# CTA-Pipelining: A Latency-Oriented Spatial Scaling Method for Multi-GPU Systems

arXiv: https://arxiv.org/abs/2607.07862 · HTML: https://mlsanity.com/paper/2607.07862/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 47 references that are in the map.
- First posted on arXiv 8 Jul 2026.

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

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness](https://mlsanity.com/paper/2205.14135/) (2022)
- [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://mlsanity.com/paper/1909.08053/) (2019)
- [SGLang: Efficient Execution of Structured Language Model Programs](https://mlsanity.com/paper/2312.07104/) (2023)
- [Efficient Large-Scale Language Model Training on GPU Clusters Using Megatron-LM](https://mlsanity.com/paper/2104.04473/) (2021)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model](https://mlsanity.com/paper/2201.11990/) (2022)
- [Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve](https://mlsanity.com/paper/2403.02310/) (2024)
- [DeepSpeed Inference: Enabling Efficient Inference of Transformer Models at Unprecedented Scale](https://mlsanity.com/paper/2207.00032/) (2022)
- [MegaBlocks: Efficient Sparse Training with Mixture-of-Experts](https://mlsanity.com/paper/2211.15841/) (2022)
- [SARATHI: Efficient LLM Inference by Piggybacking Decodes with Chunked Prefills](https://mlsanity.com/paper/2308.16369/) (2023)
- [NanoFlow: Towards Optimal Large Language Model Serving Throughput](https://mlsanity.com/paper/2408.12757/) (2024)
- [FLUX: Fast Software-based Communication Overlap On GPUs Through Kernel Fusion](https://mlsanity.com/paper/2406.06858/) (2024)
- [Mirage: A Multi-Level Superoptimizer for Tensor Programs](https://mlsanity.com/paper/2405.05751/) (2024)
- [Alpa: Automating Inter- and Intra-Operator Parallelism for Distributed Deep Learning](https://mlsanity.com/paper/2201.12023/) (2022)
- [Automatic Horizontal Fusion for GPU Kernels](https://mlsanity.com/paper/2007.01277/) (2020)
- [Nimble: Lightweight and Parallel GPU Task Scheduling for Deep Learning](https://mlsanity.com/paper/2012.02732/) (2020)
- [Taming the Titans: A Survey of Efficient LLM Inference Serving](https://mlsanity.com/paper/2504.19720/) (2025)
- [AIBrix: Towards Scalable, Cost-Effective Large Language Model Inference Infrastructure](https://mlsanity.com/paper/2504.03648/) (2025)
- [ParallelKittens: Systematic and Practical Simplification of Multi-GPU AI Kernels](https://mlsanity.com/paper/2511.13940/) (2025)
- [A Framework for Fine-Grained Synchronization of Dependent GPU Kernels](https://mlsanity.com/paper/2305.13450/) (2023)
- [Lightning: Scaling the GPU Programming Model Beyond a Single GPU](https://mlsanity.com/paper/2202.05549/) (2022)

## Abstract (arXiv)

The evolution of compute infrastructure has transformed multi-GPU systems into tightly integrated shared-memory structures. However, current software still mostly treats these coherent interconnects simply as high-speed networks. Simultaneously, the demand for serving Large Language Models under latency constraints has shifted GPU workload optimization from being throughput-driven to latency-bound, necessitating latency-oriented scaling methods beyond Tensor Parallelism (TP). Thus, we introduce CTA-pipelining, an execution paradigm designed to exploit shared-memory multi-GPU systems. As a latency-oriented spatial scaling technique, CTA-pipelining leverages dependencies at the Cooperative Thread Array level, enabling concurrent execution of dependent kernels across GPUs. We demonstrate its capability using CUTLASS, cuBLAS, and NCCL libraries on 8-GPU H200 and B200 systems. Results show on 2-layer GEMM, representing the MLP operation, CTA-pipelining reduces latency by up to 31.8% compared to micro-batching, and 29.6% compared to TP. It can also be combined with TP as an orthogonal scaling dimension to further push the latency boundary.