Hardware for artificial intelligence
<templatestyles src="Module:Message box/ambox.css"></templatestyles><templatestyles src="Multiple issues/styles.css"></templatestyles>
 | This article has multiple issues. Please help improve it or discuss these issues on the talk page. (Learn how and when to remove these messages)
Template:Expert needed<templatestyles src="Module:Message box/ambox.css"></templatestyles>
|
Specialized computer hardware is often used to execute artificial intelligence (AI) programs faster, and with less energy, such as Lisp machines, neuromorphic engineering, event cameras, and physical neural networks. Since 2017, several consumer grade CPUs and SoCs have on-die NPUs. As of 2023, the market for AI hardware is dominated by GPUs.[1]
Lisp machines
<templatestyles src="Module:Hatnote/styles.css"></templatestyles>
Lisp machines were developed in the late 1970s and early 1980s to make Artificial intelligence programs written in the programming language Lisp run faster.
Dataflow architecture
<templatestyles src="Module:Hatnote/styles.css"></templatestyles>
Dataflow architecture processors used for AI serve various purposes, with varied implementations like the polymorphic dataflow[2] Convolution Engine[3] by Kinara (formerly Deep Vision), structure-driven dataflow by Hailo,[4] and dataflow scheduling by Cerebras.[5]
Component hardware
AI accelerators
<templatestyles src="Module:Hatnote/styles.css"></templatestyles>
Since the 2010s, advances in computer hardware have led to more efficient methods for training deep neural networks that contain many layers of non-linear hidden units and a very large output layer.[6] By 2019, graphics processing units (GPUs), often with AI-specific enhancements, had displaced central processing units (CPUs) as the dominant means to train large-scale commercial cloud AI.[7] OpenAI estimated the hardware compute used in the largest deep learning projects from Alex Net (2012) to Alpha Zero (2017), and found a 300,000-fold increase in the amount of compute needed, with a doubling-time trend of 3.4 months.[8][9]
Sources
- ↑ Lua error: bad argument #1 to "get" (not a valid title).
- ↑ Lua error: bad argument #1 to "get" (not a valid title).
- ↑ Kinara (formerly Deep Vision). (2022) Retrieved 2022-12-11 from Kinara
- ↑ Hailo. Retrieved 2022-12-11 from Hailo
- ↑ Lua error: bad argument #1 to "get" (not a valid title).
- ↑ Deep Neural Networks for Acoustic Modeling in Speech Recognition. (23 October 2015) Retrieved 23 October 2015 from AIresearch.com
- ↑ Lua error: bad argument #1 to "get" (not a valid title).
- ↑ Lua error: bad argument #1 to "get" (not a valid title).
- ↑ AI and Compute. (16 May 2018) Retrieved 11 June 2020 from OpenAI
- Articles with short description
- Wikipedia articles in need of updating from November 2021
- Articles with invalid date parameter in template
- All Wikipedia articles in need of updating
- Articles with multiple maintenance issues
- Articles with hatnote templates targeting a nonexistent page
- Computer hardware
- Artificial intelligence
- Pages with script errors
