NVIDIA GeForce RTX 4070 GDDR6 vs NVIDIA GeForce RTX 3070
We compared two Desktop platform GPUs: 12GB VRAM GeForce RTX 4070 GDDR6 and 8GB VRAM GeForce RTX 3070 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA GeForce RTX 4070 GDDR6 's Advantages
Released 3 years and 11 months late
Boost Clock has increased by 43% (2475MHz vs 1725MHz)
More VRAM (12GB vs 8GB)
Larger VRAM bandwidth (480.0GB/s vs 448.0GB/s)
Lower TDP (200W vs 220W)
Score
Benchmark
FP32 (float)
GeForce RTX 4070 GDDR6
+43%
29.15 TFLOPS
GeForce RTX 3070
20.31 TFLOPS
Graphics Card
Aug 2024
Release Date
Sep 2020
GeForce 40
Generation
GeForce 30
Desktop
Type
Desktop
PCIe 4.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
1920 MHz
Base Clock
1500 MHz
2475 MHz
Boost Clock
1725 MHz
2500 MHz
Memory Clock
1750 MHz
Memory
12GB
Memory Size
8GB
GDDR6
Memory Type
GDDR6
192bit
Memory Bus
256bit
480.0GB/s
Bandwidth
448.0GB/s
Render Config
-
-
-
46
SM Count
46
5888
Shading Units
5888
184
TMUs
184
64
ROPs
96
184
Tensor Cores
184
46
RT Cores
46
128 KB (per SM)
L1 Cache
128 KB (per SM)
36 MB
L2 Cache
4 MB
-
-
-
Theoretical Performance
158.4 GPixel/s
Pixel Rate
165.6 GPixel/s
455.4 GTexel/s
Texture Rate
317.4 GTexel/s
29.15 TFLOPS
FP16 (half)
20.31 TFLOPS
29.15 TFLOPS
FP32 (float)
20.31 TFLOPS
455.4 GFLOPS
FP64 (double)
317.4 GFLOPS
Board Design
200W
TDP
220W
550 W
Suggested PSU
550 W
1x HDMI 2.1
3x DisplayPort 1.4a
Outputs
1x HDMI 2.1
3x DisplayPort 1.4a
1x 16-pin
Power Connectors
1x 12-pin
Graphics Processor
AD104
GPU Name
GA104
AD104-2XX-A1
GPU Variant
GA104-300-A1
Ada Lovelace
Architecture
Ampere
TSMC
Foundry
Samsung
5 nm
Process Size
8 nm
35.8 billion
Transistors
17.4 billion
294 mm²
Die Size
392 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
12 Ultimate (12_2)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.3
8.9
CUDA
8.6
6.8
Shader Model
6.6
