Select processor 1
Select processor 2
NVIDIA GeForce RTX 4090
NVIDIA GeForce RTX 3090 Ti

NVIDIA GeForce RTX 4090 vs NVIDIA GeForce RTX 3090 Ti. Specifications, performance, tests

User rating
star star star star star
Released
Q4/2022
NVIDIA GeForce RTX 4090
NVIDIA GeForce RTX 4090
Released
Q1/2022
User rating
star star star star star
NVIDIA GeForce RTX 3090 Ti
NVIDIA GeForce RTX 3090 Ti

What's the best choice NVIDIA GeForce RTX 4090 or NVIDIA GeForce RTX 3090 Ti? Which graphics card is faster?

Compare the available specifications of these models.

NVIDIA GeForce RTX 4090 has a base frequency of 2.23 GHz. Memory size 24 GB. Memory type GDDR6X. Released in Q4/2022.

NVIDIA GeForce RTX 3090 Ti has a base frequency of 1.56 GHz. Memory size 24 GB. Memory type GDDR6X. Released in Q1/2022.

Differences

  • Higher clock speed

    Around 43% better clock speed

    2.23 GHz left arrow 1.56 GHz

Positions in all rankings

Common positions NVIDIA GeForce RTX 4090 GPU in popular benchmarks, for comparison with other models.

  • FP32 Performance (Single-precision TFLOPS)
    12 place
  • Battlefield 5
    5 place
  • Cyberpunk 2077
    11 place

Positions in all rankings

Common positions NVIDIA GeForce RTX 3090 Ti GPU in popular benchmarks, for comparison with other models.

  • Passmark Performance Test GPU
    1 place

Specifications

Technical data
NVIDIA GeForce RTX 4090 NVIDIA GeForce RTX 4090
NVIDIA GeForce RTX 3090 Ti NVIDIA GeForce RTX 3090 Ti
General info

Comparison of basic technical data of graphics cards NVIDIA GeForce RTX 4090 and NVIDIA GeForce RTX 3090 Ti, chip, information processing units.

  • Architecture
    Ada Lovelace left arrow Ampere
  • Based on
    NVIDIA GeForce RTX 4090 left arrow GA102
  • CUDA Cores
    16384 left arrow 10752
  • GPU Chip
    AD102-300 / AD102-301 (Ada Lovelace) left arrow GA102-350-A1
  • Render Output Units
    192 left arrow 112
  • RT Cores
    128 left arrow -
  • Shader
    16384 left arrow 10752
  • Streaming Multiprocessors
    128 left arrow -
  • Texture Units
    512 left arrow 336
  • L1 Cache
    - left arrow 128 KB
  • L2 Cache
    - left arrow 6 MB
Memory settings

Video memory capacity, type, bus width and bandwidth describe different properties. These values alone do not determine graphics performance.

  • Memory Bandwidth
    1008 GB/s left arrow 1,008 GB/s
  • Memory Bus
    384 bit left arrow 384 bit
  • Memory Clock
    1313 MHz left arrow -
  • Memory Data Rate
    21 Gbps left arrow -
  • Memory Size
    24 GB left arrow 24 GB
  • Memory Type
    GDDR6X left arrow GDDR6X
  • Memory Speed (legacy frequency label)
    - left arrow 1300 MHz
Clock frequency

Published GPU clock frequencies describe different operating modes. Clock speed alone does not determine graphics performance.

  • Base Clock
    2230 MHz left arrow 1560 MHz
  • Boost Clock
    2520 MHz left arrow 1860 MHz
  • Game Clock
    - left arrow -
  • Overclocking
    Yes left arrow No
Thermal Management

Graphics card power limits and recommended system power supply describe different specifications.

  • Maximum GPU Temperature
    90 °C left arrow 92 °C
  • Maximum Junction Temperature
    90 °C left arrow -
  • System Power Supply
    850 W (System power requirement) left arrow 850 W (System power requirement)
  • TDP
    450 W left arrow 450 W
  • TDP (up)
    450 W-25 % left arrow -
  • Total Graphics Power
    450 W left arrow -
  • Graphics Card Power
    - left arrow 450 W
Cooler & Fans type

Types and sizes of coolers for graphics card cooling system NVIDIA GeForce RTX 4090 and NVIDIA GeForce RTX 3090 Ti.

  • Cooler-Type
    Air cooling left arrow -
  • Fan-Type
    Axial left arrow -
  • Fan 1
    2 x 110 mm left arrow -
  • Fan 2
    - left arrow -
  • Noise (Idle)
    0 dB / Silent left arrow -
  • Noise (Load)
    39 dB left arrow -
Interfaces

Connectivity and connections.

  • DisplayPort Ports
    3 × DisplayPort left arrow 3 × DisplayPort 1.4a
  • DVI Ports
    - left arrow -
  • HDCP Version
    2.3 left arrow 2.3
  • HDMI Ports
    1 × HDMI 2.1 left arrow 1 × HDMI 2.1
  • Max. Displays
    4 left arrow 4
  • Other Display Outputs (unparsed)
    HDMI left arrow HDMI
  • PCIe Interface
    PCIe 4.0 x16 left arrow PCIe 4.0 x16
  • Power Connectors
    3x PCIe 8-pin cables (adapter in box) OR 1x 450 W or greater PCIe Gen 5 cable left arrow 3x PCIe 8-pin cables (adapter in box) OR 450W or greater PCIe Gen 5 cable
  • USB-C Ports
    - left arrow -
  • VGA Ports
    - left arrow -
Technical data

Technical data that is used to its full potential in computer games.

  • DirectX
    12_2 (Ultimate) left arrow 12_2
  • DLSS / FSR
    Yes left arrow -
  • LED
    No LED lighting left arrow -
  • Max. resolution
    7680 × 4320 left arrow 7680 × 4320
  • Raytracing
    Yes left arrow -
  • OpenCL
    - left arrow 3.0
  • OpenGL
    - left arrow 4.6
  • Shader Model
    - left arrow 6.6
  • Vulkan
    - left arrow 1.3
Supported Video Codecs

Built-in support for video and image compression standards.

  • AV1
    Decode left arrow -
  • h264
    Decode / Encode left arrow -
  • h265 / HEVC
    Decode / Encode left arrow -
  • VP8
    Decode left arrow -
  • VP9
    Decode left arrow -
Dimensions

The difference in size, weight and slot of the compared devices

  • Height
    137 mm left arrow -
  • Length
    304 mm left arrow -
  • Slot Size
    3 PCIe-Slots left arrow -
  • Weight
    - left arrow -
  • Width
    61 mm left arrow -
More information

Comparison of interfaces and release dates for GPU NVIDIA GeForce RTX 3090 Ti and NVIDIA GeForce RTX 4090.

  • Documents
    data sheet left arrow -
  • Launch Price (historical)
    1,599 $ (Reference) left arrow -
  • Manufacturer specification scope
    NVIDIA Founders Edition/reference design. Clock specifications apply to gaming use. AIB shipping specifications and complete-system power needs may differ. left arrow NVIDIA Founders Edition/reference design. Clock specifications apply to gaming use. AIB shipping specifications and complete-system power needs may differ.
  • Part Number
    - left arrow -
  • Process Technology
    4 nm left arrow 8 nm
  • Release date
    Q4/2022 left arrow -
Theoretical Performance

  • FP16 (theoretical)
    - left arrow 40 TFLOPS
  • FP32 (theoretical)
    - left arrow 40 TFLOPS
  • FP64 (theoretical)
    - left arrow 0.625 TFLOPS
  • Pixel Rate
    - left arrow 208.3 GPixel/s
  • Texture Rate
    - left arrow 625.0 GTexel/s

Benchmarks

Performance tests GPUs

Shared tests with results for both models: 15

Blender Cycles

Each row uses one Blender version, operating system and compute backend.

Test and settingsNVIDIA GeForce RTX 4090 NVIDIA GeForce RTX 3090 TiResult difference¹
Blender 5.2.0 / OPTIX / Windows 11483.608998 samples/min Submissions: 82 5257.100902 samples/min Submissions: 13 -54.2 %
Blender 5.1.1 / OPTIX / Linux 12173.930948 samples/min Submissions: 29 5601.471369 samples/min Submissions: 7 -54.0 %
Blender 5.1.1 / OPTIX / Windows 11480.172372 samples/min Submissions: 103 5459.157697 samples/min Submissions: 5 -52.4 %
Blender 4.5.0 / OPTIX / Linux 11469.759697 samples/min Submissions: 74 5309.43807 samples/min Submissions: 24 -53.7 %
Blender 4.5.0 / OPTIX / Windows 11001.477887 samples/min Submissions: 453 5585.151795 samples/min Submissions: 115 -49.2 %
Blender 4.4.0 / OPTIX / Linux 11556.614253 samples/min Submissions: 74 4979.851473 samples/min Submissions: 8 -56.9 %
Blender 4.4.0 / OPTIX / Windows 11030.268697 samples/min Submissions: 330 5700.575741 samples/min Submissions: 51 -48.3 %
Blender 4.3.0 / OPTIX / Linux 11398.037287 samples/min Submissions: 78 5840.23445 samples/min Submissions: 8 -48.8 %
Blender 4.3.0 / OPTIX / Windows 10893.688375 samples/min Submissions: 650 5490.159484 samples/min Submissions: 35 -49.6 %
Blender 4.2.0 / OPTIX / Linux 11460.995047 samples/min Submissions: 84 5972.585335 samples/min Submissions: 15 -47.9 %
Blender 4.2.0 / OPTIX / Windows 10838.48053 samples/min Submissions: 713 5673.136912 samples/min Submissions: 51 -47.7 %
Blender 4.1.0 / OPTIX / Linux 11758.839487 samples/min Submissions: 48 6034.022292 samples/min Submissions: 7 -48.7 %
More tests: 3
Test and settingsNVIDIA GeForce RTX 4090 NVIDIA GeForce RTX 3090 TiResult difference¹
Blender 4.1.0 / OPTIX / Windows 11253.451304 samples/min Submissions: 431 5769.869986 samples/min Submissions: 32 -48.7 %
Blender 4.0.0 / OPTIX / Linux 11774.185185 samples/min Submissions: 133 6168.026652 samples/min Submissions: 16 -47.6 %
Blender 4.0.0 / OPTIX / Windows 11246.939393 samples/min Submissions: 994 5124.064923 samples/min Submissions: 111 -54.4 %

¹ Difference of the second result relative to the first in that row. No percentage is calculated for different upscaling algorithms. This compares published results, not a universal performance index.

Individual model tests

These results have no shared test of both models with the same version and settings. They are shown separately without a percentage difference.

NVIDIA GeForce RTX 4090

More results: 237

NVIDIA GeForce RTX 3090 Ti

Individual system tests

Results were obtained on the reported systems. Test versions and settings remain separate; incomplete conditions do not support calculated model performance differences.

Cinebench 2024 / GPU

Model ResultReported conditions
NVIDIA GeForce RTX 4090 31793 points System: GeForce RTX 4090
Graphics: GeForce RTX 4090
Source minimum runtime: 0 minutes.
Exact test build and complete conditions are unreported; no ranking position.

Latest comparisons