NVIDIA GeForce GTX 680 vs AMD Radeon HD 6850 1440SP Edition
We compared two Desktop platform GPUs: 2GB VRAM GeForce GTX 680 and 1024MB VRAM Radeon HD 6850 1440SP Edition to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA GeForce GTX 680 's Advantages
Boost Clock1058MHz
More VRAM (2GB vs 1024GB)
Larger VRAM bandwidth (192.3GB/s vs 128.0GB/s)
96 additional rendering cores
AMD Radeon HD 6850 1440SP Edition 's Advantages
Lower TDP (151W vs 195W)
Score
Benchmark
FP32 (float)
GeForce GTX 680
+55%
3.25 TFLOPS
Radeon HD 6850 1440SP Edition
2.088 TFLOPS
Graphics Card
Mar 2012
Release Date
Aug 2012
GeForce 600
Generation
Northern Islands
Desktop
Type
Desktop
PCIe 3.0 x16
Bus Interface
PCIe 2.0 x16
Clock Speeds
1006 MHz
Base Clock
-
1058 MHz
Boost Clock
-
1502 MHz
Memory Clock
1000 MHz
Memory
2GB
Memory Size
1024MB
GDDR5
Memory Type
GDDR5
256bit
Memory Bus
256bit
192.3GB/s
Bandwidth
128.0GB/s
Render Config
-
Compute Units
18
-
-
-
1536
Shading Units
1440
128
TMUs
72
32
ROPs
32
-
-
-
-
-
-
16 KB (per SMX)
L1 Cache
8 KB (per CU)
512 KB
L2 Cache
512 KB
-
-
-
Theoretical Performance
33.86 GPixel/s
Pixel Rate
23.20 GPixel/s
135.4 GTexel/s
Texture Rate
52.20 GTexel/s
-
-
-
3.250 TFLOPS
FP32 (float)
2.088 TFLOPS
135.4 GFLOPS
FP64 (double)
417.6 GFLOPS
Board Design
195W
TDP
151W
450 W
Suggested PSU
450 W
2x DVI
1x HDMI 1.4a
1x DisplayPort 1.2
Outputs
2x DVI
1x HDMI 1.3a
1x DisplayPort 1.1
2x 6-pin
Power Connectors
2x 6-pin
Graphics Processor
GK104
GPU Name
Cypress
GK104-400-A2
GPU Variant
Cypress PRO
(215-0735043)
Kepler
Architecture
TeraScale 2
TSMC
Foundry
TSMC
28 nm
Process Size
40 nm
3.54 billion
Transistors
2.154 billion
294 mm²
Die Size
334 mm²
Graphics Features
12 (11_0)
DirectX
11.2 (11_0)
4.6
OpenGL
4.4
3.0
OpenCL
1.2
1.1
Vulkan
N/A
3.0
CUDA
-
5.1
Shader Model
5.0
