LightOn OPU now ventures into large scale scientific computing

Original Source Here

LightOn OPU now ventures into large scale scientific computing

We are excited to announce the release of LightOn ML v1.4, now available on PyPi and on LightOn servers on LightOn Cloud and on the LightOn Appliance offering.

With this release, users of LightOn’s technology can now have at their disposal a lightning fast scalable linear random projection on any Aurora Nitro OPU. A linear transform opens the path to many interesting applications, especially in the field of HPC and randomized linear algebra.

Large Scale

For example, while computing the trace of a matrix is usually computationally cheap, computing the trace of a function of a matrix can become really expensive really fast. In such cases, we can use randomized trace estimation. An application of such algorithms is the computation of triangles in a graph, that is central in complex network analysis, such as massive social networks.

Comparison of the quality of the results of triangle estimation on the OPU with its numerical equivalents. We remark that the results obtained optically agree very well with the numerical results.

A wider selection of HPC and RandNLA applications is presented in our preprint Photonic co-processors in HPC: using LightOnOPUs for Randomized Numerical Linear Algebra, and we will go into more details in a future blog post!

Performance comparison of a linear n xn random projection on the OPU and on an NVIDIA P100 16 GB GPU.

If you want to try out your latest idea, you can register to the LightOn Cloud for a Free Trial or apply to the LightOn Cloud for Research Program!

For on-premises computations you can take a look at LightOn Appliance, the most powerful photonic AI / HPC co-processor on the market.

The LightOn Team.

About Us

LightOn is a hardware company that develops new optical processors that considerably speed up Machine Learning computation. LightOn’s processors open new horizons in computing and engineering fields that are facing computational limits. Interested in speeding your computations up? Try out our solution on LightOn Cloud! 🌈

Follow us on Twitter at @LightOnIO, subscribe to our newsletter, and/or register for our workshop series. We live stream, so you can join from anywhere. 🌍


Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot

%d bloggers like this: