Computer architecture is fun and I want to build fast computing but with extra privacy guarantees. Here I work with Prof. Satish Narayanasamy on confidential computing / hardware guranteed privacy topics.
GPU performance analysis, especially for deep learning workloads.
Specialize in: GPU architecture, memory hierarchy & multi-device communication
I take computer architecture track and work with Prof. Ron Dresklinski & Prof. Satish Narayanasamy on computer architecture and system related researches.
Course work: EECS470 Computer Architecture (A), EECS281 Data Structures and Algorithms (A), EECS312 Integrated Circuits (A), EECS482 Operating Systems (in progress), Parallele CUDA Programming (in progress)
I started my university here and gained my interest in engineering. I learned the basics about heardware and curcuits and decided to keep working on hardware related topics.
Course work: VE270 Introduction to Logic Design (A+), VE280 Programming & Elem. Data Struct (A), VE401 Probability Methods in Eng. (A+), VV186/VV285/VV286 Honors Mathematics II/III/IV (A-, A, A)
Acclerate long sequence alignment to process ultra-long full-length mRNA or cDNA reads at high througput and long genomic contigs against a large referencce database. Exploit parallelism in long sequence alignment and offload workload to GPU. Optimize scheduling to maxumize throughput
Solve privacy concerns in genomic data sharing without accuracy penalty from metadata studies or performance penalty from homomorphic encryption. Build a centralized system that collects genomic data from host institutions and analyze them on central server enclave without leaking raw or intermediate data.
Design an out-of-order, 3-way scalar processor based on R10K design using system verilog. Add additional feature load store queue, advance branch predictor and cache heriachy.
Teach out of order processor design topics including branch prediction, pipelines, prefetching, caches etc. Hold lab sessions and develop exam problems regarding OoO processor design.
Probability theory and statistics is interesting, important but often misunderstood. From a wonderful piecs of data one can draw non-sense conclusion if probabalistic methods are not used in the right way. While dealing with computer security, it’s important that we can come to a conclusion that sensitive data is “almost impossible” to leak.
I enjoy teaching and want to devote my energy towards helping students. It was a remote semster and everyone was isolated and very stressful. I was lucky to be able to help and support students and be part of this wonderful teaching team.
https://www.linkedin.com/in/juechu-dong-b0b638220/
jiaochewchew 1596892951
4844 Bob & Betty Beyster Building
2260 Hayward St
Ann Arbor, MI
48105