NVIDIA Elite Partner

NVIDIA GEFORCE RTX 4070 GPU - System Overview

Performance

Nvidia GeForce RTX 4070 GPU

The GeForce RTX 4070 is equipped with upgraded 3rd generation RT Cores and 4th generation Tensor Cores, as well as lightning fast GDDR6X memory. This GPU and its direct predecessor, the RTX 3070, share a lot of characteristics. Both support 46 raytracing (RT) cores, 184 Tensor cores, 5888 CUDA cores, and 184 texture mapping units (TMUs). They are different because the RTX 4070 has 64 raster operation units (ROPs) and a 192-bit interface. With an HDMI port and 3x DisplayPort connections, it can accommodate up to 4x independent monitors at 120Hz. Although at 240Hz, it can only support 2x 4K monitors or 8K at 60Hz with DSC using DP or HDMI.

Memory

With 12GB of GDDR6X RAM, the RTX 4070 has a 192-bit memory interface running at speeds of 1313 MHz (effectively 21 Gbps). The GPU operates at 1920 MHz with a boost speed of 2475 MHz. This graphics card does not support NVIDIA’s NVLink.

Cooling & Features

Nvidia GeForce RTX 4070 GPU fan

The dual-slot NVIDIA RTX 4070 GPU resembles the larger RTX 4080 and RTX 4090 cards, with the exception of its smaller size. On the exterior of the card there are three DisplayPort connections and one HDMI connection. Two fans that are placed at the card's opposing ends with one exposed on each side of the board. This cooling design enables a flow-through system drawing air from the bottom, through the card, and out the case's top. It draws power from 1x 16-pin power connector, with power draw rated at 200 W maximum, or 2x 8-pin connectors. Minimum required system power is listed at 650W.

Summary

The NVIDIA GeForce RTX 4070 GPU raises the bar for both gaming and artistic endeavors. The Ada Lovelace-based graphics card enables 12GB of GDDR6X memory in addition to having 5888 CUDA cores for more realalistic rendering. This VR ready GPU offers users the performance they require to produce their finest work, utilizing 46 RT cores, 184 Tensor cores, 5888 CUDA cores, and 184 TMUs.

NVIDIA GEFORCE RTX 4070 GPU - Specifications

Memory

    GPU Memory: 12GB GDDR6X
  • Memory Bus: 192-bit
  • Memory Bandwidth: Up to 504.2 GB/s

GPU Engine

  • NVIDIA CUDA Cores: 5888
  • Boost Clock: 1920 GHz
  • Base Clock: 2475 GHz

Performance

  • FP16 (half): 29.15 TFLOPS
  • FP32 (float): 29.15 TFLOPS
  • Transistor Count: 35,800 million

NVIDIA NVLink

  • Not Supported

System Interface

  • PCI Express 4.0 x16

Power Consumption

    Maximum GPU Temperature (in C)

  • 90
  • Graphics Card Power (W)

  • 200
  • Required System Power (W)

  • 650
  • Supplementary Power Connectors

  • 2x PCIe 8-pin cables (adapter in box) OR 300 W or greater PCIe Gen 5 cable
  • Certain manufacturer models may use 1x PCIe 8-pin cable

Form Factor

  • 9.6" (244 mm) x 4.4" (112 mm)
  • 2-slot

Display Support

    Display Connector

  • 3x DisplayPort
  • HDMI
  • Maximum Digital Resolution

  • 4K at 240Hz or 8K at 60Hz with DSC
  • Multi Monitor

  • Up to 4
  • HDCP

  • 2.3

Processors

  • Architecture: Ada Lovelace
  • Fabrication Process: 5nm
  • TMUs: 184
  • ROPs: 64
  • SM Count: 46
  • Tensor Cores: 184
  • RT Cores: 46

Engine

    VR Ready

  • Yes
  • Graphics Features

  • Microsoft DirectX 12 Ultimate
  • Vulkan 1.3
  • OpenGL 4.6
  • OpenCL 3.0
  • CUDA 8.9
  • Shader Model 6.7
  • Supported Technologies

  • Ray Tracing
  • NVIDIA DLSS
  • NVIDIA Ansel
  • NVIDIA Highlights
  • NVIDIA G-SYNC
  • NVIDIA Omniverse
  • NVIDIA ShadowPlay
  • NVIDIA Reflex

NVIDIA GEFORCE RTX 4070 GPU - Documentation

Let's get in touch! Get a consultation or quote on exact configuration solution you're looking for
Contact us