Tech Meridian ← ENTITY INDEX
PROMY MERIDIAN RU

TOPIC · ENTITY #8861

ml.p5.48xlarge

Related event timeline, sources and context from the news index.

EVENT TIMELINE

3

COMPANIES · 1 SOURCE · AWS Machine Learning

Amazon SageMaker HyperPod Inference adds managed support for NVIDIA Dynamo

Amazon SageMaker HyperPod Inference now supports NVIDIA Dynamo via a HyperPod Inference add-on that installs Dynamo CRDs and serving/scheduling components, enabling multi-node and disaggregated LLM inference (separate prefill and decode workers) with EFA networking, coordinated GPU scheduling, and KV-aware routing. Customers can deploy via a bring-your-own DynamoGraphDeployment or an InferenceEndpointConfig with dynamoSpec, and choose engines such as vLLM, SGLang, or TensorRT-LLM on instances like ml.p5.48xlarge using AWS DLCs.

6.0

CODING · 1 SOURCE · AWS Machine Learning

Amazon SageMaker HyperPod: three PyTorch DDP checks to detect and recover from silent data corruption

Amazon describes adding three integrity checks to PyTorch DistributedDataParallel training loops running on SageMaker HyperPod to detect silent data corruption (SDC): a non-finite gradient check, a ring replay audit that recomputes micro-batches across replicas for bitwise loss/gradient comparison, and a post-update weight fingerprint comparison. When a check flags a suspect rank, the system maps the rank to a node on EKS, contains the node, and uses HyperPod node replacement and checkpoints to recover; the team tested the approach with software-injected faults on clusters of ml.p5.48xlarge (32 H100 GPUs) and published code in a companion GitHub repo.

7.0

MODELS · 1 SOURCE · AWS Machine Learning

Amazon SageMaker HyperPod and Qumulo enable multi‑Region model training without copying data

Amazon and Qumulo validated an architecture pairing Amazon SageMaker HyperPod with Cloud Native Qumulo (CNQ) and Qumulo Cloud Data Fabric (CDF) so HyperPod clusters can train from a single dataset stored in a different AWS Region without replicating data. In a cross‑Region test (hub in us-east-2, spoke in us-west-2) running a 1.02B‑parameter LLaMA v3 job on two ml.p5.48xlarge instances per cluster (16 H100 GPUs total), the remote spoke converged to hub throughput (≈115–117 samples/sec) after a short NeuralCache warmup, achieving near‑full GPU utilization thereafter.

7.0