GeForce RTX 3070 vs Radeon RX 470
We compared two Desktop platform GPUs: 8GB VRAM GeForce RTX 3070 and 4GB VRAM Radeon RX 470 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
GeForce RTX 3070 's Advantages
Released 4 years and 1 months late
Boost Clock has increased by 43% (1725MHz vs 1206MHz)
More VRAM (8GB vs 4GB)
Larger VRAM bandwidth (448.0GB/s vs 211.2GB/s)
3840 additional rendering cores
Radeon RX 470 's Advantages
Lower TDP (120W vs 220W)
Score
Benchmark
FP32 (float)
GeForce RTX 3070
+311%
20310
Radeon RX 470
4940
3DMark Time Spy
GeForce RTX 3070
+270%
13641
Radeon RX 470
3679
3DMark Time Spy Extreme
GeForce RTX 3070
+297%
6762
Radeon RX 470
1701
Shadow of the Tomb Raider 2160p
GeForce RTX 3070
+341%
53
Radeon RX 470
12
Shadow of the Tomb Raider 1440p
GeForce RTX 3070
+280%
95
Radeon RX 470
25
Shadow of the Tomb Raider 1080p
GeForce RTX 3070
+260%
144
Radeon RX 470
40
GTA 5 1440p
GeForce RTX 3070
+208%
105
Radeon RX 470
34
GTA 5 1080p
GeForce RTX 3070
+62%
159
Radeon RX 470
98
Graphics Card
Sep 2020
Release Date
Aug 2016
GeForce 30
Generation
Arctic Islands
Desktop
Type
Desktop
PCIe 4.0 x16
Bus Interface
PCIe 3.0 x16
Clock Speeds
1500 MHz
Base Clock
926 MHz
1725 MHz
Boost Clock
1206 MHz
1750 MHz
Memory Clock
1650 MHz
Memory
8GB
Memory Size
4GB
GDDR6
Memory Type
GDDR5
256bit
Memory Bus
256bit
448.0GB/s
Bandwidth
211.2GB/s
Render Config
-
Compute Units
32
46
SM Count
-
5888
Shading Units
2048
184
TMUs
128
96
ROPs
32
184
Tensor Cores
-
46
RT Cores
-
128 KB (per SM)
L1 Cache
16 KB (per CU)
4 MB
L2 Cache
2 MB
Theoretical Performance
165.6 GPixel/s
Pixel Rate
38.59 GPixel/s
317.4 GTexel/s
Texture Rate
154.4 GTexel/s
20.31 TFLOPS
FP16 (half)
4.940 TFLOPS
20.31 TFLOPS
FP32 (float)
4.940 TFLOPS
317.4 GFLOPS
FP64 (double)
308.7 GFLOPS
Board Design
220W
TDP
120W
550 W
Suggested PSU
300 W
1x HDMI 2.1
3x DisplayPort 1.4a
Outputs
1x HDMI 2.0b
3x DisplayPort 1.4a
1x 12-pin
Power Connectors
1x 6-pin
Graphics Processor
GA104
GPU Name
Ellesmere
GA104-300-A1
GPU Variant
Polaris 10 PRO
(215-0876204)
Ampere
Architecture
GCN 4.0
Samsung
Foundry
GlobalFoundries
8 nm
Process Size
14 nm
17.4 billion
Transistors
5.7 billion
392 mm²
Die Size
232 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
12 (12_0)
4.6
OpenGL
4.6
3.0
OpenCL
2.1
1.3
Vulkan
1.2
8.6
CUDA
-
6.6
Shader Model
6.4