Floating Point Operations Compatible Streaming Elements for FPGA Accelerators

We see that in most computers and applications the CPU is taxed, first and foremost, before other pieces of hardware are. As this is seen in most general usage cases, especially if someone has a strong CPU, there are others where it might be smart for your computer to use other components in your system. This is where hardware acceleration comes into play. In computing, hardware acceleration is the use of computer hardware to perform some functions more efficiently than is possible in software running on a more general purpose CPU. To perform operations at high speeds we require accelerators which can boost the speed of the circuit. In order to achieve this, we have custom circuits where flexibility of the circuit is not possible and soft process approach where there is only register to register transfer is present. To overcome these faults in the existing system a high performance, fine grained streaming processor, known as a streaming accelerator element, is proposed which realizes accelerators as large scale custom multicore networks. By implementing this approach with advanced program control and memory addressing capabilities, we can see that the program inefficiencies can be almost eliminated to enable performance and cost, which are not possible among other software programmable solutions. When used to realize accelerators for matrix multiplication it is shown how the proposed architecture enables real time performance. To support floating point operations we add Floating Point Unit FPU to the ALU of processing elements which performs IEEE754 2008 single precision floating point operations addition, multiplication, and subtraction. Chinta Sravani | Dr. Prasad Janga | Mrs. S. SriBindu "Floating Point Operations Compatible Streaming Elements for FPGA Accelerators" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-2 | Issue-5 , August 2018, URL: https://www.ijtsrd.com/papers/ijtsrd15853.pd