Intel seems to agree.
‘Myriad X combines dedicated imaging, computer vision processing and – thanks to the industry-first Neural Compute Engine – high-performance deep learning inference within the same chip, and the results are opening up new realms of possibility, ‘ claims Intel’s Remi El-Ouazzane in the company’s announcement.
Apart from the Neural Computer Engine, which can run deep neural networks in real time, Myriad X’s 16 Programmable 128-bit VLIW Vector Processors can run multiple concurrent applications.
Intel swallowed California-based chip outfit Movidius in September to improve robots’ awareness of the real world. By buying the firm, which specialises in low-power-chip design for machine intelligence algorithms, Intel hoped to enhance its RealSense human-computer interaction camera technology, which provides face, gesture and speech recognition, as well as augmented-reality capabilities. And if you’re wondering what TOPS is, it refers to a trillion operations per second (TOPS).
Movidius, founded by David Moloney and Sean Mitchell a decade ago, was set up to bring deep learning and AI to devices such as smartphones.
In July, Intel and Movidius revealed a new $79 Neural Compute Stick, described by Intel as the world’s first USB-based deep learning inference kit and self-contained AI accelerator.
“Enabling devices with humanlike visual intelligence represents the next leap forward in computing”.
Named Myriad X, this VPU aims to deliver artificial intelligence capabilities in a low-power and high-performance package. Myriad X is an on-chip hardware block created to run deep neural networks at high speed and low power without compromising accuracy, enabling devices to see, understand and respond to their environments in real time.
16 MIPI lanes to connect up to 8 eight high resolution RGB cameras to Myriad X, supporting up to 700Mpixel/s image signal processing throughput. Where the Myriad 2 offers a claimed 100-150 gigaflops – 100-150 billion floating-point operations per second – of compute performance, the Myriad X is said to be ten times faster at over a teraflop of peak compute performance.
Intel is offering two chip packages for Myriad X. While one package has no memory, the other one as 4 Gbit LPDDR4 memory.
The tiny SOC also comes with 2.5MB of homogeneous on-chip memory, allowing for up to 450Gbps of internal bandwidth. The Myriad X looks about the size of a small coin, meaning placing it on the most compact of devices shouldn’t be too hard.