NVIDIA GeForce RTX 4070 vs NVIDIA GeForce RTX 3090
We compared two Desktop platform GPUs: 12GB VRAM GeForce RTX 4070 and 24GB VRAM GeForce RTX 3090 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA GeForce RTX 4070 's Advantages
Released 2 years and 7 months late
Boost Clock has increased by 46% (2475MHz vs 1695MHz)
Lower TDP (200W vs 350W)
NVIDIA GeForce RTX 3090 's Advantages
More VRAM (24GB vs 12GB)
Larger VRAM bandwidth (936.2GB/s vs 504.2GB/s)
4608 additional rendering cores
Score
Benchmark
FP32 (float)
GeForce RTX 4070
29.15 TFLOPS
GeForce RTX 3090
+22%
35.58 TFLOPS
3DMark Time Spy
GeForce RTX 4070
17887
GeForce RTX 3090
+11%
19893
3DMark Time Spy Extreme
GeForce RTX 4070
7907
GeForce RTX 3090
+29%
10258
3DMark Speed Way
GeForce RTX 4070
+2%
5508
GeForce RTX 3090
5383
Shadow of the Tomb Raider 2160p
GeForce RTX 4070
82
GeForce RTX 3090
+13%
93
Shadow of the Tomb Raider 1440p
GeForce RTX 4070
+3%
160
GeForce RTX 3090
154
Shadow of the Tomb Raider 1080p
GeForce RTX 4070
+28%
256
GeForce RTX 3090
200
Cyberpunk 2077 2160p
GeForce RTX 4070
54
GeForce RTX 3090
+22%
66
Cyberpunk 2077 1440p
GeForce RTX 4070
+17%
93
GeForce RTX 3090
79
Cyberpunk 2077 1080p
GeForce RTX 4070
+14%
130
GeForce RTX 3090
114
Battlefield 5 2160p
GeForce RTX 4070
98
GeForce RTX 3090
+18%
116
Battlefield 5 1440p
GeForce RTX 4070
172
GeForce RTX 3090
+9%
189
Battlefield 5 1080p
GeForce RTX 4070
200
GeForce RTX 3090
199
GTA 5 2160p
GeForce RTX 4070
132
GeForce RTX 3090
131
GTA 5 1440p
GeForce RTX 4070
144
GeForce RTX 3090
+26%
182
GTA 5 1080p
GeForce RTX 4070
190
GeForce RTX 3090
+10%
209
Horizon Forbidden West 1080p
GeForce RTX 4070
101
GeForce RTX 3090
+3%
105
Horizon Forbidden West 1440p
GeForce RTX 4070
78
GeForce RTX 3090
+16%
91
Horizon Forbidden West 2160p
GeForce RTX 4070
49
GeForce RTX 3090
+14%
56
Graphics Card
Apr 2023
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
1395 MHz
2475 MHz
Boost Clock
1695 MHz
1313 MHz
Memory Clock
1219 MHz
Memory
12GB
Memory Size
24GB
GDDR6X
Memory Type
GDDR6X
192bit
Memory Bus
384bit
504.2GB/s
Bandwidth
936.2GB/s
Render Config
-
-
-
46
SM Count
82
5888
Shading Units
10496
184
TMUs
328
64
ROPs
112
184
Tensor Cores
328
46
RT Cores
82
128 KB (per SM)
L1 Cache
128 KB (per SM)
36 MB
L2 Cache
6 MB
-
-
-
Theoretical Performance
158.4 GPixel/s
Pixel Rate
189.8 GPixel/s
455.4 GTexel/s
Texture Rate
556.0 GTexel/s
29.15 TFLOPS
FP16 (half)
35.58 TFLOPS
29.15 TFLOPS
FP32 (float)
35.58 TFLOPS
455.4 GFLOPS
FP64 (double)
556.0 GFLOPS
Board Design
200W
TDP
350W
550 W
Suggested PSU
750 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
GA102
AD104-250-A1
GPU Variant
GA102-300-A1
Ada Lovelace
Architecture
Ampere
TSMC
Foundry
Samsung
5 nm
Process Size
8 nm
35.8 billion
Transistors
28.3 billion
295 mm²
Die Size
628 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
6.7
Shader Model
6.6
