Researchers build World’s First KiloCore Processor, Each Core Runs at 1.78 GHz
The world’s first KiloCore processor, which has a whopping 1000 independent programmable cores, has been created, with each core running at a maximum computation rate of 1.78 GHz.
The new 1000-core processor is built by a team at the University of California, Davis, Department of Electrical and Computer Engineering. The CPU packs 621 million transistors and was fabricated by IBM on their relatively ancient 32nm CMOS technology, which means you won’t be seeing mass production anytime soon.
Up until now, there has been no multiple-processor chip that has more than 300 processors, with most of them created for research purposes and only a few being sold commercially. “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,” said Bevan Baas, professor of electrical and computer engineering, who led the team that designed the chip architecture.
This KiloCore processor adopts a much more flexible approach than the usual Single Instruction Multiple Data (SIMD) which is utilized by processors such as GPUs. Each of the 1,000 processor cores can run their own small program independently of the others, which according to Baas, helps break an application up into many small pieces. Each piece then runs in parallel on different processors, allowing for high throughput and lower energy use.
Since each processor is independently clocked, it is capable of completely shutting itself down when not in use, saving more energy. What’s more impressive is that each core operates at an average maximum clock frequency of 1.78 GHz, as they are designed to transfer data to each other rather than using a pooled memory area that can be a bottleneck for data.
As for applications, the KiloCore processor can be used in wireless coding/decoding, video processing, encryption, and others that require large amounts of parallel data such as scientific data applications and datacenter record processing.
Source: UC Davis.