Computing power increased five-fold

Posted Jul 27, 2017 by Tim Sandle
Computer scientists have devised a method that increases, by a factor of five, the computing power of a standard algorithm. This is when the algorithm is processed using a standard chip called FPGA.
Stanford engineers have created a four-layer prototype high-rise chip. In this representation  the b...
Stanford engineers have created a four-layer prototype high-rise chip. In this representation, the bottom and top layers are logic transistors. Sandwiched between them are two layers of memory. The vertical tubes are nanoscale electronic "elevators" that connect logic and memory, allowing them to work together to solve problems.
Max Shulaker
Linköping University, in Sweden, houses a leading computer science laboratory. Here researchers have developed a new method to accelerate computing power. The five-fold increase in power relates to a programmable integrated circuit referred to as "FPGA," (or ‘field-programmable gate array’). An FPGA is an integrated circuit that can be configured by a customer or a designer after manufacturing (which accounts for the "field-programmable" part of the name). The FPGA configuration is specified using a hardware description language, which is similar to that used with an application-specific integrated circuit. FPGAs contain an array of programmable logic blocks together with a hierarchy of reconfigurable interconnects. These allow the blocks to be "wired together" so that many logic gates become inter-wired in different configurations.
The logic blocks can be configured to perform complex combinational functions. The blocks can be reprogrammed an unlimited number of times. The market in FPGAs is expected to reach $9.8 billion by 2020. This means the research is not only of advantage to businesses, it is also potentially lucrative.
With the new research, the technologists successfully increased the speed of an algorithm, within the chips, called the "fast Fourier transform." This algorithm is central to spectral analysis, radar technology and telecommunication.
Some business uses of FPGAs
FPGAS are used across many sectors, including:
Aerospace and Defense: Radiation-tolerant FPGAs.
Automotive: Automotive silicon and IP solutions for gateway and driver assistance systems, Consumer Electronics: Lower-cost solutions enabling next generation, full-featured consumer applications, such as converged handsets and digital flat panel displays.
Data Center: To aid high-bandwidth, low-latency servers, networking, and storage applications.
Medical: For diagnostic, monitoring, and therapy applications.
Until recently it was understood that an FPGA was at maximum capacity, according to lead researcher Dr. Oscar Gustafsson. This was achieved by creating a frequency greater than 450 MHz, by making a subtle change to the signal route. This has led to FPGAs that can potentially function five times as fast, or to deal with five times the number of calculations, compared with current systems.
The potential is huge, allowing major industrial sectors the opportunity to save large sums for demanding calculations in industry without needing to replace the hardware. The research is published in the journal IEEE Transactions on Very Large Scale Integration (VLSI) Systems, under the title “Efficient FPGA Mapping of Pipeline SDF FFT Cores.”