diff --git a/python/sglang/srt/distributed/bootstrap.py b/python/sglang/srt/distributed/bootstrap.py index 2f025b88ec..6b3932995d 100644 --- a/python/sglang/srt/distributed/bootstrap.py +++ b/python/sglang/srt/distributed/bootstrap.py @@ -168,6 +168,19 @@ def init_parallel_runtime( ): _prewarm_tp_lm_head_all_to_all() + # EPLB uses the default process group, not the TP/EP coordinators. + # Limit warmup to static, single-node full-EP layouts for now. + if ( + device == "cuda" + and backend == "nccl" + and get_exec().moe.enable_eplb + and get_exec().moe.elastic_ep_backend is None + and parallel.nnodes == 1 + and parallel.pp_size == 1 + and parallel.moe_ep_size == parallel.tp_size > 1 + ): + _prewarm_eplb_p2p() + logger.info(f"Init parallel ends. elapsed={time.perf_counter() - tic:.2f} s") @@ -350,6 +363,41 @@ def _prewarm_nccl() -> None: ) +def _prewarm_eplb_p2p() -> None: + """Materialize EPLB's default-group NCCL transports before KV sizing.""" + warmup_start = time.perf_counter() + rank = dist.get_rank() + world_size = dist.get_world_size() + free_before, _ = torch.cuda.mem_get_info() + # Exercise bulk transfers as well as communicator creation. + bytes_per_peer = 4 << 20 + send = torch.empty(bytes_per_peer, dtype=torch.uint8, device="cuda") + recv = torch.empty( + (world_size - 1, bytes_per_peer), dtype=torch.uint8, device="cuda" + ) + ops = [] + for slot, peer in enumerate(p for p in range(world_size) if p != rank): + ops.extend( + [ + dist.P2POp(dist.isend, send, peer), + dist.P2POp(dist.irecv, recv[slot], peer), + ] + ) + for work in dist.batch_isend_irecv(ops): + work.wait() + current_platform.synchronize() + del ops, send, recv + current_platform.empty_cache() + free_after, _ = torch.cuda.mem_get_info() + logger.info( + "EPLB default-group P2P warmup completed in %.3fs " + "(world_size=%d, free memory delta=%.2f MiB)", + time.perf_counter() - warmup_start, + world_size, + (free_before - free_after) / (1 << 20), + ) + + def _prewarm_tp_lm_head_all_to_all() -> None: """Materialize PyNCCL P2P resources before model-memory accounting.""" warmup_start = time.perf_counter()