Built on a Different Level
AI Performance
4,000 TOPS AI compute (FP4 with sparsity)
120 TFLOPS FP32 compute
5th Gen Tensor Cores 3× faster than previous gen
Massive Memory
96GB GDDR7 with ECC error correction
1,597 GB/s memory bandwidth
70B models on a single card — no multi-GPU needed
Core Configuration
24,064 CUDA Cores
PCIe Gen 5 interface — 2× bandwidth of Gen 4
600W sustained server-grade performance
Advanced Ray Tracing
355 TFLOPS ray tracing performance
4th Gen RT Cores 2× ray-triangle intersection rate
DLSS 4 Multi Frame Generation — up to 3× faster frames
The Performance Numbers
96 GB GDDR7. 4,000 AI TOPS. Built for Production.
The RTX PRO 6000 Blackwell Server Edition is the most capable single-GPU instance on E2E Cloud — engineered for sustained production AI workloads, not burst experiments. Run 70B parameter models at FP8 on a single card with 26 GB of KV cache headroom remaining. Partition into up to four isolated 24 GB MIG instances for concurrent tenants. Deploy on hardened, monitored infrastructure backed by a production SLA.
Built for Every Professional AI Workload
From local LLM inference to engineering simulation — the RTX PRO 6000 handles it all on a single card.
Full-Spectrum LLM Support. 7B to 141B. One Server GPU.
Stop splitting models across two GPUs. The RTX PRO 6000 Server Edition runs the full range — from 7B up to Mixtral 8×22B (141B total parameters) — on a single server GPU.
| Model | VRAM Usage |
|---|---|
7B Small — fast inference | ~14 GB 15% of 96GB |
13B Balanced quality | ~26 GB 27% of 96GB |
30–34B High quality | ~18 GB 19% of 96GB |
70B Production frontier | ~70 GB 73% of 96GB |
70B Max throughput | ~38 GB 40% of 96GB |
8×22B 141B total · MoE | ~71 GB 74% of 96GB |
Pricing for NVIDIA RTX PRO 6000
Access NVIDIA's most powerful server GPU with Blackwell architecture, 96GB GDDR7 memory, and cutting-edge AI performance.
On-demand — ₹180/hr per GPU
Instant access to RTX PRO 6000 with 96GB GDDR7, PCIe Gen5, and up to 4,000 TOPS AI performance. A typical 70B model fine-tuning run takes 4–8 hours on a single card.
Sign up to consoleDetailed Pricing Options
View all pricing tiers and configurations for RTX PRO 6000
| Configuration | Hourly/On-Demand | Monthly | Annually |
|---|---|---|---|
1x NVIDIA RTXPRO6000Most Popular | ₹180/hr | ₹1,15,320 | ₹13,03,400 |
2x NVIDIA RTXPRO6000 | ₹360/hr | ₹2,30,640 | ₹26,06,800 |
4x NVIDIA RTXPRO6000 | ₹720/hr | ₹4,61,280 | ₹52,13,600 |
8x NVIDIA RTXPRO6000 | ₹1,440/hr | ₹9,22,560 | ₹1,04,27,200 |
Unleash AI At Scale
Deploy RTX PRO 6000 GPUs for AI training, fine-tuning, simulation, and professional graphics — from a single card to an 8-GPU cluster. INR billing. Indian data centres. No commitment needed to start.