NVIDIA RTX A6000 GPU: Specs, VRAM, Price & Benchmarks (2026)
The NVIDIA RTX A6000 (48GB GDDR6 with ECC) launched in 2020. This guide covers its full specs, VRAM, SXM/PCIe differences where they apply, AI performance, and live per-GPU rental pricing on Aquanode.
Short answer on cost: the RTX A6000 rents from $0.550 per GPU per hour on Aquanode.
How much VRAM does the RTX A6000 have?
The RTX A6000 has 48GB GDDR6 with ECC, with 768 GB/s of peak memory bandwidth.
- VRAM: RTX A6000 48GB GDDR6 with ECC vs L40S 48GB GDDR6 with ECC
- VRAM: RTX A6000 48GB GDDR6 with ECC vs RTX 4090 24GB GDDR6X
What fits in 48GB GDDR6 with ECC of VRAM
| Model | Precision | Fits? |
|---|---|---|
| Llama 3 8B | BF16 | ~16GB. Fits with headroom. |
| Llama 3.1 70B | BF16 | ~140GB. Does not fit on one card. |
| Llama 3.1 70B | INT4 | ~35-40GB. Does not fit on a single 48GB card without further offload. |
| Mistral 7B | BF16 | ~14GB. Fits comfortably. |
Approximate, based on published parameter counts and standard bytes-per-parameter rules of thumb (FP16 ≈ 2 bytes/param, INT4 ≈ 0.5-0.6 bytes/param). Real footprint also depends on KV-cache size and framework overhead.
What can the RTX A6000 run?
Popular open models from small to frontier scale, with the memory each needs and how many RTX A6000 cards (48GB GDDR6 with ECC each) that takes.
| Model | As published | FP8 | INT4 |
|---|---|---|---|
| Qwen/Qwen3-8B 8.2B | BF16: ~18.3 GB, 1 GPU | FP8: not supported | INT4: ~4.6 GB, 1 GPU |
| Qwen/Qwen2.5-14B-Instruct 14.8B | BF16: ~33 GB, 1 GPU | FP8: not supported | INT4: ~8.3 GB, 1 GPU |
| Qwen/Qwen3-32B 32.8B | BF16: ~73.2 GB, 2 GPUs | FP8: not supported | INT4: ~18.3 GB, 1 GPU |
| Qwen/Qwen-72B 72.3B | BF16: ~162 GB, 4 GPUs | FP8: not supported | INT4: ~40.4 GB, 1 GPU |
| MiniMaxAI/MiniMax-M2.7 228.7B | FP8: not supported | – | INT4: ~128 GB, 3 GPUs |
| deepseek-ai/DeepSeek-R1 684.5B | FP8: not supported | – | INT4: ~383 GB, 8 GPUs |
Estimates: weights at the stated precision plus a flat 20% for KV cache and overhead, at a moderate context length. A dash means the precision is not offered for that model (it is already published at that size). INT4 needs a published quantized checkpoint. Open any model for a per-GPU breakdown, or use the RTX A6000 VRAM calculator.
RTX A6000 VRAM calculator: check which models fit in its memory at each precision.
All models that fit in 48 GB: the open models whose weights and overhead fit, at native, FP8 and INT4 precision.
RTX A6000 specs
| Architecture | NVIDIA, launched 2020 |
| VRAM | 48GB GDDR6 with ECC |
| Memory bandwidth | 768 GB/s |
| Interconnect | NVLink, 2-way |
| TDP | 300W |
| Form factor | PCIe, dual-slot |
Specs sourced from the vendor's public datasheet/product page. See the source.
Related reading: RTX A6000 guide.
GPU Glossary: What is VRAM?, Tensor Cores, CUDA Cores, TFLOPS, NVLink vs PCIe
RTX A6000 AI performance
A workstation-class card with 48GB VRAM at a lower TDP (300W) than the data-center accelerators, making it a reasonable single-card fit for 7-13B fine-tuning and inference without needing a rack-scale power/cooling budget.
- Memory bandwidth: 768 GB/s
Ampere-generation part with no FP8 tensor cores (FP8 arrived with the following Ada/Hopper generation), so FP8-quantized serving isn't an option, and 768 GB/s bandwidth is well below what the data-center HBM parts offer for large-batch throughput.
The cheapest RTX A6000 offer right now ($0.550/GPU/hr) is about 12% below the market median of $0.627/GPU/hr.
RTX A6000 price: what does it cost?
Buying. $4,650 (launch MSRP). The newer Ada-generation card ("6000ADA" / RTX 6000 Ada below) launched separately at a higher price point.
Renting. The cheapest current on-demand rate for the RTX A6000 on Aquanode is $0.550/GPU/hr. Live rates range from $0.550 to $0.831 per GPU per hour, with a median of $0.627/GPU/hr. Billed by the offer's own terms; the table below shows every live rate.
| Region | $/GPU/hr | Available | VRAM | vCPU | RAM |
|---|---|---|---|---|---|
| Kansas City, Us | $0.550 | 1 | 48 GB | 6 | 32 GB |
| Canada | $0.556 | 1 | 48 GB | 28 | 58 GB |
| Romania | $0.655 | 2 | 48 GB | 4.5 | 47 GB |
| Finland | $0.911 | 4 | 48 GB | 10 | 60 GB |
RTX A6000 price history
Aquanode stores one snapshot of its GPU price index per UTC day. For the RTX A6000 that is 26 days so far, 2026-10-06 to 2026-10-31, so this is a short history, not a long-run trend. The lowest per-GPU rate went from $0.363 on 2026-10-06 to $0.550 on 2026-10-31.
| Day (UTC) | Lowest per GPU hour | Median per GPU hour | Data-center lowest per GPU hour | Offers |
|---|---|---|---|---|
| 2026-10-31 | $0.550 | $0.627 | $0.550 | 20 |
| 2026-10-30 | $0.550 | $0.621 | $0.550 | 25 |
| 2026-10-29 | $0.550 | $0.627 | $0.550 | 19 |
| 2026-10-28 | $0.550 | $0.627 | $0.550 | 22 |
| 2026-10-27 | $0.550 | $0.627 | $0.550 | 18 |
| 2026-10-26 | $0.550 | $0.627 | $0.550 | 19 |
| 2026-10-25 | $0.550 | $0.627 | $0.550 | 18 |
| 2026-10-24 | $0.550 | $0.627 | $0.550 | 17 |
| 2026-10-23 | $0.550 | $0.616 | $0.550 | 18 |
| 2026-10-22 | $0.550 | $0.605 | $0.550 | 17 |
| 2026-10-21 | $0.550 | $0.625 | $0.550 | 24 |
| 2026-10-20 | $0.550 | $0.627 | $0.550 | 17 |
| 2026-10-19 | $0.550 | $0.625 | $0.550 | 19 |
| 2026-10-18 | $0.550 | $0.616 | $0.550 | 14 |
| 2026-10-17 | $0.550 | $0.616 | $0.550 | 18 |
| 2026-10-16 | $0.363 | $0.627 | $0.550 | 27 |
| 2026-10-15 | $0.363 | $0.627 | $0.550 | 36 |
| 2026-10-14 | $0.363 | $0.627 | $0.550 | 32 |
| 2026-10-13 | $0.363 | $0.627 | $0.550 | 35 |
| 2026-10-12 | $0.363 | $0.627 | $0.550 | 39 |
| 2026-10-11 | $0.363 | $0.627 | $0.550 | 35 |
| 2026-10-10 | $0.363 | $0.627 | $0.550 | 37 |
| 2026-10-09 | $0.363 | $0.627 | $0.550 | 33 |
| 2026-10-08 | $0.363 | $0.583 | $0.550 | 35 |
| 2026-10-07 | $0.363 | $0.583 | $0.550 | 36 |
| 2026-10-06 | $0.363 | $0.583 | $0.550 | 32 |
Each row is the stored daily snapshot of the live index, copied as recorded. A dash means that day stored no figure. The same series is available as JSON and summarised in the monthly GPU price report.
How this price is calculated
All prices on this page are normalized to a per-GPU hourly rate using each offer's authoritative GPU count, so that raw price is divided by the number of GPUs it actually covers; some offers report price as already per-GPU, so those are used as-listed. An offer with a missing, zero, or invalid GPU count is excluded entirely rather than published at a guessed rate.
No offers were excluded from this snapshot for a missing or invalid price. No offers were dropped as price outliers in this snapshot.
Only the cheapest qualifying offer per provider is shown in the table above. This page regenerates at most once per hour.
RTX A6000 vs L40S: how do they compare?
- VRAM: RTX A6000 48GB GDDR6 with ECC vs L40S 48GB GDDR6 with ECC
- Memory bandwidth: 768 GB/s vs 864 GB/s
On Aquanode right now, RTX A6000 starts at $0.550/GPU/hr against L40S's $0.825/GPU/hr, about 33% less.
RTX A6000 vs RTX 4090: how do they compare?
- VRAM: RTX A6000 48GB GDDR6 with ECC vs RTX 4090 24GB GDDR6X
- Memory bandwidth: 768 GB/s vs 1,008 GB/s
On Aquanode right now, RTX 4090 starts at $0.393/GPU/hr against RTX A6000's $0.550/GPU/hr, about 29% less.
Compare the RTX A6000 with other GPUs
Compare RTX A6000 with
- B200 vs RTX A6000
- H200 vs RTX A6000
- H100 vs RTX A6000
- A100 vs RTX A6000
- DGX A100 vs RTX A6000
- RTX A6000 vs V100
- AMD MI300X vs RTX A6000
- L40S vs RTX A6000
- L40 vs RTX A6000
- A40 vs RTX A6000
- RTX A5000 vs RTX A6000
- RTX A4000 vs RTX A6000
- L4 vs RTX A6000
- RTX A6000 vs T4
- RTX 6000 Ada vs RTX A6000
- RTX A6000 vs RTX PRO 6000
- RTX A6000 vs RTX PRO 6000 WS
- RTX A6000 vs RTX PRO 6000 SE
- RTX A6000 vs RTX PRO 5000
- RTX 5090 vs RTX A6000
- RTX 5080 vs RTX A6000
- RTX 5070 Ti vs RTX A6000
- RTX 5070 vs RTX A6000
- RTX 5060 Ti vs RTX A6000
- RTX 4090 vs RTX A6000
- RTX 4080 Super vs RTX A6000
- RTX 4080 vs RTX A6000
- RTX 4070 Ti vs RTX A6000
- RTX 4070 Super vs RTX A6000
- RTX 4070 vs RTX A6000
- RTX 4060 Ti vs RTX A6000
- RTX 3090 vs RTX A6000
- RTX 3080 vs RTX A6000
- RTX 3070 vs RTX A6000
- RTX 3060 vs RTX A6000
- B300 vs RTX A6000
- AMD MI355X vs RTX A6000
- AMD MI325X vs RTX A6000
- H100 NVL vs RTX A6000
- RTX A6000 vs RTX PRO 4500
- RTX A6000 vs RTX PRO 4500 SE
- RTX A6000 vs RTX PRO 4000
- RTX 5880 Ada vs RTX A6000
- RTX 5000 Ada vs RTX A6000
- RTX 4000 Ada vs RTX A6000
- RTX 4000 SFF Ada vs RTX A6000
- RTX 2000 Ada vs RTX A6000
- RTX 4070 Ti Super vs RTX A6000
- RTX A4500 vs RTX A6000
- RTX 2080 Ti vs RTX A6000
- RTX 3080 Ti vs RTX A6000
- RTX 3070 Ti vs RTX A6000
- RTX 3060 Ti vs RTX A6000
- GTX 1080 Ti vs RTX A6000
- RTX 4060 vs RTX A6000
- RTX 5060 vs RTX A6000
- GTX 1660 Super vs RTX A6000
- RTX 2060 vs RTX A6000
- Radeon RX 9070 XT vs RTX A6000
- Radeon RX 6700 XT vs RTX A6000
- RTX 6000 vs RTX A6000
- A16 vs RTX A6000
- P40 vs RTX A6000
- P4 vs RTX A6000
- Quadro P2000 vs RTX A6000
- Quadro M4000 vs RTX A6000
Get notified when the price drops
GPU supply moves hourly. Tell us what you're waiting for and we'll email you when a matching offer appears across any provider we track.
Rent a RTX A6000 on Aquanode
- On-demand instances from $0.550/GPU/hr, billed by the provider's own terms, with no hardware procurement or long-term commitment.
- 24 live offers across 4 regions today.
- Set a price/availability alert above to hear the moment a cheaper or newly-available RTX A6000 offer appears.
- Compare every RTX A6000 offer side by side, or browse the full multi-provider GPU marketplace.
Good for
A workstation-class card with 48GB VRAM at a lower TDP (300W) than the data-center accelerators, making it a reasonable single-card fit for 7-13B fine-tuning and inference without needing a rack-scale power/cooling budget.
Not good for
Ampere-generation part with no FP8 tensor cores (FP8 arrived with the following Ada/Hopper generation), so FP8-quantized serving isn't an option, and 768 GB/s bandwidth is well below what the data-center HBM parts offer for large-batch throughput.
RTX A6000 FAQs
How much VRAM does the RTX A6000 have?
The RTX A6000 has 48GB GDDR6 with ECC, with 768 GB/s of peak memory bandwidth.
What is the NVIDIA RTX A6000?
The NVIDIA RTX A6000 is a GPU released in 2020, with 48GB GDDR6 with ECC of memory and a 300W power envelope. See the full spec table above for interconnect, form factor and tensor-throughput details.
How much does it cost to rent a RTX A6000?
Live RTX A6000 rental prices currently range from $0.550 to $0.831 per GPU per hour, with a median of $0.627 per GPU per hour.
What's the cheapest RTX A6000 rate?
The lowest current RTX A6000 rate on Aquanode is $0.550 per GPU per hour in Kansas City, Us.
How does the RTX A6000 compare to the L40S?
VRAM: RTX A6000 48GB GDDR6 with ECC vs L40S 48GB GDDR6 with ECC Memory bandwidth: 768 GB/s vs 864 GB/s On Aquanode right now, RTX A6000 starts at $0.550/GPU/hr against L40S's $0.825/GPU/hr, about 33% less. See the full RTX A6000 vs L40S comparison for a shared-provider price breakdown.
How does the RTX A6000 compare to the RTX 4090?
VRAM: RTX A6000 48GB GDDR6 with ECC vs RTX 4090 24GB GDDR6X Memory bandwidth: 768 GB/s vs 1,008 GB/s On Aquanode right now, RTX 4090 starts at $0.393/GPU/hr against RTX A6000's $0.550/GPU/hr, about 29% less. See the full RTX A6000 vs RTX 4090 comparison for a shared-provider price breakdown.
What is the RTX A6000 good for?
A workstation-class card with 48GB VRAM at a lower TDP (300W) than the data-center accelerators, making it a reasonable single-card fit for 7-13B fine-tuning and inference without needing a rack-scale power/cooling budget.
What are the RTX A6000's limitations?
Ampere-generation part with no FP8 tensor cores (FP8 arrived with the following Ada/Hopper generation), so FP8-quantized serving isn't an option, and 768 GB/s bandwidth is well below what the data-center HBM parts offer for large-batch throughput.
Is renting cheaper than buying?
Renting avoids the upfront hardware cost and lets you match spend to actual usage. A rented RTX A6000 at $0.550/hr only costs money while it's running, whereas buying ties up capital in hardware that keeps depreciating whether it's in use or not. Buying outright runs $4,650 (launch MSRP). Which is cheaper depends on how continuously you'd run it; short or bursty workloads usually favor renting.
How is the RTX A6000 price calculated?
All prices are normalized to a per-GPU hourly rate using each offer's authoritative GPU count, which the raw price is divided by; some offers report price as already per-GPU. Offers whose price can't be safely normalized, or whose rate is an extreme outlier against the rest of the market, are excluded.
RTX A6000 price by region
Related guides
Other models in the same generation, then the rest of the GPU index.
