AMD Radeon Pro Duo technical specs and benchmarks

AMD Radeon Pro Duo

Summary

AMD started Radeon Pro Duo sales 26 April 2016 at a recommended price of 1,499 USD. This is a high-end GCN 3.0 architecture desktop card based on 28 nm manufacturing process and primarily aimed at gamers. 8192 GB of High Bandwidth Memory (HBM) memory clocked at 500 GHz are supplied, and together with 2x 4096 Bit memory interface this creates a bandwidth of 512 GB/s.

Compatibility-wise, this is 2-slot card attached via PCIe 3.0 x16 interface. Its manufacturer default version has a length of 277 mm. An additional 3 8-pin power connectors are required, and power consumption is at 350 Watt.

It provides good gaming and benchmark performance at 70.36 percent of a leader's which is NVIDIA Quadro P6000.

General info

Info on video card type (desktop or notebook), its architecture and sales start time and price.
Place in performance rating
 
Perfomance to price
28.18
 
Architecture
GCN 3.0
 
Code name
Capsaicin
 
Type
Desktop
 
Design
reference
 
Release date
26 April 2016
 
Launch price (MSRP)
1,499 USD
 
of 14999 (Quadro Plex 7000)
Price
819.99 USD
 
of 4808 (Quadro P6000)

Technical specs

Quantitative parameters such as shaders number, GPU core clock, manufacturing process, texturing and calculation speed. These parameters indirectly say of video card performance, but to be more precise you have to review it's benchmark and gaming test results.
Pipelines
2x 4096
 
Compute units
128
 
Boost Clock
1000 MHz
 
Transistor count
8,900 million
 
of 21100
Manufacturing process technology
28 nm
 
of 12
Power consumption (TDP)
350 Watt
 
of 900 (Tesla S2050)
Texture fill rate
2x 256.0 billion/sec
 
Floating-point performance
2x 8,192 gflops
 

Compatibility, dimensions and requirements

Information on video card compatibility with other computer components. Useful when choosing a future computer configuration or upgrading an existing one. For desktop video cards it's interface and bus (motherboard compatibility), additional power connectors (power supply compatibility). For notebook video cards it's notebook size, connection slot and bus - if the video card is inserted into the slot instead of being soldered to the notebook motherboard.
Bus support
PCIe 3.0
 
Interface
PCIe 3.0 x16
 
Length
277 mm
 
Width
2-slot
 
Supplementary power connectors
3x 8-pin
 

Memory specs

Parameters of memory installed on the video card - type, size, bus, clock and resulting bandwidth. GPUs integrated into processors don't have dedicated memory and use a shared part of system RAM.
Memory type
High Bandwidth Memory (HBM)
 
Maximum RAM amount
8 GB
 
of 32 (Tesla M10)
Memory bus width
2x 4096 Bit
 
Memory clock speed
500 MHz
 
Memory bandwidth
512 GB/s
 
of 1024 (FirePro S9300 X2)

Video outputs and ports

Video connectors present on the video card. Mostly actual for desktop video cards since notebook ones' available video connectors vastly depend on the notebook model.
Display Connectors
1x HDMI, 3x DisplayPort
 
Eyefinity
+
 
Number of Eyefinity displays
6
 
HDMI
+
 
DisplayPort support
+
 

Technologies

Technological solutions supported by the GPU. You'll probably need this information if you require some particular technology.
AppAcceleration
+
 
CrossFire
+
 
Enduro
+
 
FRTC
+
 
FreeSync
+
 
HD3D
+
 
LiquidVR
+
 
PowerTune
+
 
TressFX
+
 
TrueAudio
+
 
ZeroCore
+
 
UVD
+
 
VCE
+
 

API support

APIs supported by video card, sometimes including their particular versions.
DirectX
DirectX® 12
 
OpenGL
4.5
 
OpenCL
2.0
 
Vulkan
+
 
Mantle
+
 

Video card benchmarks performance

Non-gaming benchmarks performance. Overall benchmark performance is measured in points in 0-100 range.
Overall benchmark performance
70.36
 
 
Passmark
9472
 
 
3DMark Fire Strike Graphics Score
27110
 
 

Mining hashrate

Video card cryptocurrency mining performance. Usually measured in megahashes per second.
Bitcoin / BTC (SHA256)
809 Mh/s
 
 

Game benchmarks (in FPS)

How good for gaming a particular GPU is. Gaming benchmark results are measured in FPS. Overall gaming performance is a number in 0-100 range, and the older the game, the less impact it has on the overall grade of the video card.
Overall gaming performance
76.34
 
 

Relative perfomance

Overall video card performance compared to nearest competitors. All GPU types participate.

NVIDIA equivalent

We believe that the nearest equivalent to Radeon Pro Duo from NVIDIA is GeForce GTX 1070 Ti, which is slower by 5%.

Choose a GPU

Compare GPUs

Compare Radeon Pro Duo with others

Other GPUs

Comments

Here you can ask a question, agree or disagree with our judgements, or report an error or mismatch.