---
title: "In 2016, Nvidia’s Deep Learning GPU Will Have 17 Billion Transistors"
description: "GPUs are increasingly being used to run artificial intelligence software, and it should fly on the next-generation Pascal chip from Nvidia, which reportedly will sport 17 billion transistors."
date: "2015-07-27"
modified: "2015-08-13"
authors:
  - name: "Kif Leswing"
    link: "https://futurism.com/authors/kif_leswing"
url: "https://futurism.com/in-2016-nvidias-best-deep-learning-gpu-will-have-17-billion-transistors"
categories:
  - "Artificial Intelligence"
tags:
  - "artificial intelligence"
---

# In 2016, Nvidia’s Deep Learning GPU Will Have 17 Billion Transistors

![](<https://futurism.com/wp-content/uploads/2015/07/PlanarFinFET.jpg>)

- The big increase in number of transistors comes because Nvidia is using a 16nm FinFET process. The chips will be fabricated by TSMC.
- The current generation, Maxwell, is based on a 28nm process. Pascal is expected to be commercially available in 2016.
- Pascal processors will support 3D memory and mixed prevision floating point calcuation, which are both useful for deep learning.