NVIDIA GeForce RTX 3070 Ti vs NVIDIA GeForce GTX 670

Compare NVIDIA GeForce RTX 3070 Ti 8 GB vs NVIDIA GeForce GTX 670 4 GB, specs and GPU benchmark score. Which is the better graphics card for the money?

Benchmark Score
Other Benchmarks
Specifications
Comments (0)
Popular Comparisons

GPU Comparison

Quickly search and compare graphics cards

NVIDIA GeForce GTX 670

Price

The cheapest price from our partner retailers

$ 639.00

$ 614.25 significantly less expensive

$ 1253.25

Benchmark Score

Overall Score

General gaming and workstation score

8845 points

1324% significantly better overall score

621 points

Flux Core frame rate

Volumetric ray casting test, a computationally expensive method of rendering high-quality scenes

210 FPS

180 FPS significantly higher Flux Core frame rate

30 FPS

Electron frame rate

Randomly generated noise sphere test

104 FPS

86 FPS significantly higher Electron frame rate

18 FPS

City frame rate

Procedurally generated city scene with voxel rendering

151 FPS

133 FPS significantly higher City frame rate

18 FPS

Clouds frame rate

Real-time noise calculation and ray marching test

83 FPS

69 FPS significantly higher Clouds frame rate

14 FPS

Free GPU Benchmark

Want to compare your graphics card against the GeForce RTX 3070 Ti and the GeForce GTX 670? Download our free and quick PC Performance Test.

Download GPU Benchmark

Other Benchmarks

3DMark Graphics score

High-end graphics benchmark

14857 points

707% significantly higher 3DMark Graphics score

1841 points

Blender score

Cycles Render (Samples per minute)

3754.04 points

1936% significantly higher Blender score

184.30 points

Geekbench 5 (CUDA) score

CUDA compute benchmark

165866 points

1285% significantly higher Geekbench 5 (CUDA) score

11969 points

Geekbench 5 (OpenCL) score

OpenCL compute benchmark

147200 points

895% significantly higher Geekbench 5 (OpenCL) score

14780 points

Geekbench 5 (Vulkan) score

Vulkan compute benchmark

104828 points

630% significantly higher Geekbench 5 (Vulkan) score

14358 points

Geekbench 6 (OpenCL) score

OpenCL compute benchmark

138062 points

801% significantly higher Geekbench 6 (OpenCL) score

15314 points

Geekbench 6 (Vulkan) score

Vulkan compute benchmark

131952 points

711% significantly higher Geekbench 6 (Vulkan) score

16269 points

Specifications

Core Clock Speed

Nominal chip frequency

1575 MHz

72% significantly higher core clock speed

915 MHz

Boost Clock Speed

Higher turbo frequency when power conditions are met

1770 MHz

80% significantly higher boost clock speed

980 MHz

Memory

Onboard memory size for textures and vertices

8 GiB

100% significantly more memory

4 GiB

Memory Bus Width

Number of parallel lines to the memory chips

256 Bit

Identical

256 Bit

Memory Bandwidth

Data transfer speed between GPU core and memory

608.3 GB/s

216% significantly higher memory bandwidth

192.3 GB/s

TDP

Thermal Design Power: Measure of heat generated by the GPU

290 W

170 W

70% significantly lower TDP

Pixel Rate

Number of pixels that can be rendered per second

169.9 Gigapixels/s

519% significantly higher pixel rate

27.44 Gigapixels/s

Texture Rate

Number of textured pixels that can be rendered per second

339.8 Gigatexels/s

209% significantly higher texture rate

109.8 Gigatexels/s

Floating Point Performance

Raw number of floating point operations per second

21.75 TFLOPS

725% significantly better floating point performance

2.63 TFLOPS

Shading Units

Number of processors dedicated to shader processing

6144

357% significantly more shading units

1344

Texture Mapping Units

Number of processors dedicated to applying textures

192

71% significantly more texture mapping units

112

Render Output Processors

Number of processors dedicated to final pixel rendering

200% significantly more render output processors

Other details

Rank

Ranking in the hardwareDB database

45th of 586

308th of 586

Release date

The official date of release of this chip

2021 May

Memory Type

The type of memory used by this chip

GDDR6X

GDDR5

DirectX Support

Maximum version of DirectX supported

12.0

11.0

OpenGL Support

Maximum version of OpenGL supported

4.6

4.5

GeForce RTX 3070 Ti vs GeForce GTX 670 benchmarks

For gaming, the GeForce RTX 3070 Ti graphics card is better than the GeForce GTX 670 in our tests.

The GeForce RTX 3070 Ti has a significantly higher core clock speed. This is the frequency at which the graphics core is running at. While not necessarily an indicator of overall performance, this metric can be useful when comparing two GPUs based on the same architecture. In addition, the GeForce RTX 3070 Ti also has a significantly higher boost clock speed: the maximum frequency the chip can reach if power delivery and thermals allow.

Next up, in terms of GPU memory, the GeForce RTX 3070 Ti has significantly more memory with 8 GiB of memory compared to 4 GiB. Memory size doesn't directly affect performance, but too little memory will certainly degrade gaming performance.

In addition, the GeForce GTX 670 has a significantly lower TDP at 170 W when compared to the GeForce RTX 3070 Ti at 290 W. TDP (Thermal Design Power) measures total heat output from the chip.

According to the results of the hardwareDB benchmark utility, the GeForce RTX 3070 Ti is faster than the GeForce GTX 670.