Hardware🖥️
NVIDIA A100 80GB (SXM)
Released May 14, 2020by NVIDIA
80GB HBM2e datacenter GPU: the standard for datacenter training and inference. Requires dedicated servers and enterprise power.
GPUNVIDIADatacenterTraining
Back to resources
Organizations🟢
NVIDIA
NVIDIA: CUDA-architecture GPUs are the de facto standard for local AI inference. VRAM determines which models run locally: 24GB (RTX 4090) covers ~30B quantized, 80GB (A100/H100) covers 70B+ models at full precision. The GPUs in the catalog are the child resources.
Features
- CUDA: the most widely supported inference toolkit
- Consumer GeForce: from 12GB (RTX 3060) to 32GB (RTX 5090)
- Data center: A100 80GB and H100 80GB for large models
Resources by NVIDIA
Hardware⚡
NVIDIA DGX Spark (GB10)
La workstation desktop NVIDIA per l'IA locale: superchip Grace Blackwell GB10 con 128GB di memoria unificata a 273 GB/s e circa 1 petaFLOP in FP4. CUDA completo, ~12 tok/s su modelli 27B a 4 bit, prefill veloce su contesti lunghi. Prezzo al lancio 3.999 USD, poi 4.699 USD per la carenza di memoria.
NVIDIAMemoria unificataEdge AIDesktop
Hardware🖥️
NVIDIA H100 80GB (SXM)
Released Mar 20, 2022by NVIDIA
The top-tier datacenter GPU with 80GB HBM3: the top-of-the-line for large-model training and inference. Enterprise configurations only.
GPUNVIDIADatacenterTraining
Models🎙️
NVIDIA NemotronLabs VoiceChat 11B
11B end-to-end full-duplex voice model: audio understanding and generation in a single architecture, without the cascaded ASR→LLM→TTS stack. First open full-duplex model with tool calling during conversation, ~450ms responses, interruptions, and natural turns.
VocaleOpen SourceLocaleNVIDIA
Hardware🖥️
NVIDIA RTX 3060 12GB
Released Dec 8, 2021by NVIDIA
The most accessible consumer GPU with 12GB VRAM: the entry point for local inference. Comfortably runs 7-8B quantized models.
GPUNVIDIAInferenzaBudget
Hardware🖥️
NVIDIA RTX 3090 24GB
Released Sep 17, 2020by NVIDIA
24GB VRAM at very favorable used-market prices: the cheapest way to reach 24GB for 30B models. Discontinued.
GPUNVIDIAUsato24GB
Hardware🖥️
NVIDIA RTX 4060 Ti 16GB
Released Nov 20, 2023by NVIDIA
16GB VRAM at an entry-level price: the right choice for quantized 13-14B models. Good headroom for personal use.
GPUNVIDIAInferenzaMid-range
Hardware🖥️
NVIDIA RTX 4070 Ti Super 16GB
Released Jan 31, 2024by NVIDIA
16GB VRAM with 672 GB/s bandwidth: high-end performance at a contained price. Among the best VRAM/price ratios.
GPUNVIDIAInferenzaRapporto qualità/prezzo
Hardware🖥️
NVIDIA RTX 4090
Released Oct 12, 2022by NVIDIA
The world's most powerful consumer GPU for local training and high-speed inference.
GPUTrainingInferenzaConsumer
Hardware🎮
NVIDIA RTX 5060 Ti 16GB
Il 16GB d'ingresso della generazione Blackwell: GB206, 16GB GDDR7 a 448 GB/s, 180W, PCIe 5.0 x8. Lancio a 429 USD: il modo più economico di far girare Qwen3.8-27B e i modelli 30B+ in Q4 su una singola scheda. Prezzo street 2026 più alto per la carenza di GDDR7.
NVIDIABlackwellGPU
Hardware🎮
NVIDIA RTX 5070 12GB
Blackwell GB205 di fascia media: 12GB GDDR7 su bus 192-bit (672 GB/s), 6.144 CUDA core, 250W. Supporto completo ai formati FP4/FP8 per l'inferenza. Lancio a 549 USD, marzo 2025.
NVIDIABlackwellGPU
Hardware🎮
NVIDIA RTX 5070 Ti 16GB
Blackwell GB203 da fascia alta: 16GB GDDR7 su bus 256-bit (896 GB/s), 8.960 CUDA core, 300W. Il 16GB rende possibile Qwen3.8-27B o i modelli 30B+ in Q4 con contesto ampio. Lancio a 749 USD, febbraio 2025.
NVIDIABlackwellGPU
Hardware🖥️
NVIDIA RTX 5080 16GB
Released Mar 5, 2025by NVIDIA
16GB GDDR7 with 960 GB/s bandwidth: the mid-range Blackwell GPU. Decode speed well above prior generations.
GPUNVIDIABlackwellInferenza
Hardware🖥️
NVIDIA RTX 5090 32GB
Released Jan 30, 2025by NVIDIA
The fastest consumer GPU on the planet: 32GB GDDR7 and 1,792 GB/s bandwidth. The top pick for local inference without going to the datacenter.
GPUNVIDIABlackwellHigh-end
Tags
Discover NVIDIA
Visit the official site to download, learn more and start using NVIDIA.