Latest Research in Accelerator Microarchitecture Optimization

Top Research Topics in Accelerator Microarchitecture Optimization

• Distributed, Parallel, and Cluster Computing — 46 papers
• Hardware Architecture — 40 papers
• Parallel Computing and Optimization Techniques — 24 papers
• Machine Learning — 10 papers
• Performance — 7 papers
• Programming Languages — 7 papers
• Scientific Computing and Data Management — 6 papers
• Information Theory — 6 papers
• Mathematical Software — 5 papers
• Networking and Internet Architecture — 5 papers

Highest-Cited Papers

1. AI-coupled HPC Workflow Applications, Middleware and Performance (11 citations)
2. VoltGrid: Microsecond Collective Interposition for Transient dI/dt Mitigation in Multi-Accelerator Training Clusters
3. Twelve quick tips for designing AI-driven HPC workflows
4. VoltGrid: Microsecond Collective Interposition for Transient dI/dt Mitigation in Multi-Accelerator Training Clusters
5. The Price of the Bottleneck: A Pre-Registered, Multi-Domain Deployment Characterization of a Deterministic Edge Decision Token
6. Machine-Learned Mismatch and Task Preservation Beliefs in CoSMA DAI for Common Knowledge Aware Semantic Alignment
7. Comparative Analysis of Classical Machine Learning Vs. Deep Learning for Resource-Constrained Edge Devices
8. PipeGNN: A Bandwidth-Efficient GNN Accelerator with Node-Level Pipelined Push Execution
9. Scaling Fourier-Based Sparse Matrix Analysis on GPUs
10. Chiral Topological Edge Routing for Thermal Mitigation and Bisection Decongestion in 2D Mesh Networks-on-Chip
11. SemABR: Measuring Video Semantic Fidelity with Multimodal LLMs for Adaptive Bitrate Streaming
12. Zero-I/O Fault Recovery for Sharded Deep Learning via Dynamic Framework Dependency Rebinding
13. Chiral Topological Edge Routing for Thermal Mitigation and Bisection Decongestion in 2D Mesh Networks-on-Chip
14. cosmokdtree: a flexible OpenMP-parallelized k-d tree for computational astrophysics applications
15. Ozaki Scheme II: A GEMM-oriented emulation of floating-point matrix multiplication using an integer modular technique
16. Quantifying the Effect of HCLs on a Fixed-Microarchitecture MXFP4 Accelerator
17. MeshKV: A Network-on-Chip KV Cache Fabric for Scalable Transformer Decoding Accelerators
18. Automated Instruction Encoding Synthesis for Modern GPU ISA Compression
19. Accelerating Stateful Network Applications with Performance Prediction on SoC SmartNICs
20. COMPASS-ABS: Reducing Fragmentation in Shared GPU Clusters for Deep Learning Training Workloads

Sub-Regions

• Hardware Architecture — 56 papers
• Information Theory — 54 papers
• Machine Learning — 44 papers
• Distributed, Parallel, and Cluster Computing — 30 papers
• Parallel Computing and Optimization Techniques — 22 papers
• Software Testing and Debugging Techniques — 8 papers