The Biggest Computer Chip in the World
US-based California-based Cerberus Systems claims to have created the largest computer processor. This processor, called the ‘wafer scale engine’, is slightly larger than the iPad. The company says a processor is sufficient to use sophisticated artificial intelligence technology, such as driverless cars and surveillance software.
Why innovation is important?
As the days go by, the size of the processor is getting smaller. Now, many chips are made together and then cut apart.
The most powerful central processing units (CPUs) on a computer have 4 processor cores. At the same time, each of them can work separately. Graphics processor units (GPUs) usually have more cores. That’s why GPUs are given priority over artificial intelligence technology. The higher the core, the more information can be processed at the same time.
The most powerful GPU has about five thousand cores. And Cerebrus’s new chip has four million cores. There is a high speed connection between the cores. Having more cores can easily work for complex machine learning technologies, despite the low power consumption. According to the manufacturer’s comment, the wafer scale engine will reduce from months to minutes while processing complex data.
Some different opinions
There is no doubt that the chip will work fast if the information on Cerberus Systems is accurate. However, Ian Qutres, editor of the information-based online magazine AndTech, seems a bit worried. He says the advantage of a small computer chip is that it works on low power, cooling down fast. Working with big chips will require special infrastructure. I mean, not everyone is benefiting from it. And for that reason it is said to use artificial intelligence.
Prior to the wafer scale engine, the first chip specialized for artificial intelligence has been developed. Google built the tensor processing unit in 2016. Huawei plans to add neural processing unit facilities to Kirin processors next year.