understandably, UltraSPARC microprocessor had higher clock frequencies than SuperSPARC microprocessor. In terms of functional units, there was a noticeable difference. To achieve higher clock frequencies than SuperSPARC, UltraSPARC microprocessor has simpler units. For example, this was achieved by not cascading the ALU units to make sure that the clock frequency was not restricted. SuperSPARC microprocessor had 3.1 transistors, while UltraSPARC had 3.8 transistors. UltraSPARC microprocessor had a bigger L2 cache compared to SuperSPARC’s L2. Overall, UlatraSPARC provided a higher performance in all areas compared to SuperSPARC.