NVIDIA
Architecture
Hopper is NVIDIA's datacenter-focused architecture succeeding Ampere. Built on TSMC 4N, it introduces the Transformer Engine with automatic FP8/FP16 mixed-precision training, HBM3/HBM3e memory, and NVLink 4.0 for multi-GPU scaling. The H100 flagship delivers up to 3x the AI training performance of A100.
AI Relevance
The Transformer Engine automatically manages FP8 precision for optimal training speed without accuracy loss. With up to 141 GB HBM3e (H200), Hopper GPUs can hold the largest open-weight models entirely in GPU memory, making them the workhorse of AI datacenters.
Agentic Coding
BThis model is still usable for agentic-coding, but it is not the most specialized pick. It belongs to a current frontier family for local AI. It fits natively with comfortable headroom. Known channels: huggingface, lm-studio.
Chat
BThis model is a direct match for chat. It belongs to a current frontier family for local AI. It fits natively with comfortable headroom. Known channels: huggingface, lm-studio.
Coding
BThis model is a direct match for coding. It belongs to a current frontier family for local AI. It fits natively with comfortable headroom. Known channels: huggingface, lm-studio.
RAG
BThis model is a direct match for rag. It sits in the middle of the current model mix. It fits natively with comfortable headroom. Known channels: huggingface, ollama, lm-studio.
Reasoning
BThis model is a direct match for reasoning. It belongs to a current frontier family for local AI. It fits natively with comfortable headroom. Known channels: huggingface, lm-studio.
Mixtral 8x22B
Command A 111BMistral Small 4 119BDevstral 2 123B Instruct
Qwen3.5 122B A10B
Qwen 2.5 72BQwen 2.5 VL 72B
Llama 3.3 70BQwen3-Coder-NextQwen3.5 35B A3BQwen 2.5 Coder 32BQwen3.5 27Bgemma 3 27b itQwen3-Coder 30B A3B InstructQwen3-VL 30B A3B InstructDevstral Small 2 24B InstructDevstral Small 1.1Codestral 2 25.08Qwen3.5 9BDeepSeek R1 0528 Qwen3 8BQwen3.5 4BQwen2.5 3B Instruct
gemma 2bQwen2.5 1.5B InstructQwen 3 235B A22B
DeepSeek R1 671B
GLM-5
Kimi K2.5Mistral Large 3Qwen3-Coder 480B A35B InstructDeepSeek V3 671BQwen3.5 397B A17BLlama 4 Maverick 17B 128EJust out of reach
High-quality models that need a bit more memory
Upgrade paths
See what you unlock with more powerful hardware
Upgrade options
~$8,000 MSRP
