NVIDIA GeForce RTX 4090
3 laptops with NVIDIA GeForce RTX 4090 in comparion
Outstanding equipment
The Nvidia GeForce RTX 4090 was released on 03.01.2023 and is the fastest mobile GPU on the market at this time. With 16 GB GDDR6 and a 256-bit graphics bus operating at 20 GB/s, the RTX 4090 offers performance of the highest standard. With the new 4 nm manufacturing process, the computing core density is higher than ever before. Like all GPUs in the RTX series, it offers support for ray tracing and G-Sync.
According to NVIDIA, there are different variants of the Nvidia GeForce RTX 4090, which are scaled between 80 - 150 watts of power consumption. The performance that can be achieved depends on this TGP and the cooling system used. The Nvidia GeForce RTX 4090 also supports the new version of the DirectX API: DirectX Ultimate, which makes ray tracing possible.
New architecture
The GPU uses an AD103 Ada-Lovelace chip, which, according to NVIDIA, enables improved performance in ray tracing and should also offer an additional performance boost in shader calculations. NVIDIA promises a doubling of the bandwidth, the ray tracing cores and the tensor cores. Compared to the desktop version of the RTX 4090, the mobile version is logically much slower. The Nvidia GeForce RTX 4090 should be comparable to the Nvidia GeForce RTX 4080 desktop version, but benchmarks to verify this do not yet exist.
Too new for benchmarks
Due to the newness of the graphics card, only a few statements can be made about its performance. However, it is certainly the most powerful graphics card currently available from NVIDIA. However, it is still too early to determine its exact performance. Nevertheless, the graphics card is already being installed in gaming laptops. It is combined with the best CPUs of the new generation from Intel, such as the Intel Core i9-13950HX, which is also new on the market. Razer in particular is already installing the new graphics card.
(Status: 3.2.2023)
Core Information
Codename
GN21-X11
GPU Name
AD103
Model Series
NVIDIA GeForce RTX 40
Release Date
03.01.2023
GPU Specifications
GPU Architecture
Ada Lovelace
Base Clock
1335 MHz
Boost Clock
2040 MHz
Integrated GPU
No
MXM Module
No
Compute Units
Render Output Units
112
Texture Mapping Units
304
Shader Units
9728
Tensor Cores
304
Hardware Ray Tracing
Yes
Ray Tracing Units
76
Memory
Memory Type
GDDR6
Memory Clock
2250 MHz
Effective Memory Data Rate
20 Gbps
Memory Bandwidth
576 GB/s
System-Dependent Memory Bandwidth
No
Memory Size
16 GB
Memory Interface
256 Bit
Shared Memory
No
Dedicated Graphics Memory
Yes
Unified Memory
No
Cache
L2 Cache
64 MB
Manufacturing
Foundry
TSMC
Process Node
5 nm
Chip Specifications
Chip Integration
Monolithic
Power Consumption
Base Power Consumption
150 Watt
Base Power Type
TGP
Minimum Power Consumption
80 Watt
Minimum Power Type
TGP
Maximum Power Consumption
150 Watt
Maximum Power Type
TGP
Supported Standards
DirectX
12
OpenGL
4.6
OpenCL
3.0
Vulkan
1.3
Shader Model
6.7
Intel Iris Xe Graphics G7 80 EUs
AMD Radeon 610M
NVIDIA GeForce RTX 5070 Ti
Intel Graphics 4 Xe3 2.45 GHz (Panther Lake)
NVIDIA RTX PRO 500 Blackwell
Intel Arc B390
Intel Graphics 4 Xe3 2.5 GHz (Panther Lake)
Intel Graphics 2 Xe3 (Wildcat Lake)
Qualcomm Adreno X2-90 1.7GHz
Qualcomm Adreno X2-85
NVIDIA GeForce RTX 5060
NVIDIA GeForce RTX 5070
Start Laptop Configurator
Start Interactive Buying Guide