News
IndustryTSMC’s grip on advanced chipmaking will face competitionIndustryCadence joins European Chip Design Platform to support semiconductor startups and SMEsMaking chipsMarc Hijink on ASML’s Culture, Supply Chain, and GeopoliticsProcessorsIntel's post-Nova Lake CPUs take shape as Razor Lake desktop and laptop variants surfaceAI chipsHow Google’s New AI Chip Software Helps It Battle NvidiaIndiaMore IC Packaging Plants Emerge In IndiaIndustryUS Tightens China AI Chip Access as Nvidia, TSMC and Semiconductor Stocks SlideDesignAxelera AI moves into HPC with TitaniaMaking chipsChinese envoy says chip equipment export curbs threaten EU-China trade thawIndiaSEMICON India 2026 Day 2 Sees 25 MoUs, Strategic Collaborations Across Semiconductor Value ChainIndustryTSMC’s grip on advanced chipmaking will face competitionIndustryCadence joins European Chip Design Platform to support semiconductor startups and SMEsMaking chipsMarc Hijink on ASML’s Culture, Supply Chain, and GeopoliticsProcessorsIntel's post-Nova Lake CPUs take shape as Razor Lake desktop and laptop variants surfaceAI chipsHow Google’s New AI Chip Software Helps It Battle NvidiaIndiaMore IC Packaging Plants Emerge In IndiaIndustryUS Tightens China AI Chip Access as Nvidia, TSMC and Semiconductor Stocks SlideDesignAxelera AI moves into HPC with TitaniaMaking chipsChinese envoy says chip equipment export curbs threaten EU-China trade thawIndiaSEMICON India 2026 Day 2 Sees 25 MoUs, Strategic Collaborations Across Semiconductor Value Chain
EveryChip
Nvidia GeForce RTX 2080 (TU104 | Turing) (Package | Close-Up)

Turing

Nvidia GPU microarchitecture with real-time ray tracing and AI cores

Photo: Fritzchens Fritz, CC0, via Wikimedia Commons

Turing is a graphics processing unit microarchitecture developed by Nvidia and launched in 2018. It introduced dedicated ray tracing and artificial intelligence processors to consumer hardware, enabling real-time visual effects and deep learning acceleration.

Share on WhatsApp

  • First consumer GPU with dedicated ray tracing cores
  • Manufactured on TSMC 12FFC process node
  • Includes Tensor cores for AI and deep learning
  • Supports GDDR6, GDDR5, and HBM2 memory types
  • Successor to Pascal and predecessor to Ampere

What it is

Turing is a graphics processing unit microarchitecture created by Nvidia. It was first revealed in August 2018 and reached consumers the following month. This design marked a significant shift by integrating hardware specifically for ray tracing and artificial intelligence tasks. Unlike previous generations that relied solely on traditional rasterization, Turing uses specialized cores to handle complex lighting calculations and matrix operations. It serves as the foundation for several product lines, including gaming cards and professional workstation accelerators.

How it came about

Nvidia named the architecture after mathematician Alan Turing. Development built upon the Volta design, which was reserved for high-performance computing. The company introduced Turing at SIGGRAPH 2018 for professional Quadro RTX cards. One week later, they unveiled the consumer GeForce 20 series at Gamescom. This rapid rollout brought advanced rendering techniques to the mainstream market. In early 2019, Nvidia expanded the family with the GeForce 16 series, which used the same underlying design but omitted the new specialized cores to offer a lower-cost option.

Inside the chip

The architecture combines three main types of processing units. Standard CUDA cores handle traditional shading and compute tasks, supporting concurrent integer and floating-point operations. New RT cores accelerate ray tracing by processing bounding volume hierarchies and triangle collisions. Tensor cores manage artificial intelligence workloads, such as deep learning super sampling for image upscaling. The design supports GDDR6, GDDR5, and HBM2 memory. It also includes NVENC for video encoding and DisplayPort 1.4a outputs. Compute Capability is rated at 7.5.

How it was made

TSMC manufactures the chips using its 12 nanometre FinFET process, referred to as 12FFC. The largest die, known as TU102, contains 18.6 billion transistors. Memory components come from suppliers like Samsung Electronics and Micron Technology. The initial consumer launch used Micron chips, but production shifted to Samsung parts by late 2018. The fabrication process allows for high transistor density while maintaining power efficiency suitable for desktop and laptop forms.

Turing
A drawing made by EveryChip from what this page says. It is a diagram, not a picture of the real chip.

Where it went

This microarchitecture powers the GeForce 20 series and GeForce 16 series for personal computers and laptops. It also drives the Quadro RTX line for professional workstations and the Tesla T4 for servers. Gamers use it for real-time reflections and shadows, while professionals leverage it for rendering and AI inference. The architecture supports APIs like DirectX 12 Ultimate, Vulkan, and OpenGL 4.6. It enables features such as VirtualLink for virtual reality headsets and hardware-accelerated video decoding.

Versions and family

Key variants include the high-end TU102 and other dies in the TU10x and TU11x families. The GeForce 16 series represents a cut-down version without RT or Tensor cores. Professional cards include the Quadro RTX 3000 through 8000. Server applications use the Tesla T4. The architecture succeeds Pascal and precedes Ampere. It supports a wide range of products from entry-level laptop GPUs to flagship desktop cards like the RTX 2080 Ti and Titan RTX.

Why it matters

Turing brought real-time ray tracing to consumer hardware, a goal long pursued by the graphics industry. It increased ray tracing performance significantly compared to the previous Pascal generation. The inclusion of Tensor cores allowed for AI-driven image enhancement techniques like DLSS. This design influenced subsequent architectures by establishing dedicated hardware for non-rasterization tasks. It set a new standard for visual fidelity in games and professional visualization applications.

Questions people ask about Turing

What makes Turing different from previous Nvidia architectures?

Turing introduces dedicated RT cores for ray tracing and Tensor cores for artificial intelligence. Previous architectures like Pascal relied on standard CUDA cores for all tasks. This allows Turing to handle complex lighting and AI upscaling much more efficiently.

Which products use the Turing microarchitecture?

Turing powers the GeForce 20 series, GeForce 16 series, Quadro RTX cards, and Tesla T4 servers. It covers a wide range from gaming laptops to professional workstations and data centre accelerators.

Does every Turing chip support ray tracing?

No. The GeForce 16 series uses the Turing design but lacks the specific RT and Tensor cores required for hardware-accelerated ray tracing. Only the RTX-branded cards and Quadro RTX models include these specialized units.

Good reading and sources

Where this page comes from

The facts on this page come from the Wikipedia article "Turing (microarchitecture)", written by its contributors and used under the CC BY-SA 4.0 licence, and from its Wikidata record (CC0). EveryChip rewrote the text in its own words, with AI assistance, and compares every page with the article to make sure no passage is copied. It may be reused under the same CC BY-SA 4.0 licence with a link back.

It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.

Photo credits on this page

The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

The weekly digest

Every Monday: the chip of the day, new pages on EveryChip, and the conferences and anniversaries of the week ahead.