LLMpediaThe first transparent, open encyclopedia generated by LLMs

GeForce RTX 40 series

Generated by DeepSeek V3.2
⚠Note: This article was automatically generated by a large language model (LLM) from purely parametric knowledge (no retrieval). It may contain inaccuracies or hallucinations. This encyclopedia is part of a research project currently under review.
Article Genealogy
Parent: Nvidia Hop 3
Expansion Funnel Raw 66 → Dedup 27 → NER 15 → Enqueued 12
1. Extracted66
2. After dedup27 (None)
3. After NER15 (None)
Rejected: 12 (not NE: 12)
4. Enqueued12 (None)
Similarity rejected: 3
GeForce RTX 40 series
NameGeForce RTX 40 series
ManufacturerNVIDIA
CodenameAda Lovelace
FabTSMC
Generation4th Generation RTX
ReleasedSeptember 20, 2022
PredecessorGeForce 30 series
SuccessorGeForce 50 series
TypeGraphics processing unit
MemoryGDDR6X
InterfacePCI Express

GeForce RTX 40 series. The GeForce RTX 40 series is a line of graphics processing units developed by NVIDIA, succeeding the GeForce 30 series. Based on the Ada Lovelace architecture and manufactured by TSMC using a custom 4N process, these GPUs introduced significant generational leaps in performance and new features like DLSS 3. The launch lineup included models such as the GeForce RTX 4090 and GeForce RTX 4080, targeting enthusiasts and high-end PC gaming.

Overview

The series was officially announced by NVIDIA CEO Jensen Huang during a keynote at GTC in September 2022, with the flagship GeForce RTX 4090 launching first. This generation marked a substantial architectural shift, leveraging the advanced Ada Lovelace design to deliver major improvements in ray tracing and AI-driven performance. The release was strategically positioned against competing products from AMD's Radeon RX 7000 series and aimed to push forward technologies like the Metaverse and real-time path tracing in titles such as Cyberpunk 2077.

Specifications

The GPUs utilize TSMC's 4N manufacturing process, a custom variant of its 5 nm process, allowing for increased transistor density and power efficiency. Memory configurations predominantly feature high-speed Micron GDDR6X VRAM, with the GeForce RTX 4090 equipped with 24 GB. Key specifications include a significant increase in CUDA core counts, larger L2 cache sizes compared to the Ampere generation, and support for the latest PCI Express 4.0 interface. The Founders Edition cards maintained a distinct cooling design pioneered with the GeForce 30 series.

Architecture

The underlying Ada Lovelace architecture introduces several groundbreaking technologies. The new fourth-generation Tensor Cores are optimized for the FP8 data format, accelerating the optical flow accelerator that enables DLSS 3's Frame Generation. Third-generation RT Cores offer up to twice the ray-triangle intersection throughput, enhancing performance in ray tracing workloads for games like Portal with RTX. Additional architectural improvements include an enhanced Shader Execution Reordering (SER) scheduler to better utilize CUDA cores and increased L2 cache bandwidth.

Performance

Performance benchmarks demonstrated a substantial leap over the preceding GeForce 30 series, particularly in 4K resolution gaming and ray tracing titles. The GeForce RTX 4090 often doubled the performance of the GeForce RTX 3090 Ti in certain scenarios, especially when utilizing DLSS 3. In professional and creative applications from Adobe and Autodesk, the series also showed marked improvements in GPU acceleration tasks. Reviews from outlets like Tom's Hardware and AnandTech highlighted the generational performance gain, though noted increased power consumption in flagship models.

Features

A headline feature is DLSS 3, which combines Super Resolution, NVIDIA Reflex, and new AI-based Frame Generation to dramatically boost frame rates. The series also introduced AV1 encoding support via the eighth-generation NVENC encoder, beneficial for streamers using OBS Studio. Enhanced support for NVIDIA Broadcast and NVIDIA Omniverse was included, alongside full hardware compliance with the latest Microsoft DirectX 12 Ultimate and Vulkan API standards. NVIDIA Studio Driver support ensures optimization for creative software from Blender to DaVinci Resolve.

Product lineup

The initial launch in late 2022 featured the GeForce RTX 4090 and two versions of the GeForce RTX 4080, though the 12 GB model was later relaunched as the GeForce RTX 4070 Ti. Subsequent releases included the GeForce RTX 4070, GeForce RTX 4060 Ti, and GeForce RTX 4060, broadening the market reach. These cards were produced as Founders Edition models by NVIDIA and by Add-in board partners such as ASUS, MSI, and Gigabyte. The lineup was designed to compete across various segments against AMD's Radeon RX 7000 series offerings.

Category:NVIDIA graphics processing units