1 citations · 2 across the 9 of their papers we have counts for
10 papers
Agentic AI Workload Characteristics
Yichao Yuan, Ankita Nayak, Souvik Kundu +1
Agentic AI shifts LLM serving from isolated prompt-generation requests to stateful, multi-turn executions that repeatedly invoke the model, call tools, and grow context over time.…
KAIROS: Stateful, Context-Aware Power-Efficient Agentic Inference Serving
Yichao Yuan, Mosharaf Chowdhury, Nishil Talati
Power has become a central bottleneck for AI inference. This problem is becoming more urgent as agentic AI emerges as a major workload class, yet prior power-management techniques…
BlazingAML: High-Throughput Anti-Money Laundering (AML) via Multi-Stage Graph Mining
Haojie Ye, Arjun Laxman, Yichao Yuan +2
Money laundering detection faces challenges due to excessive false positives and inadequate adaptation to sophisticated multi-stage schemes that exploit modern financial networks.…
ZKProphet: Understanding Performance of Zero-Knowledge Proofs on GPUs
Tarunesh Verma, Yichao Yuan, Nishil Talati +1
Zero-Knowledge Proofs (ZKP) are protocols which construct cryptographic proofs to demonstrate knowledge of a secret input in a computation without revealing any information about t…
GraphMend: Code Transformations for Fixing Graph Breaks in PyTorch 2
Savini Kashmira, Jayanaka Dantanarayana, Thamirawaran Sathiyalogeswaran +3
This paper presents GraphMend, a compiler technique that automatically fixes FX graph breaks in PyTorch 2 programs. Although PyTorch 2 introduced TorchDynamo and TorchInductor to e…
MoE-Lens: Towards the Hardware Limit of High-Throughput MoE LLM Serving Under Resource Constraints
Yichao Yuan, Lin Ma, Nishil Talati
Mixture of Experts (MoE) LLMs, characterized by their sparse activation patterns, offer a promising approach to scaling language models while avoiding proportionally increasing the…
MoDM: Efficient Serving for Image Generation via Mixture-of-Diffusion Models
Yuchen Xia, Divyam Sharma, Yichao Yuan +2
Diffusion-based text-to-image generation models trade latency for quality: small models are fast but generate lower-quality images, while large models produce better images but are…
Vortex: Overcoming Memory Capacity Limitations in GPU-Accelerated Large-Scale Data Analytics
Yichao Yuan, Advait Iyer, Lin Ma +1
Despite the high computational throughput of GPUs, limited memory capacity and bandwidth-limited CPU-GPU communication via PCIe links remain significant bottlenecks for acceleratin…
Palermo: Improving the Performance of Oblivious Memory using Protocol-Hardware Co-Design
Haojie Ye, Yuchen Xia, Yuhan Chen +6
Oblivious RAM (ORAM) hides the memory access patterns, enhancing data privacy by preventing attackers from discovering sensitive information based on the sequence of memory accesse…
Everest: GPU-Accelerated System For Mining Temporal Motifs
Yichao Yuan, Haojie Ye, Sanketh Vedula +2
Temporal motif mining is the task of finding the occurrences of subgraph patterns within a large input temporal graph that obey the specified structural and temporal constraints. D…