Papers
8
Total Citations
116
H-Index
6
About
Viktor K. Prasanna is a distinguished researcher specializing in hardware acceleration for artificial intelligence, with a particular focus on Reinforcement Learning (RL) and heterogeneous computing platforms. His work sits at the compelling intersection of machine learning algorithms and high-performance computing, where he has made significant contributions to accelerating some of the most computationally demanding AI training workloads. Prasanna's most impactful contribution, "Accelerating Proximal Policy Optimization on CPU-FPGA Heterogeneous Platforms" (46 citations), pioneered the use of FPGA-based hardware to dramatically speed up Proximal Policy Optimization, a state-of-the-art RL algorithm widely applied in robotics, game playing, and finance. His subsequent work, including the QTAccel framework (19 citations) and PPOAccel (11 citations), extended this vision by developing generic, reusable hardware design frameworks for multiple RL algorithm families, from Q-Table methods to deep policy gradient approaches. Beyond single-agent systems, Prasanna has pushed boundaries into Multi-Agent Reinforcement Learning and scalable parallel training architectures, demonstrating a comprehensive command of the field. With over 100 combined citations across his recent publications, his research offers invaluable guidance for anyone seeking to make RL training practical, efficient, and deployable at scale.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5QTAccel8 citations · 2020
- 6
- 7
- 8Accelerating Multi-Agent DDPG on CPU-FPGA Heterogeneous Platform4 citations · 2023