NVIDIA GeForce RTX 4070 GDDR6 vs ATI Radeon HD 4890
We compared two Desktop platform GPUs: 12GB VRAM GeForce RTX 4070 GDDR6 and 1024MB VRAM Radeon HD 4890 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 15 years and 4 months late
Boost Clock2475MHz
More VRAM (12GB vs 1024GB)
Larger VRAM bandwidth (480.0GB/s vs 124.8GB/s)
5088 additional rendering cores
ATI Radeon HD 4890 's Advantages
Lower TDP (190W vs 200W)
Score
Benchmark
FP32 (float)
GeForce RTX 4070 GDDR6
+2043%
29.15 TFLOPS
Radeon HD 4890
1.36 TFLOPS
Graphics Card
Aug 2024
Release Date
Apr 2009
GeForce 40
Generation
Radeon R700
Desktop
Type
Desktop
PCIe 4.0 x16
Bus Interface
PCIe 2.0 x16
Clock Speeds
1920 MHz
Base Clock
-
2475 MHz
Boost Clock
-
2500 MHz
Memory Clock
975 MHz
Memory
12GB
Memory Size
1024MB
GDDR6
Memory Type
GDDR5
192bit
Memory Bus
256bit
480.0GB/s
Bandwidth
124.8GB/s
Render Config
-
Compute Units
10
46
SM Count
-
5888
Shading Units
800
184
TMUs
40
64
ROPs
16
184
Tensor Cores
-
46
RT Cores
-
128 KB (per SM)
L1 Cache
16 KB (per CU)
36 MB
L2 Cache
256 KB
-
-
-
Theoretical Performance
158.4 GPixel/s
Pixel Rate
13.60 GPixel/s
455.4 GTexel/s
Texture Rate
34.00 GTexel/s
29.15 TFLOPS
FP16 (half)
-
29.15 TFLOPS
FP32 (float)
1360 GFLOPS
455.4 GFLOPS
FP64 (double)
272.0 GFLOPS
Board Design
200W
TDP
190W
550 W
Suggested PSU
450 W
1x HDMI 2.1
3x DisplayPort 1.4a
Outputs
2x DVI
1x S-Video
1x 16-pin
Power Connectors
2x 6-pin
Graphics Processor
AD104
GPU Name
RV790
AD104-2XX-A1
GPU Variant
RV790 XT
(215-0758016)
Ada Lovelace
Architecture
TeraScale
TSMC
Foundry
TSMC
5 nm
Process Size
55 nm
35.8 billion
Transistors
0.959 billion
294 mm²
Die Size
282 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
10.1 (10_1)
4.6
OpenGL
3.3
3.0
OpenCL
1.1
1.3
Vulkan
N/A
8.9
CUDA
-
6.8
Shader Model
4.1
