Latest Content
DeepSeek R1 0528 icon

DeepSeek R1 0528

NVIDIA
DeepSeek-R1-0528 is a reasoning-optimized Mixture-of-Experts (MoE) large language model developed by DeepSeek AI, serving as an upgraded version of the original DeepSeek-R1. It features a 671B parameter architecture with ~37B activated parameters, built on a 61-layer transformer with 128 attention heads and a 7,168 hidden size, utilizing 256 routed experts with 8 experts activated per token. The model focuses on deep reasoning, mathematics, coding, and logical inference, with improved hallucination control. It supports a 128K token context window using YaRN-based rotary scaling, enabling extended reasoning and problem solving.
TypeMoE LLM
CapabilitiesText Generation, Instruction Following, Reasoning, Mathematical Reasoning+5 more
Paper/Blog
Paper|Blog|HF Model Card
LicenseMIT

Inference Instructions

Deploy and run this model on NVIDIA B200 GPUs using the command below. Copy the command to get started with inference.

CONSOLE
docker run --gpus all 
 --shm-size 128g 
 -p 8000:8000 
 -v ~/.cache/huggingface:/root/.cache/huggingface 
 -e HF_TOKEN='YOUR_HF_TOKEN' 
 --ipc=host 
 lmsysorg/sglang:v0.5.8-cu130 
 python3 -m sglang.launch_server 
 --model-path deepseek-ai/DeepSeek-R1-0528 
 --host 0.0.0.0 
 --port 8000 
 --max-running-requests 1024 \ --tp 8 
 --ep 8 
 --mem-fraction-static 0.90 
 --trust-remote-code

Model Benchmarks

Each model was tested with a fixed input size and total token volume while increasing concurrency to measure serving performance under load.

ITL vs Concurrency

Time to First Token

Throughput Scaling

Total Tokens/sec vs Avg TTFT

Vultr Cloud GPU

NVIDIA HGX B200

Deploy NVIDIA B200 on Vultr Cloud GPU

Deploy NVIDIA B200 instancesContact Sales