GeForce GTX 670 and GeForce RTX 3090 - video card comparison
Passmark - benchmark score GeForce GTX 670 vs GeForce RTX 3090
GeForce GTX 670
5491
GeForce RTX 3090
27597
GeForce GTX 670
GeForce RTX 3090
Architecture
Kepler (2012−2018)
Ampere (2020−2022)
GPU code name
GK104
Ampere GA102
Market segment
Desktop
Desktop
Release date
May 2012
September 2020
Bus support
PCI Express 3.0
-
Interface
PCIe 3.0 x16
PCIe 4.0 x16
Supplementary power connectors
Two 6-pin
1x 12-pin
SLI options
+
-
Boost clock speed
980 MHz
1700 MHz
Core clock speed
915 MHz
1400 MHz
CUDA cores
1344
-
Floating-point performance
2,459.5 gflops
-
Manufacturing process technology
28 nm
8 nm
Number of transistors
3,540 million
28,300 million
Pipelines / CUDA cores
1344
10496
Power consumption (TDP)
170 Watt
350 Watt
Maximum GPU temperature
97 °C
-
Texture fill rate
102.5 billion/sec
556.0
Memory type
GDDR5
GDDR6X
Maximum RAM amount
2 GB
24 GB
Memory bus width
256-bit GDDR5
384 Bit
Memory clock speed
6.0 GB/s
19500 MHz
Memory bandwidth
192.2 GB/s
936.2 GB/s
Shared memory
-
-
Display Connectors
One Dual Link DVI-I, One Dual Link DVI-D, One HDMI, One DisplayPort
1x HDMI, 3x DisplayPort
Multi monitor support
4 displays
-
HDMI
+
+
HDCP
+
-
Maximum VGA resolution
2048x1536
-
Audio input for HDMI
Internal
-
DirectX
12 (11_0)
12 Ultimate (12_2)
CUDA
+
8.5
OpenCL
1.2
2.0
OpenGL
4.2
4.6
Shader Model
5.1
6.5
Vulkan
1.1.126
1.2
3D Blu-Ray
+
-
3D Gaming
+
-
3D Vision
+
-