Unverified50% confidenceFactExact time
A user running 16× RTX 3090 GPUs distributed across multiple Lenovo P620-based nodes interconnected via a 100Gbit network is able to run MiniMax M2, Qwen 235B, and Qwen 397B.
1
Sources
50%
Confidence
Long-term
Relevance
10/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA mid-range machine equipped with a 5060Ti graphics card, 16GB of VRAM, and 32GB DDR4 RAM can fully run the ComfyUI workflow collection combining Qwen3 and MiniMax H3.72% similarUnverifiedQwen3 27B features 64 layers, a hidden dimension of 5120, and a native context length of 262,144 tokens.72% similarUnverifiedUsing NVIDIA B300 chips in a 1-megawatt data center can achieve a total throughput of approximately 450,000 tokens/s, which at 340 tokens/s per user amounts to approximately 1,324 concurrent active users.70% similarVerifiedThe RTX 4090 has 24GB of VRAM and can smoothly run most 13B–34B quantized models70% similarUnverifiedQwen3-8B Flash performs comparably to DeepSeek's 27B model, has a native context of 256K expandable to 1M, and the official website already provides an API interface with 1 million context.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/974690API
curl https://kongchang.com/api/v1/knowledge/claims/974690MCP
get_claim(id=974690)