NVIDIA GeForce RTX 4090 vs NVIDIA GeForce RTX 4070 Ti

Comparative analysis of NVIDIA GeForce RTX 4090 and NVIDIA GeForce RTX 4070 Ti videocards for all known characteristics in the following categories: Essentials, Technical info, Video outputs and ports, Compatibility, dimensions and requirements, API support, Memory. Benchmark videocards performance analysis: PassMark - G2D Mark, PassMark - G3D Mark, 3DMark Fire Strike - Graphics Score, CompuBench 1.5 Desktop - Face Detection (mPixels/s), CompuBench 1.5 Desktop - T-Rex (Frames/s), CompuBench 1.5 Desktop - Video Composition (Frames/s), CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s), Geekbench - OpenCL, CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s).

NVIDIA GeForce RTX 4090

NVIDIA GeForce RTX 4070 Ti

Differences

Reasons to consider the NVIDIA GeForce RTX 4090

2.1x more pipelines: 16384 vs 7680
2x more maximum memory size: 24 GB vs 12 GB
Around 8% better performance in PassMark - G2D Mark: 1294 vs 1195
Around 21% better performance in PassMark - G3D Mark: 38529 vs 31827
Around 60% better performance in 3DMark Fire Strike - Graphics Score: 36466 vs 22839
Around 38% better performance in CompuBench 1.5 Desktop - T-Rex (Frames/s): 93.23 vs 67.426
Around 52% better performance in CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s): 4413.025 vs 2895.895
Around 55% better performance in Geekbench - OpenCL: 317521 vs 205236

Specifications (specs)
Pipelines	16384 vs 7680
Maximum memory size	24 GB vs 12 GB
Benchmarks
PassMark - G2D Mark	1294 vs 1195
PassMark - G3D Mark	38529 vs 31827
3DMark Fire Strike - Graphics Score	36466 vs 22839
CompuBench 1.5 Desktop - T-Rex (Frames/s)	93.23 vs 67.426
CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s)	4413.025 vs 2895.895
Geekbench - OpenCL	317521 vs 205236

Reasons to consider the NVIDIA GeForce RTX 4070 Ti

Videocard is newer: launch date 3 month(s) later
Around 3% higher core clock speed: 2310 MHz vs 2235 MHz
Around 4% higher boost clock speed: 2610 MHz vs 2520 MHz
485.6x more texture fill rate: 626.4 GTexel/s vs 1,290 GTexel/s
Around 58% lower typical power consumption: 285 Watt vs 450 Watt
2x better performance in CompuBench 1.5 Desktop - Face Detection (mPixels/s): 939.363 vs 461.456
Around 73% better performance in CompuBench 1.5 Desktop - Video Composition (Frames/s): 348.259 vs 200.733

Specifications (specs)
Launch date	3 Jan 2023 vs 20 Sep 2022
Core clock speed	2310 MHz vs 2235 MHz
Boost clock speed	2610 MHz vs 2520 MHz
Texture fill rate	626.4 GTexel/s vs 1,290 GTexel/s
Thermal Design Power (TDP)	285 Watt vs 450 Watt
Benchmarks
CompuBench 1.5 Desktop - Face Detection (mPixels/s)	939.363 vs 461.456
CompuBench 1.5 Desktop - Video Composition (Frames/s)	348.259 vs 200.733

Compare benchmarks

GPU 1: NVIDIA GeForce RTX 4090
GPU 2: NVIDIA GeForce RTX 4070 Ti

PassMark - G2D Mark

GPU 1

GPU 2

1294

1195

PassMark - G3D Mark

GPU 1

GPU 2

38529

31827

3DMark Fire Strike - Graphics Score

GPU 1

GPU 2

36466

22839

CompuBench 1.5 Desktop - Face Detection (mPixels/s)

GPU 1

GPU 2

461.456

939.363

CompuBench 1.5 Desktop - T-Rex (Frames/s)

GPU 1

GPU 2

93.23

67.426

CompuBench 1.5 Desktop - Video Composition (Frames/s)

GPU 1

GPU 2

200.733

348.259

CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s)

GPU 1

GPU 2

4413.025

2895.895

Geekbench - OpenCL

GPU 1

GPU 2

317521

205236

Name	NVIDIA GeForce RTX 4090	NVIDIA GeForce RTX 4070 Ti
PassMark - G2D Mark	1294	1195
PassMark - G3D Mark	38529	31827
3DMark Fire Strike - Graphics Score	36466	22839
CompuBench 1.5 Desktop - Face Detection (mPixels/s)	461.456	939.363
CompuBench 1.5 Desktop - T-Rex (Frames/s)	93.23	67.426
CompuBench 1.5 Desktop - Video Composition (Frames/s)	200.733	348.259
CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s)	4413.025	2895.895
Geekbench - OpenCL	317521	205236
CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s)	0

Compare specifications (specs)

	NVIDIA GeForce RTX 4090	NVIDIA GeForce RTX 4070 Ti
Essentials
Architecture	Ada Lovelace	Ada Lovelace
Code name	AD102	AD104
Launch date	20 Sep 2022	3 Jan 2023
Place in performance rating	10	11
Technical info
Boost clock speed	2520 MHz	2610 MHz
Core clock speed	2235 MHz	2310 MHz
Manufacturing process technology	4 nm	4 nm
Peak Double Precision (FP64) Performance	1,290 GFLOPS (1:64)	626.4 GFLOPS (1:64)
Peak Half Precision (FP16) Performance	82.58 TFLOPS (1:1)	40.09 TFLOPS (1:1)
Peak Single Precision (FP32) Performance	82.58 TFLOPS	40.09 TFLOPS
Pipelines	16384	7680
Pixel fill rate	443.5 GPixel/s	208.8 GPixel/s
Texture fill rate	1,290 GTexel/s	626.4 GTexel/s
Thermal Design Power (TDP)	450 Watt	285 Watt
Transistor count	76300 million	35800 million
Video outputs and ports
Display Connectors	1x HDMI 2.1, 3x DisplayPort 1.4a	1x HDMI 2.1, 3x DisplayPort 1.4a
Compatibility, dimensions and requirements
Form factor	Triple-slot	Dual-slot
Height	61 mm, 2.4 inches	42 mm, 1.7 inches
Interface	PCIe 4.0 x16	PCIe 4.0 x16
Length	304 mm, 12 inches	285 mm, 11.2 inches
Recommended system power (PSU)	850 Watt	600 Watt
Supplementary power connectors	1x 16-pin	1x 16-pin
Width	137 mm, 5.4 inches	112 mm, 4.4 inches
API support
DirectX	12 Ultimate (12_2)	12 Ultimate (12_2)
OpenCL	3.0	3.0
OpenGL	4.6	4.6
Shader Model	6.7	6.7
Vulkan
Memory
Maximum RAM amount	24 GB	12 GB
Memory bandwidth	1,008 GB/s	504.2 GB/s
Memory bus width	384 bit	192 bit
Memory clock speed	1313 MHz, 21 Gbps effective	1313 MHz, 21 Gbps effective
Memory type	GDDR6X	GDDR6X