# ViBE: Co-Optimizing Workload Skew and Hardware Variability for MoE Serving

arXiv: https://arxiv.org/abs/2606.00735 · HTML: https://mlsanity.com/paper/2606.00735/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Efficient MoE training and inference systems, one of 12 directions in Mixture-of-experts.
- Builds on 24 of its 43 references that are in the map.
- First posted on arXiv 30 May 2026.

## Placed in

- Mixture-of-experts / [Efficient MoE training and inference systems](https://mlsanity.com/moe/efficient-moe-training-and-inference-systems/)

## Builds on (most cited first, 24 of 24)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://mlsanity.com/paper/1909.08053/) (2019)
- [GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding](https://mlsanity.com/paper/2006.16668/) (2020)
- [gpt-oss-120b & gpt-oss-20b Model Card](https://mlsanity.com/paper/2508.10925/) (2025)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [Splitwise: Efficient generative LLM inference using phase splitting](https://mlsanity.com/paper/2311.18677/) (2023)
- [Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve](https://mlsanity.com/paper/2403.02310/) (2024)
- [DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale](https://mlsanity.com/paper/2201.05596/) (2022)
- [FastMoE: A Fast Mixture-of-Expert Training System](https://mlsanity.com/paper/2103.13262/) (2021)
- [Accelerating Distributed MoE Training and Inference with Lina](https://mlsanity.com/paper/2210.17223/) (2022)
- [Comet: Fine-grained Computation-communication Overlapping for Mixture-of-Experts](https://mlsanity.com/paper/2502.19811/) (2025)
- [TAPAS: Thermal- and Power-Aware Scheduling for LLM Inference in Cloud Platforms](https://mlsanity.com/paper/2501.02600/) (2025)
- [Exploiting Inter-Layer Expert Affinity for Accelerating Mixture-of-Experts Model Inference](https://mlsanity.com/paper/2401.08383/) (2024)
- [MoETuner: Optimized Mixture of Expert Serving with Balanced Expert Placement and Token Routing](https://mlsanity.com/paper/2502.06643/) (2025)
- [Reducing Energy Bloat in Large Model Training](https://mlsanity.com/paper/2312.06902/) (2023)
- [SLO-aware GPU Frequency Scaling for Energy Efficient LLM Inference Serving](https://mlsanity.com/paper/2408.05235/) (2024)
- [Not All GPUs Are Created Equal: Characterizing Variability in Large-Scale, Accelerator-Rich Systems](https://mlsanity.com/paper/2208.11035/) (2022)
- [Characterizing the Efficiency of Distributed Training: A Power, Performance, and Thermal Perspective](https://mlsanity.com/paper/2509.10371/) (2025)
- [Integrated Hardware Architecture and Device Placement Search](https://mlsanity.com/paper/2407.13143/) (2024)
- [Lit Silicon: A Case Where Thermal Imbalance Couples Concurrent Execution in Multiple GPUs](https://mlsanity.com/paper/2511.09861/) (2025)

## Built on by (most cited first, 1 of 1)

- [TEMPO: Makespan-Aware Expert-Parallel Load Balancing Across Memory- and Compute-Bound Regimes](https://mlsanity.com/paper/2608.13057/) (2026)

## Abstract (arXiv)

In distributed Mixture-of-Experts (MoE) inference, input-dependent token routing interacts with GPU performance variability to create persistent stragglers under synchronized execution, where the slowest GPU determines layer latency. This performance variability is inherent to modern accelerators: manufacturing variation, power limits, and thermal conditions introduce measurable execution-time differences across nominally identical GPUs. The core challenge is that MoE execution-time imbalance arises from the interaction of workload skew and hardware asymmetry. Token routing produces uneven and layer-varying expert loads, while GPU throughput depends on device-specific operating characteristics and workload intensity. Prior work mitigates routing skew but assumes homogeneous hardware, optimizing token balance rather than execution latency. As a result, even balanced token assignments can leave hardware-induced stragglers unaddressed. Thus, we propose Variability-Informed Binning of Experts (ViBE), a hardware-aware expert placement framework that minimizes execution-time imbalance across GPUs. ViBE combines per-GPU performance modeling with expert activation profiling to assign high-load experts to faster devices and low-load experts to slower ones, reducing layer-level stragglers without modifying model semantics or hardware. Because both workload characteristics and effective GPU throughput can shift across serving conditions, ViBE supports lightweight recalibration under workload/performance drift to refresh its routing and performance estimates when needed. Results show that ViBE consistently reduces execution-time imbalance and improves SLO attainment by 14%, while lowering P90 TTFT by up to 45%. We further show that the impact of hardware variability increases at scale, making variability-aware placement important for efficient, high-utilization LLM serving.