Zhongrui Chen

Hi there. I am a fourth-year PhD candidate in Computer Science at UNC Chapel Hill, advised by Ben Berg.

I work at the intersection of Queueing Theory and Computer Systems. Namely,

• Multiresource job scheduling: Scheduling jobs that request resources like (4 CPUs, 8GiBs, 2 GPUs, 10Gbps NIC). Hosts run many jobs in parallel subject to resource limits.
• Multiserver job scheduling: Multiresource scheduling with a single type of resource.
• LLM inference job scheduling: Resource allocation for inference jobs to improve system throughput and lower latency.

Our trilogy on throughput-optimal data center scheduling is as follows:

• Markovian Service Rate (MSR) Policies: [SIGMETRICS ‘25], [MAMA ‘24]
  A throughput-optimal multiresource scheduling policy with mean response time analysis. Compatible preemption modes: preemptive, non-preemptive, preemptive with setup costs.
• Quickswap: [Performance ‘25]
  A throughput-optimal non-preemptive multiserver scheduling policy with provably better performance in theory and in trace simulations.
• Scaling Cycles: [MAMA ‘26]
  A Quickswap-like policy for multiresource jobs with latency-related optimality results on fluid scale.