Bo Yuan
Rutgers University-New Brunswick, NJ·Engineering·Early-career
Computers now run tasks that demand huge amounts of computation, such as AI models, robot planning, and quantum chemistry simulations. General-purpose hardware often cannot keep up efficiently, and the tricks used to speed things up can quietly introduce errors.
- Fields
Engineering
Computer Science
- Topics
computer architecture
hardware acceleration
AI accelerators
- Grants
- 3 known grants, through Apr 2029
What they work on
- Develop design principles for building domain expertise into learning-based robot motion planners, improving their reliability and performance in complex, resource-limited real-world settings.
- Build hardware primitives and dataflow and microarchitecture optimizations for motion planning, and automate design and mapping, testing on robotic simulators, FPGA boards, and real robots.
- Create HexAI, a framework that automatically explores architecture and mapping choices for AI accelerators and adapts hardware to real workloads such as training and multi-tenant inference.
- Jointly improve power, area, and timing of AI accelerator designs using predictive register-transfer level modeling and multi-fidelity Bayesian optimization, aiming for about tenfold better energy efficiency.
- Build a formal verification framework that uses floating-point SMT encodings and counterexample-guided abstraction refinement to prove error bounds for mixed-precision tensor contraction sequences.
- Extend verification to reordered and fused tensor contractions with incremental SMT solving and equality saturation, and use it to find implementations that best balance operation count, memory use, and verified error.
More about the research
This research designs computing systems by considering algorithms, hardware, and design automation together. The aim is computation that is fast, energy-efficient, and trustworthy. Three projects apply this approach: a specialized processor for robot motion planning, automated design of AI accelerator chips, and mathematical proofs that sped-up scientific computations remain accurate.
Why it matters
Faster, lower-power motion planning hardware could help autonomous cars, warehouse robots, and surgical robots make real-time decisions. Automated accelerator design could make energy-efficient AI hardware quicker and cheaper to produce. Verified tensor computations could give scientists in quantum chemistry, physics, and materials science confidence that sped-up simulations remain accurate and reproducible, supported by open-source tools.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Collaborative Research: CS2: Formally Verified and Performance-Optimized Tensor Contraction Sequences in Quantum Many-Body Computations
NSF · May 2026 – Apr 2029
Collaborative Research: SHF: Small: HexAI: Holistic Exploration for Design Automation of Efficient AI Accelerators
NSF · Oct 2025 – Sep 2028
CAREER: SHF: Chimp: Algorithm-Hardware-Automation Co-Design Exploration of Real-Time Energy-Efficient Motion Planning
NSF · Jan 2023 – Dec 2027
Closest labs
Jiajia Li
North Carolina State University, NC·Computer Science
Build a multi-level auto-tuning system that selects GPU execution strategies for ragged tensors, whose pieces vary in size, keeping overhead low via efficient sampling and search-space pruning.
Nathaniel Bleier
University of Michigan, MI·Computer Science
Combine learning-guided program synthesis with automated logical reasoning to generate instruction sequences, accelerator control codes and microcoded expansions that are formally proven equivalent to the original program.
Cristinel Ababei
Marquette University, WI·Computer Science
Explore the design space of systems pairing multi-core CPUs with SmartSSD computational storage devices to identify the computational kernels and parameters that give the best acceleration.
The school
Rutgers University-New Brunswick: 57 labs on PI Pages, $837M a year on research, 713 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.