NVIDIA · October 2022

GeForce RTX 4090

The NVIDIA GeForce RTX 4090 Founders Edition is a high-end consumer GPU designed for enthusiasts and professionals. It is part of the Ada Lovelace architecture, offering significant performance improvements over previous generations. Targeted at gamers and content creators, it features advanced ray tracing and AI capabilities, making it ideal for demanding applications and next-gen gaming experiences.

GeForce RTX 4090 — illustration of the card's form factor
VRAM
24 GB
FP32 TFLOPS
82.6 TFLOPS
CUDA Cores
16,384
TDP
450 W

Provider Marketplace

Cheapest
$0.16/hour
Starting from
Best Value
$0.34/hour
Starting from
Enterprise Choice
$0.66/hour
Starting from

All Cloud Providers

7 Options available
SaladCloud logo
SaladCloudCheapest
On-Demand
$0.16/ hour
Estimated Cost
Provision
$0.32/ hour
Estimated Cost
Provision
CloudRift logo
Reserved · commitment
$0.33/ hour
Estimated Cost
Provision
RunPod logo
On-Demand
$0.34/ hour
Estimated Cost
Provision
TensorDock logo
On-Demand
$0.35/ hour
Estimated Cost
Provision
Oblivus logo
Reserved · commitment
$0.38/ hour
Estimated Cost
Provision
Runcrate logo
On-Demand
$0.66/ 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Estimates onlyrates are collected automatically from public provider pages and may be out of date.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 .

Compute Performance

FP3282.6 TFLOPS
FP8660.6/1321.2 TFLOPS

Architecture

MicroarchitectureAda Lovelace
CUDA Cores16384
Tensor Cores4th Generation
RT Cores3rd Generation
Base Clock2230 MHz
Boost Clock2520 MHz

Memory & VRAM

Memory TypeGDDR6X
Total Capacity24 GB
Bandwidth1008 GB/sec
Bus Width384-bit
Memory PoolingNot Supported

Connectivity & Scaling

PCIe InterfaceGen 4

Power & Efficiency

TDP450 W
Peak Power450
Idle Power19
PSU Required850
Connectors3x PCIe 8-pin cables (adapter in box) OR 450 W or greater PCIe Gen 5 cable
Thermal LimitsMaximum GPU Temperature: 90 C

Physical Design

Form FactorPCIe
Slot Width3-Slot
Dimensions304 mm x 137 mm x 61 mm

Thermals & Cooling

DC Heat450 W

Software Ecosystem

Driver StabilityGame Ready and Studio Drivers; NVIDIA Studio drivers engineered for maximum stability

Server & Deployment

Ref ArchitecturesThis GPU is represented in NVIDIA's Founders Edition or reference graphics card design.

System Compatibility

Required PCIePCI Express Gen 4
OS CompatWindows 11 x64

Benchmarks & Throughput

Structured Sparsity

Supported (FP8 performance with structured Sparsity enabled)

Transformer Throughput

Delivering over 1.3 PetaFLOPS of tensor processing in the RTX 4090

Scaling Efficiency

No (NVIDIA NVLink/SLI not supported)

Multi-GPU Scalability

Scaling Efficiency

Single GPUover 2x more performance (at same power as RTX 3090 Ti)

Scaling Characteristics

ParallelismShader Execution Reordering (SER); No NVLink (SLI-Ready)

Workload Readiness

Market Authority

Key Strengths

Excels in high-resolution gaming and content creation tasks.

  • ·4K Gaming: Delivers exceptional performance in 4K gaming with ray tracing.
  • ·Content Creation: Accelerates rendering and video editing tasks.
  • ·AI Workloads: Supports AI development with Tensor Cores.

Limitations

High cost and power requirements are key considerations.

  • ·Cost: Premium pricing limits accessibility to enthusiasts.
  • ·Power Demand: High power draw necessitates a robust power supply.
  • ·Size: Large physical size may not fit in all cases.

Expert Insight

The GeForce RTX 4090 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.