GPU and Multi-GPU Architecture
Scalable memory systems and new memory technologies
Scalable on-chip and off-chip interconnects
Chip-level and system-level scheduling
Power, performance, and energy-efficiency in large-scale systems
Specialized accelerators for workloads like AI algorithms, crypto algorithms, databases, etc.
Hardware-software co-design
Systems Infrastructure for large LLM training and inference
Systems/AI algorithms codesign (e.g., for sparsity)
AI/ML for systems (hardware design, code optimization, etc.)
ML for EDA
GPU-accelerated algorithms
Languages and programming models for parallel computing
Optimizing compilers and AI-based performance assistants
Distributed runtime systems
Systems software and operating system interfaces
Optimizing GPU-accelerated workloads
Compilers and code verification
Large-scale GPU networking
Topologies, routing, and congestion control
Networking techniques at the intersection of scale-out and scale-up
GPU Accelerated EDA
