Nvidia

RTX A6000

The RTX A6000 is Nvidia’s 48GB graphics card with a 300W board power rating, built on the Ampere GA102 architecture. Below you’ll find what it can run locally, its benchmark scores and its full spec sheet. No shop we track in Canada has a board version in stock right now, but the specs and benchmarks below still apply.

Prices update regularly and may differ at the retailer. This page contains affiliate links.

Benchmarks

Standardized benchmark scores for this chip, each compared to the fastest card we track and linked to its source. They apply to every board version.

Render

RTX A60005,285 samples/min
RTX PRO 6000 Blackwell (best)16,780 samples/min

Compute & AI

Geekbench 7 OpenCLGeekbench Browser ↗
RTX A6000193,937 points
GeForce RTX 5090 (best)365,480 points

3D graphics

3DMark Fire Strike GraphicsUL 3DMark ↗
RTX A600040,885 points
GeForce RTX 4090 (best)75,641 points
3DMark Steel NomadUL 3DMark ↗
RTX A60004,869 points
RTX PRO 6000 Blackwell (best)15,797 points

Local AI performance

What the RTX A6000 can run on your own machine, and how fast.

Largest model it can run (8K context)

Q4 (the usual choice, near-original quality)
~68B
Q8 (closer to original, half the model size)
~40B
FP16 (unquantised, for training and research)
~21B

Parameter counts the card holds with room for an 8K conversation, counting the model weights plus its attention cache. A longer conversation needs more memory, so it lowers these figures.

Speed & value

Speed, 8B model at Q4
99-116 tok/s
Memory bandwidth
768 GB/s
Price per GB of VRAM
—

Generation speed tracks memory bandwidth, because every word produced reads the whole model from memory. Shown as a range.

Models that fit, at Q4

ModelSizeMemory neededEst. speed
Qwen2 Math 72B72.7B43.9GB12-14 tok/s
Dolphin Llama3 70B71.0B42.7GB12-14 tok/s
Llama3 Chatqa 70B71.0B42.7GB12-14 tok/s

Popular models small enough for this card at Q4, one per family. Speed is left blank for mixture-of-experts models, where a bandwidth estimate is not reliable enough to publish.

Choosing a card for AI? Compare every GPU for local AI, including setups that pool the memory of two or more cards.

RTX A6000 availability

Live retailer stock for this model.

Currently out of stock

None of the board versions of the RTX A6000 are in stock at the shops we track in Canada right now. Prices refresh daily, so check back.

Browse in-stock graphics cards for AISee US prices for this card

Buying a used, refurbished or open-box RTX A6000

None of the shops we track in Canada has a new RTX A6000 in stock right now, so used, refurbished and open-box listings are the main way to buy one.

Open box:
unused or returned, sold as new but no longer sealed.
Refurbished:
repaired or checked by a seller or the maker, often with a warranty of its own.
Used:
sold as the seller describes it. Warranty depends on the brand and the seller.

On eBay

What to check before buying

  • The RTX A6000 came out in December 2020, so a used one can be over 5 years old. Ask how long and how hard it was used.
  • Ask for photos of the card itself, not stock images. Check the fans, the PCIe connector and the power sockets for damage or scorch marks.
  • Run a game or a stress test as soon as it arrives, while you can still return it. eBay's buyer protection covers an item that arrives faulty or doesn't match its listing.
  • Ask whether the warranty transfers. Some brands cover a second owner; others cover only the first buyer.
  • On a refurbished listing, check who did the work and how long their warranty runs.

More in our guide to buying new vs used PC parts.

Specifications

The reference RTX A6000 measures 267 mm long.

Performance

Boost clock
1800 MHz
Base clock
1410 MHz
Shader cores
10,752
SM count
84
Texture units (TMUs)
336
Render outputs (ROPs)
112
RT cores
84 · 2nd gen
Tensor / AI cores
336 · 3rd gen

Memory

VRAM
48GB GDDR6
Memory bus
384-bit
Bandwidth
768 GB/s
Memory speed
16 Gbps
L2 cache
6 MB

GPU die

Architecture
Ampere GA102
Process node
Samsung 8N
Foundry
Samsung
Die size
628 mm²
Transistors
28.3 billion
Generation
RTX A-series

Power

TDP
300W
Power connectors
1 8-pin

Display & outputs

Max resolution
7680 × 4320
Max displays
4
HDCP
2.2
Outputs
4 DisplayPort 1.4a
PCIe interface
PCIe 4.0 ×16

APIs & software

DirectX
12 Ultimate (12.2)
Vulkan
1.3
OpenGL
4.6
OpenCL
3.0
CUDA
Compute 8.6
Shader model
6.6

Video encoding & decoding

CodecEncodeDecode
H.264 / AVCSupportedSupported
H.265 / HEVCSupportedSupported
AV1Not supportedSupported
VP9Not supportedSupported

Encoder: NVENC 7th gen · Decoder: NVDEC 5th gen

Features

Upscaling
DLSS Super Resolution, Ray Reconstruction
Also runs
FSR 3, FSR 3 Frame Generation, XeSS 2, XeSS 2 Frame Generation
Multi-GPU
NVLink
ECC memory
Yes

Physical

Reference length
267 mm
Slot width
2 slots
Released
2020-12-15

10,752 shader cores feed a 384-bit memory interface delivering 768 GB/s of peak bandwidth. 84 RT cores handle real-time ray tracing; 336 tensor cores accelerate AI workloads and DLSS Super Resolution. Compare the RTX A6000 against consumer and workstation cards on the ranking of graphics cards for AI.

How the RTX A6000 compares

The RTX A6000 has no offer in stock in Canada right now, so these are the closest cards you can buy today, picked on memory, estimated generation speed and live prices. Gaming cards are included wherever they hold as much for less. Open a card for its full spec sheet.

Affiliate disclosure: Some links on this page are affiliate links — MaxMyBuild may earn a commission at no extra cost to you if you buy through them. As an Amazon Associate, MaxMyBuild earns from qualifying purchases. For Amazon, current price and availability are shown on the Amazon product page; product prices and availability are accurate as of the date and time of purchase and are subject to change. Prices shown for other retailers are approximate and may differ at checkout. “Search” links are non-affiliate and provided for convenience only.

Advertisement