Nvidia unveiled its latest flagship graphics card, the GeForce RTX 5090 at CES 2025 on January 6, 2025. This powerful GPU boasts impressive specs and aims to set new standards for gaming and creative performance. The RTX 5090 features 21,760 CUDA cores and 32GB of GDDR7 memory, making it the most powerful GeForce GPU ever created.
Part of Nvidia’s Blackwell lineup, the RTX 5090 will release on January 30, 2025 (along with the RTX 5080) and should be the best GPU you can buy for a PC for at least the next 2-3 years. Nvidia has priced this high-end graphics card at $1,999, so it’s definitely one of the more expensive options out there. But they claim that it is up to 2x faster than the previous flagship card (the RTX 4090) and are gearing it as a premium option for high-performance enthusiasts and professionals. The new GPU uses TSMC’s 4nm fabrication process, which promises improved efficiency and performance over previous generations.
Nvidia’s RTX 5090 launch is part of a wider release of the RTX 50 series. The lineup includes more affordable options like the RTX 5080 ($999), RTX 5070 Ti ($749), and RTX 5070 ($549). These cards offer varying levels of performance to suit different budgets and needs, giving consumers a range of choices in the latest GPU technology.
NVIDIA RTX 5090 Specs
| Feature | GeForce RTX 5090 |
|---|---|
| NVIDIA CUDA® Cores | 21760 |
| Shader Cores | Blackwell |
| Tensor Cores (AI) | 5th Gen, 3352 AI TOPS |
| Ray Tracing Cores | 4th Gen, 318 TFLOPS |
| Boost Clock (GHz) | 2.41 |
| Base Clock (GHz) | 2.01 |
| Standard Memory Config | 32 GB GDDR7 |
| Memory Interface Width | 512-bit |
| Technology Support | Blackwell |
| Ray Tracing | Yes |
| NVIDIA DLSS | DLSS 4, Super Resolution, DLAA, Ray Reconstruction, Frame Generation, Multi Frame Generation |
| NVIDIA Reflex | Reflex 2, Low Latency Mode, Frame Warp (Coming Soon) |
| NVIDIA Broadcast | Yes |
| PCI Express Gen 5 | Yes |
| Resizable BAR | Yes |
| NVIDIA App | Yes |
| NVIDIA Ansel | Yes |
| NVIDIA FreeStyle | Yes |
| NVIDIA ShadowPlay | Yes |
| NVIDIA Highlights | Yes |
| NVIDIA G-SYNC® | Yes |
| Game Ready Drivers | Yes |
| NVIDIA Studio Drivers | Yes |
| NVIDIA Omniverse | Yes |
| RTX Remix | Yes |
| Microsoft DirectX® 12 Ultimate | Yes |
| NVIDIA GPU Boost™ | Yes |
| NVIDIA NVLink™ (SLI-Ready) | No |
| Vulkan 1.4, OpenGL 4.6 | Yes |
| NVIDIA Encoder (NVENC) | 3x Ninth Gen |
| NVIDIA Decoder (NVDEC) | 2x Sixth Gen |
| AV1 Encode | Yes |
| AV1 Decode | Yes |
| CUDA Capability | 12.8 |
| VR Ready | Yes |
| Display Support | 4K at 480Hz or 8K at 120Hz with DSC |
| Standard Display Connectors | 3x DisplayPort(2), 1x HDMI(3) |
| Multi Monitor | up to 4 |
| HDCP | 2.3 |
| Card Dimensions | 304 mm |
| Length | 304 mm |
| Width | 137 mm |
| Slots | 2-Slot |
| SFF-Ready Enthusiast GeForce Card | Founders Edition Yes, Varies |
| Thermal and Power Specs | 90 |
| Maximum GPU Temperature (in C) | 90 |
| Total Graphics Power (W) | 575 |
| Required System Power (W) | 1000 |
| Supplementary Power Connectors | 4x PCIe 8-pin OR 1x 600W PCIe Gen 5 |
Pricing and Release Date
The NVIDIA RTX 5090 launched with an MSRP of $1999 USD. It was announced on January 6, 2025, and the official release date is January 30, 2025.
However, due to high demand and potential supply constraints, actual availability and pricing may vary. It’s recommended to check with retailers and online stores for the latest information on pricing and stock.
Key Takeaways
- The RTX 5090 offers 21,760 CUDA cores and 32GB GDDR7 memory for unparalleled performance
- Nvidia will release the RTX 5090 on January 30, 2025, with a price tag of $1,999
- The RTX 50 series includes more affordable options to cater to various user needs and budgets
NVIDIA RTX 5090 Overview
The NVIDIA GeForce RTX 5090 is the most powerful consumer GPU ever created. It boasts impressive specifications, incredible performance, and a premium price tag.
Specifications and Architecture
The RTX 5090 is built on NVIDIA’s new Blackwell architecture. It features over 20,000 CUDA cores, surpassing even the server-class NVIDIA H100 GPU. The base clock speed is 2.01 GHz, with a higher boost clock for demanding tasks.
The GPU uses GDDR7 memory on a 512-bit bus, providing massive bandwidth. It supports PCIe Gen 5 for lightning-fast data transfer. The RTX 5090 includes advanced Tensor Cores for AI operations and RT Cores for ray tracing.
Key specs:
- CUDA Cores: 20,000+
- Base Clock: 2.01 GHz
- Memory: GDDR7
- Memory Bus: 512-bit
- PCIe: Gen 5
Performance and Gaming
The RTX 5090 delivers unparalleled gaming performance. It easily handles 4K gaming at high frame rates and can even tackle 8K gaming in some titles. The GPU excels in ray tracing tasks, providing lifelike lighting and reflections.
NVIDIA’s DLSS 4 technology further boosts performance through AI upscaling. This allows gamers to enjoy higher resolutions and frame rates without sacrificing visual quality.
In games like Cyberpunk 2077, the RTX 5090 provides smooth gameplay at maximum settings. Its AI capabilities also enhance NPC behavior and environmental details.
Pricing, Availability, and Market Position
The GeForce RTX 5090 comes with a premium price tag of $1,999. This positions it as NVIDIA’s flagship consumer GPU, aimed at enthusiasts and professionals who demand the best performance.
The GPU became available in January 2025, launching alongside other RTX 50-series cards like the 5080, 5070 Ti, and 5070. These offer various price points and performance levels to suit different budgets and needs.
The RTX 5090 faces competition from AMD’s high-end offerings. However, its superior ray tracing and AI capabilities give it an edge in the premium GPU market.
Availability may be limited at launch due to high demand. The GPU is compatible with high-end desktop PCs and some premium laptops.
