GeForce 16 series
Entry to mid-range Turing GPUs without ray tracing cores
Nvidia launched this graphics processor line in February 2019 for budget gamers. Built on the Turing architecture, it omits dedicated ray tracing hardware found in higher-end models. The series covers entry and mid-range segments, serving as a cost-effective alternative to the RTX 20 lineup.
- Based on Turing architecture without RT or Tensor cores
- Manufactured by TSMC on a 12 nm FinFET process
- Launched in February 2019 with the GTX 1660 Ti
- Includes models from GTX 1630 to GTX 1660 Ti
- Last GeForce series without hardware ray tracing
What it is
The GeForce GTX 16 series comprises graphics processing units designed by Nvidia for personal computers. Announced in early 2019, these chips target entry-level and mid-range markets. They utilize the Turing microarchitecture but exclude the specialized cores for artificial intelligence and real-time ray tracing. This distinction separates them from the concurrent RTX 20 series. Manufactured by TSMC using a 12 nanometre process, the lineup includes models like the GTX 1650 and GTX 1660. It serves users seeking modern performance without the premium cost of high-end features.
How it came about
Nvidia introduced the series in February 2019, starting with the GTX 1660 Ti. The launch aimed to fill the gap left by the expensive RTX 20 cards. Media outlets frequently compared these new chips to AMD’s Radeon RX 500 series. In April 2019, Nvidia enabled software-based ray tracing via driver updates for select models. The GTX 1630 arrived much later in June 2022. Support for Windows 7 ended in September 2024, marking the end of legacy driver updates for this generation.
Inside the chip
These GPUs share the underlying Turing design with the RTX 20 series. However, they lack Tensor and RT cores. Dedicated integer cores remain, allowing simultaneous integer and floating-point calculations. The TU117 variant uses an older NVENC encoder from the Volta generation. It also lacks support for Nvidia Optical Flow. Memory options include GDDR5 and faster GDDR6 types. The interface connects via PCIe 3.0 x16. Feature support extends to DirectX 12, Vulkan 1.4, and OpenGL 4.6.
How it was made
TSMC manufactures the silicon using its 12 nm FinFET process technology. Transistor counts vary by model, ranging from 4.7 billion in the TU117 die to 10.8 billion in the TU106. The TU116 variant contains 6.6 billion transistors. This production method balances performance and cost for the intended market segment. The physical cards connect to motherboards through standard PCIe slots. Power requirements differ, with some models drawing up to 100 watts.
Where it went
Primary applications include gaming on desktop PCs and laptops. The GTX 1650 saw integration into portable systems in April 2019. Reviewers noted its suitability for small, power-efficient builds. The series handles modern titles at moderate settings. It supports APIs required for current games, including Direct3D 12 and Vulkan. While not intended for professional AI workloads due to missing Tensor cores, it serves general consumer graphics needs effectively.
Versions and family
The lineup spans from the low-end GTX 1630 to the mid-range GTX 1660 Ti. Models include the GTX 1650, 1650 Super, 1660, and 1660 Super. The GTX 1650 exists in both GDDR5 and GDDR6 memory configurations. Laptop-specific variants like the GTX 1650 Ti were also released. The series sits between the older GeForce 10 family and the newer RTX 30 generation. It represents the final GeForce line without hardware-accelerated ray tracing.
Why it matters
This series provided a budget-friendly entry point into the Turing architecture. It allowed gamers to access newer driver features without buying RTX cards. The GTX 1650 Super was praised for its value proposition in 2019. However, the late arrival of the GTX 1630 drew criticism for poor price-to-performance ratios. The series helped Nvidia maintain market share against AMD’s competitive offerings in the lower tiers.
Problems and reception
Reviews were mixed across the range. The GTX 1650 Super received strong recommendations for its speed boost over the base model. Critics cited limited video memory as a drawback. The original GTX 1650 faced competition from AMD’s RX 570. The GTX 1630 was poorly received in 2022, with reviewers calling it outdated silicon. Some experts advised buyers to choose older, faster models instead. The series generally met expectations for its price bracket, except for the final release.
Questions people ask about GeForce 16 series
Does the GeForce 16 series support ray tracing?
It lacks dedicated hardware for real-time ray tracing. However, Nvidia enabled software-based DirectX Raytracing via driver updates for certain models like the GTX 1660. Performance is limited compared to RTX cards with dedicated RT cores.
What is the difference between GTX 1650 and 1650 Super?
The Super variant is significantly faster, offering around thirty percent better performance. It uses GDDR6 memory and has higher power consumption. The original model may use GDDR5 and draws less power, fitting systems without extra PCIe cables.
Who manufactures the GeForce 16 series chips?
Nvidia designs the GPUs, while TSMC manufactures them. The production uses a 12 nanometre FinFET process. This partnership allows Nvidia to produce cost-effective chips for the entry and mid-range market segments.
Is the GTX 1630 a good buy today?
Reviews from its 2022 launch were negative. Critics noted it performed worse than older cards like the RX 470. Experts recommended the GTX 1650 or 1650 Super instead, citing better value and performance for similar prices.
Good reading and sources
- Official websiteOfficial site
- Nvidia GeForce GTX 1660 Ti could launch on February 15 with a sub-$300 priceAnandTech, Article
- Nvidia GeForce GTX 1650 4GB Review: This Turing Fails the TestTom's Hardware, Article
- GeForce GTX 16 series on WikipediaEncyclopedia
- More photos on Wikimedia CommonsPhotos
- Wikidata recordOpen data
The facts on this page come from the Wikipedia article "GeForce GTX 16 series", 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.
Like GeForce 16 series
All graphics processorsMore from Nvidia
All of them
🇺🇸 United StatesApple A13 Bionic
64-bit ARM system on a chip for mobile devicesSoC · Arm · 2019
🇯🇵 JapanFujitsu A64FX
48-core ARM supercomputer CPU with 512-bit SVE and HBM2, built for FugakuCPU · Arm · 2019
🇺🇸 United StatesApple U1
Ultra-wideband system on a chip for spatial awarenessSoC · Arm · 2019