NVIDIA GeForce RTX 4090 Max-Q

NVIDIA GeForce RTX 4090 Max-Q — 16 Гб GDDR6, 9,728 ядер, GPI 14.4

Loading...

Performance Rating

The GeForce RTX 4090 Max-Q is a mobile graphics chip by NVIDIA, launched on January 3rd, 2023. Built on the 5 nm process, and based on the AD103 graphics processor, in its GN21-X11 variant, the chip supports DirectX 12 Ultimate. This ensures that all modern games will run on GeForce RTX 4090 Max-Q. Additionally, the DirectX 12 Ultimate capability guarantees support for hardware-raytracing, variable-rate shading and more, in upcoming video games. The AD103 graphics processor is a large chip with a die area of 379 mm² and 45,900 million transistors. Unlike the fully unlocked GeForce RTX 4080 SUPER, which uses the same GPU but has all 10240 shaders enabled, NVIDIA has disabled some shading units on the GeForce RTX 4090 Max-Q to reach the product's target shader count. It features 9728 shading units, 304 texture mapping units, and 112 ROPs. Also included are 304 tensor cores which help improve the speed of machine learning applications. The card also has 76 raytracing acceleration cores. NVIDIA has paired 16 GB GDDR6 memory with the GeForce RTX 4090 Max-Q, which are connected using a 256-bit memory interface. The GPU is operating at a frequency of 930 MHz, which can be boosted up to 1455 MHz, memory is running at 2250 MHz (18 Gbps effective).

A100 A100
H200 H200
MI325X MI325X

NVIDIA GeForce RTX 4090 Max-Q

14.4

NVIDIA GeForce RTX 4090 Max-Q

14.4

Memory

Memory Size

16 GB

Memory Type

Memory Bandwidth

576.0 GB/s

Memory Bus Width

256 бит

ML Performance

FP16 (Half Precision)

28.31 TFLOPS

BF16 (Brain Float)

No TFLOPS
311.84 TFLOPS
(NVIDIA A800 SXM4 80 GB)

TF32 (TensorFloat)

Compute Power

FP32 (Single Precision)

28.31 TFLOPS

FP64 (Double Precision)

0.4423 TFLOPS

CUDA Cores

Architecture & Compatibility

GPU Architecture

Ada Lovelace

SM (Streaming Multiprocessor)

PCIe Version

PCIe 4.0 x16

ML Software Support

CUDA Version

CUDA Toolkit (first supported)

CUDA Toolkit 11

CUDA Toolkit status

Supported

Clocks & Performance

Base Clock

Boost Clock

Memory Clock

Power Consumption

Recommended PSU

Power Connector

None

TDP/TGP

80 W
unknown
(NVIDIA CMP 70HX)

Rendering

Texture Units (TMU)

L2 Cache

64 MB

Additional

Slots

IGP
SXM Module
(NVIDIA B300)

Release Date

Jan. 3, 2023
March 12, 2026
(Intel Arc B770)

Display Outputs

Portable Device Dependent
Portable Device Dependent
(NVIDIA GeForce RTX 5090 Mobile)

Renting is cheaper than buying