Experimental framework improves training efficiency of large language models across multiple data centers, highlighting reduced communication overhead.
The proliferation of Large Language Models (LLMs) with exponentially growing parameters is making cross-data center (DC) training an inevitable trend. However, viable strategies for extending single-DC training frameworks to multi-DC environments remain underdeveloped. We experimentally demonstrate, for the first time, a high-performance geo-distributed LLMs training framework across multiple DCs interconnected by a lossless, remote direct memory access (RDMA) enabled Datacenter Optical Transport Network (DC-OTN). An enhanced pipeline parallelism scheme is implemented within the Ascend full-stack environment of Huawei, which effectively eliminates the impact of cross-DC communication overhead on training efficiency. The overlapped computation and cross-DC communication is achieved with constraint cross-DC bandwidth and High Bandwidth Memory (HBM), reducing computation bubble ratio by up to 78.91%.
No takes yet. Share an insight, caveat, or question.
Dai et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: