Saturday, February 2, 2019
Cray SuperComputer :: essays research papers
The Cray X-MP/22 manufactured by Cray interrogation Incorpo dictated (CRI) of Minneapolis, Minnesota was delivered and installed at the U of Toronto this September. The Cray is a well respect computer - mainly for its extremely fast rate of mathematical floating-point calculation. As the university states in its July/August computer magazine "ComputerNews", the Crays "level of performance should modify researchers with large computational requirements at the university of Toronto and other Ontario universities to compete effectively against the high hat in the world in their respective fields." The Cray X-MP/22 has two aboriginal Processing Units (CPUs) - the first 2 in the 22. The Cray operates at a clock rate of 105 MHz (the regular, run-of-the-mill IBMPC has a clock rate of 4.77 MHz). By quick calculations, you would be led to believe the Cray is only about 20 times faster that the PC. Obviously, this is not the case. The Cray handles data considerably diffe rently than the PC. The Crays circuits stick out an array of data (known as a vector) to be processes as a SINGLE entity. So, where the IBMPC may require several clock cycles to multiply two numbers, the Cray performs eitherthing in one clock cycle. This power is measured in Millions of vagrant Point Operations Per Second (MFLOPS) - which is to say the rate at which floating-point operations earth-closet be performed. The Cray MFLOPS vary as it does many activities, but a rate of up to 210 MFLOPS (per CPU) can be achieved. The second 2 in the X-MP/22 title refers to the two million 64-bit words (16Mb) of shared central memory. This can be expanded to four million words in the future(a) if the need arises. But it doesnt stop there The Cray can pipe culture back and forth between the CPU memory and the Input/ takings Subsystem (IOS). The IOS then takes it upon itself the store the information in any of the four shop devices i) one of the four 1200 Mb disk drives (at a rate of 5 .9Mb every second), ii) one of two standard 200ips 6250bpi tape drives, iii) a Solid nominate Storage Device (SSD) (which is much like a 128Mb RAM disk), or iv) through to a front-end computer (the U of T uses two the IBM4381 and a DEC VAX). These computers would be programmed (usually in FORTRAN) and the information passed onto the Cray. The results would then be transfered back to the front end computers.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment