Content by cormacgarvey (3)

CormacGarvey examines the deployment and benchmarking of the DeepSeek R1 reasoning model on Azure ND_H100_v5 nodes using vLLM, providing practical insights into infrastructure demands and performance.

Cormac Garvey evaluates the inference performance and cost-efficiency of Llama 3.1 8B using vLLM across Azure GPU and CPU virtual machines, offering actionable benchmarks and deployment strategies for enterprise AI workloads.

Cormac Garvey offers an in-depth benchmarking study of Llama 3.1 8B model inference using vLLM on Azure ND-H100-v5 GPUs. The article explains AI inference stages, optimization techniques, and reports on throughput, latency, and deployment tips for enterprise AI practitioners.