NVIDIA

GeForce RTX 5090

The NVIDIA GeForce RTX 5090 is a high-end consumer graphics card targeting gamers and content creators who demand top-tier performance. Built on the Ada Lovelace architecture, it offers significant improvements in ray tracing and AI-driven tasks. With enhanced CUDA cores and advanced RT and Tensor cores, it is designed for 4K gaming and complex rendering tasks.

GeForce RTX 5090 — illustration of the card's form factor
VRAM
32 GB
CUDA Cores
21,760
TDP
575 W
Memory
GDDR7

Provider Marketplace

Cheapest
$0.25/hour
Starting from
Best Value
$0.69/hour
Starting from
Enterprise Choice
$0.72/hour
Starting from

All Cloud Providers

4 Options available
SaladCloud logo
SaladCloudCheapest
On-Demand
$0.25/ hour
Estimated Cost
Provision
CloudRift logo
Reserved · commitment
$0.55/ hour
Estimated Cost
Provision
RunPod logo
On-Demand
$0.69/ hour
Estimated Cost
Provision
Runcrate logo
On-Demand
$0.72/ hour
Estimated Cost
View Provider

Estimates onlyrates are collected automatically from public provider pages and may be out of date. Prices vary by region, commitment term, and availability, and typically exclude storage, egress, and tax. Confirm current pricing with the provider before purchasing. Last collected .

Architecture

MicroarchitectureBlackwell
CUDA Cores21760
Tensor Cores5th Generation, 3352 AI TOPS
RT Cores4th Generation, 318 TFLOPS
Matrix EngineSecond-gen transformer model
Base Clock2.01 GHz
Boost Clock2.41 GHz
Transformer EngineYes (second-gen transformer model)
Dynamic PrecisionSupported (FP4)

Memory & VRAM

Memory TypeGDDR7
Total Capacity32 GB
Bandwidth1792 GB/sec
Bus Width512-bit

Connectivity & Scaling

PCIe InterfaceGen 5

Power & Efficiency

TDP575 W
PSU Required1000
Connectors4x PCIe 8-pin cables (adapter in box) OR 1x 600 W PCIe Gen 5 cable
Thermal LimitsMaximum GPU Temperature 90 C; Cooling: 3D Vapor Chamber with integrated heat pipes and active fins; dual-slot cooler (2-Slot)

Physical Design

Form FactorPCIe
Slot Width2-Slot
Dimensions304 mm x 137 mm
CoolingActive

Software Ecosystem

CUDA12.0
PyTorchYes
Driver StabilityYes

Server & Deployment

Ref ArchitecturesIn NVIDIA's Founders Edition or reference graphics card design.

System Compatibility

CPU PairingMinimum system power estimation is based on a PC configured with a Ryzen 9 9950X processor.
Required PCIePCI Express Gen 5
OS CompatWindows PC

Benchmarks & Throughput

Scaling Efficiency

NVIDIA NVLink (SLI-Ready): No

Multi-GPU Scalability

Scaling Characteristics

ParallelismCUDA cores; Tensor Cores (AI); PCI Express Gen 5; Resizable BAR; NVLink not supported; CUDA Capability 12.0

Workload Readiness

Market Authority

Key Strengths

The RTX 5090 excels in high-performance gaming and creative workloads.

  • ·4K Gaming: Delivers exceptional performance in 4K gaming with high frame rates.
  • ·Ray Tracing: Advanced RT cores provide realistic lighting and shadows.
  • ·AI Tasks: Enhanced Tensor cores accelerate AI-driven applications.
  • ·Content Creation: Optimized for video editing and 3D rendering tasks.

Limitations

High performance comes with increased power and space requirements.

  • ·Power Requirements: Demands a powerful PSU, increasing overall system cost.
  • ·Size: Large size may not fit in smaller cases.

Expert Insight

The GeForce RTX 5090 represents a strategic leap in AI compute. When comparing cloud providers, consider not just the hourly rate, but also the interconnect bandwidth (InfiniBand/NVLink) and regional availability which can significantly impact total cost of ownership for large-scale training.

Glossary Terms

FP32 TFLOPS
VRAM
TDP
Cores
Information updated daily. Cloud pricing subject to vendor availability.