Startup integrates optical processing within data center
LightOn, founded in 2016, has worked with in partnership with Paris-based cloud computing service provider OVH Group and claims performance improvements for certain machine learning tasks.
The OPU uses laser light that is shone on to a digital micromirror device (DMD) to encode for 1s and 0s with the light then redirected through a lens and random scattering medium assembly before being polarize and read by a conventional camera. This allows very large matrices to be manipulated in parallel. One of the operations that can be done is kernel classification. Typically, the DMD can handle matrices of the order of 1k by 1k.
For a task called transfer learning the OPU showed six-fold speed up at five times greater energy efficiency than a GPU-based solution. This translates to 30x less power consumed. Another benchmark on time-series analysis with a recurrent neural network demonstrated a 200x speedup over conventional CPUs with large RAMs.
LightOn intends to make its technology available as a service that will open up for beta customers in Spring 2018.
LightOn Cloud service will be accessed through a user interface that provides access to virtual machines that feature a combination of CPU, GPU and LightOn’s OPU. The LightOn application programming interface (API) will work with PyTorch and Scikit-Learn. The company is working on providing compatibility with other popular machine learning frameworks such as TensorFlow.
LightOn is one of several startup companies seeking to use the low energy and parallelism of optical processing to accelerate machine learning.
Related links and articles:
Random projections through multiple optical scattering: Approximating kernels at the speed of light
News articles:
MIT optical AI spin-off raises funds from Baidu
MIT spin-off raises funds for optical processor
Nanoscale photonics switch could enable 10-100 Terabit/s data transfer
Solid-state lidar maker opens factory