Concept: Parallel Computing and Optimization Techniques
Adaptive CPU frequency scaling for energy-efficient and sustainable edge computing under renewable energy uncertainty
Optimizing CPU performance on renewable-powered edge servers using machine learning
Thoth: Uncovering Data-Dependent Memory Access Patterns via Annotation-Directed Load Sampling
Hardware prefetcher for irregular memory access in sparse data structures
CXL-SpecKV: A Disaggregated FPGA Speculative KV-Cache for Datacenter LLM Serving
Offloading memory to remote accelerators improves LLM inference speed and reduces costs
It’s about Time: Temporal Abstractions for Asynchronous GPU Tensor Computations
Managing timing and coordination in asynchronous GPU tensor operations
SLAWS: Spatial Locality Analysis and Workload Orchestration for Sparse Matrix Multiplication
Improving efficiency of sparse matrix multiplication through adaptive locality analysis
PAT: Accelerating LLM Decoding via P refix- A ware A t tention with Resource Efficient Multi-Tile Kernel
Accelerating language model inference by reusing shared prompt cache across concurrent requests
Performance and energy consumption optimization of ternary optical computers based on the M/G/1 queuing model
Energy-efficient scheduling for optical computers using queuing theory
Dynamic Latency-Throughput Balancing in Distributed Large Model Inference with Interleaved Parallelism
Dynamically balancing speed and throughput for serving large AI models across multiple GPUs
WaSC: Hardening WebAssembly Sandboxes via System Interface Decoupling
Improving WebAssembly security through isolated system interface management
Integrating Quantum Software Tools with(in) MLIR
Using MLIR to create unified quantum software toolchains
