Lightning Talk- Accelerated ML on Intel® FPGAs using oneAPI
In this demo, we will present how users can run up to 40x faster their ML applications using oneAPI. As a use case we will show how Logistic Regression training can run 42x faster on an Arria 10 FPGA card using the oneAPI interface. The logistics Regression ML Model reference code was ported to DPC++ code and the results were compared. The implementation was done on Intel FPGA using oneAPI FPGA development flow.
Speaker(s)
Chris Kachris
Chris Kachris is the co-founder and CEO of InAccel that helps company’s speedup their applications using hardware accelerators (FPGAs) in the cloud or on-prem. Chris is the editor of the book ‘Hardware Accelerators in Data Centers’. He has over 15 years of experience on FPGAs (reconfigurable computing), digital design, embedded systems (SoCs), and HW/SW co-design for machine learning, network processing and data processing. Chris is the author of more than 70 papers on the domain of FPGA and reconfigurable computing and has more than 2000 citations for his work and he has served in the program committee of FPL, FCCM, SAMOS, and TRETS.
