About me
I am a fourth-year Ph.D. student affiliated with the research group led by Prof. Peipei Zhou at the Brown University.
My academic pursuits center on optimizing heterogeneous computing systems to efficiently support data-intensive applications, including large-scale graph analytics, machine learning, and emerging AI workloads. More broadly, I am interested in heterogeneous architecture design, software-hardware co-design, and developing programming abstractions that simplify the use of complex systems while preserving performance.
What i'm doing
-
GPU-initiated I/Os
GPU-initiated I/O enables the GPU to directly issue and manage storage operations, eliminating CPU intervention to reduce latency and improve overlap between computation and data movement.
-
Heterogeneous Systems
Coordinating different hardware platforms including GPUs, SSDs, FPGAs, etc., together for tasks with heterogeneous workloads.
-
Applications
Several data-intensive applications have been explored, including DLRM and graph analytics, along with computational workloads such as large integer multiplication, Rivest-Shamir-Adleman (RSA), Mandelbrot set generation, and homomorphic encryption.
-
Hardware Accelerator
It involves developing hardware and software together to better integrate and optimize the overall system for improved performance and efficiency.
