A team at the University of California, Davis, Department of Electrical Computer Engineering have hit a world’s first by designing the first “KiloCore” processor. This processor has 1000 independent programmable processors, with a maximum computation rate of 1.78 trillion instructions per second!
The processor features 621 million transistors and was show off at the recent Symposium on VLSI Technology and Circuits in Honolulu on June 16. Bevan Baas, Professor of Electrical and Computer Engineering said, “To the best of our knowledge, it is the world’s first 1,000-processor chip and it is the highest clock-rate processor ever designed in a university”.
The highest core chips we have seen have had no more than 300 cores, with most of them being made research and only just a few sold commercially. The new 1000-core processor was fabricated by IBM on their 32 CMOS technology, so it is not something super advanced like the 14nm and 10nm nodes we are seeing in current processors.
Each of the 1000 cores can run a program independently of others. This is a much more flexible approach than using the Single Instruction Multiple Data (SIMD), which is the way GPUs work with computational data. The whole idea behind this is to break an application into many smaller pieces, with each of them running parallel on a different processors, allowing for higher throughput and lower power consumption.
The 1000 cores are independently clocked, many of them can completely shut themselves down when they are not needed. Each core has an average maximum clock frequency of 1.78 GHz. This new processor is reported to be the most energy-efficient “many-core” processor ever created. The processor can execute 115 billion instructions per second, while only dissipating 0.7W, which is low enough to be powered by a single AAA battery. The processor executes instructions 100x more efficient than modern laptop CPUs.
Applications are already in development for the KiloCore processor, they include wireless encoding / decoding, video processing, encryption, and other that require large amounts of parallel data. The scientific and data center communities and quite excited by the potential of this processor.