ORCAS Lab
The Open-Source Research in Computer Architecture and Systems (ORCAS) lab at UCLA builds open GPU hardware and software, and uses it to explore new GPU architectures for AI, graphics, and ray tracing. Our main platform is Vortex, a full-stack, open-source RISC-V GPU. Students at every level, from first-year undergraduates to PhD candidates, make real research contributions on it.
People
Current Students
- Yanggon Kim
- Chengxuan Wang
- Nikhil Rout
- Jacob Leigh
- Xinle Song
- Mingxin Liu
- Pritam Mukhopadhyaya
Former Students
- Kuan Fu Chen
- Mark Diaz
- Sagar Jain
- Brandon Kam
- Jacob Levinson
- Randall Liu
- Swastik Purathepparambil
- Injae Shin
- Michael Srouji
- Chenchen Tang
- Alex Tao
- Georgios Triantafyllou
- Nathan Wei
- Eliot Yoon
- Steve Zang
Join ORCAS
We welcome students who want to build real hardware and software. If you are interested in GPU architecture, compilers, or open-source hardware, email me at blaisetine@cs.ucla.edu with your CV and a short note about what you would like to work on. Prospective PhD students should also apply to the UCLA Computer Science PhD program.
Lab Publications
VESICA: Fused Vector Extension for RISC-V Embedded GPUs
IEEE/ACM International Conference on Computer-Aided Design (ICCAD), 2026
PRISM: Accelerating Ray Tracing on RISC-V GPU
IEEE International Conference on Computer Design (ICCD), 2026
Supporting Async Barrier and TMA for Vortex GPGPU (POSTER)
Workshop on Open-Source Computer Architecture Research (OSCAR), 2026
Thread-Scalable 2:4 Structured Sparsity on an Open RISC-V GPU
Workshop on Open-Source Computer Architecture Research (OSCAR), 2026
RTL Design and Analysis of Sparse Tensor Cores for Unstructured Sparsity in RISC-V GPUs
Workshop on Open-Source Computer Architecture Research (OSCAR), 2026
Workshop on Open-Source Computer Architecture Research (OSCAR), 2026
PRISM: Accelerating Ray Tracing on RISC-V GPU (POSTER)
Workshop on Open-Source Computer Architecture Research (OSCAR), 2026
Hardware Acceleration of GPU Kernel Scheduling on RISC-V GPU (POSTER)
Workshop on Open-Source Computer Architecture Research (OSCAR), 2026
gem5-Vortex: Heterogeneous Cross-ISA Integration of a RISC-V GPGPU in gem5 (POSTER)
Workshop on Open-Source Computer Architecture Research (OSCAR), 2026
CycleFence: Precise Cycle-Level Profiling for RISC-V GPUs (POSTER)
Workshop on Open-Source Computer Architecture Research (OSCAR), 2026
Inside VOLT: Designing an Open-Source GPU Compiler
ACM SIGPLAN International Conference on Compiler Construction (CC), 2026 · DOI
Multiport Support for Vortex OpenGPU Memory Hierarchy
1st Workshop on Open Source Solutions for Massively Parallel Integrated Circuits (OSSMPIC), co-located with DATE 2025, 2025 · PDF