GitHub - PrimitiveContext/blackwell: Production LLM deployment specs for NVIDIA Blackwell GPUs (RTX Pro 6000, DGX Spark). Includes vLLM configurations, benchmarks, load balancer, and throughput calculators for NVFP4/FP8/MoE models.

PrimitiveContext / blackwell Public

Notifications You must be signed in to change notification settings
Fork 0
Star 1

Production LLM deployment specs for NVIDIA Blackwell GPUs (RTX Pro 6000, DGX Spark). Includes vLLM configurations, benchmarks, load balancer, and throughput calculators for NVFP4/FP8/MoE models.

1 star 0 forks Branches Tags Activity

Star

Notifications

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 36 Commits
DgxSpark		DgxSpark
Rtx4090Ada		Rtx4090Ada
RtxPro6000		RtxPro6000
SM12x.NVFP4w-NVFP4kv.MoE.GQA.TrtLLM-Fork/.research		SM12x.NVFP4w-NVFP4kv.MoE.GQA.TrtLLM-Fork/.research
TrtLLM.Peak.Optimized		TrtLLM.Peak.Optimized
load-balancer		load-balancer
scripts		scripts
strategies/map-reduce		strategies/map-reduce
theoretical		theoretical
.gitattributes		.gitattributes
.gitignore		.gitignore
litellm-config.yaml.template		litellm-config.yaml.template
llm-deployment.spec.yaml		llm-deployment.spec.yaml

About

Production LLM deployment specs for NVIDIA Blackwell GPUs (RTX Pro 6000, DGX Spark). Includes vLLM configurations, benchmarks, load balancer, and throughput calculators for NVFP4/FP8/MoE models.