Research Focus Areas#
Hardware Acceleration for Neural Networks#
Development of specialized hardware architectures to accelerate neural network computation, with emphasis on optimizing throughput, latency, and power consumption.
FPGA-Based Prototyping#
Design and implementation of FPGA-based prototypes for neural network acceleration, enabling rapid validation of novel architectural concepts.
Neural Processing Unit (NPU) Design#
Custom NPU architectures optimized for specific neural network workloads, incorporating parallelization techniques and quantization strategies.
Hardware-Software Co-Design#
Integrated design methodologies that consider both hardware constraints and software requirements to achieve optimal system-level performance.
Embedded AI Systems#
Deployment of AI models on resource-constrained embedded devices for real-time applications in robotics, machine vision, and industrial automation.
Current Research Projects#
- Novel architectures for hardware-aware neural network optimization
- Low-latency inference pipelines for real-time AI applications
- Parallelization and quantization strategies for embedded neural networks
- AI-embedded systems for industrial automation and autonomous control
