• Taeyoon is awarded Best Poster by KIISE for our work FusionFlow!

  • Xinyue Ma is accepted to the intern program at Microsoft Research Redmond!

    She will work under the RiSE group in MSR Redmond for a 3-month internship program from September.

  • “FusionFlow" is accepted to appear in VLDB 2024!

    "orchestrating data preprocessing tasks across CPUs and GPUs while minimizing interference with GPU-based model training"

  • Xinyue Ma is accepted to the intern program at Microsoft Research Asia!

    She will work under the Intelligent Cloud and Edge group in MSRA for a 3-month internship program.

  • "Blaze" is accepted to appear at EuroSys 2024!

    "An unified caching mechanism that integrates the separate operational layers together."

Our Research

Our research goal is to advance the state of the art in emerging large-scale system platforms by making them more efficient, responsive, intelligent, and programmable. Our current research topics lie in the following areas:

Systems and AI: We build systems support for improving machine learning frameworks and prediction-serving systems, as well as leverage machine learning in producing intelligent system software.

Bigdata analytics: We build data processing pipelines for real-time big data analytics at cloud/IoT scale that enable system operators to promptly troubleshoot system anomalies, improving the performance and reliability of their services.

Systems for new HW: We produce substantially better system software in the face of the recent explosion of hardware features and heterogeneity, such as accelerators and processor/memory tailored to improve efficiency, density, and performance predictability.

READ MORE

Selected Publications

FusionFlow: Accelerating Data Preprocessing for Machine Learning with CPU-GPU Cooperation

Taeyoon Kim, Chanho Park, Mansur Mukimbekov, Heelim Hong, Minseok Kim, Ze Jin, Changdae Kim, Ji-Yong Shin, Myeongjae Jeon

Cost-effective On-device Continual Learning over Memory Hierarchy with Miro

Xinyue Ma, Suyeon Jeong, Minjia Zhang, Di Wang, Jonghyun Choi, Myeongjae Jeon

CarM: Hierarchical Episodic Memory for Continual Learning

Soobee Lee, Minindu Weerakoon, Jonghyun Choi, Minjia Zhang, Di Wang, Myeongjae Jeon

Jarvis: Large-scale Server Monitoring with Adaptive Near-data Processing

Atul Sandur, ChanHo Park, Stavros Volos, Gul Agha, Myeongjae Jeon

Zico: Efficient GPU Memory Sharing for Concurrent DNN Training

Gangmuk Lim, Jeongseob Ahn, Wencong Xiao, Youngjin Kwon, Myeongjae Jeon

Analysis of Large-Scale Multi-Tenant GPU Clusters for DNN Training Workloads

Myeongjae Jeon, Shivaram Venkataraman, Amar Phanishayee, Junjie Qian, Wencong Xiao, Fan Yang