Home > Press > Light meets deep learning: computing fast enough for next-gen AI
|
A team of Greek academic researchers and California entrepreneurs benchmarked their Silicon Photonic (SiPho) neural network technology against processing unit currently on the market and six-year-old technology with projections. CREDIT
Authors of publication |
Abstract:
Artificial intelligence (AI) models are essential for sophisticated image classification, the most important part of digital analysis. The researchers who recently published “Universal Linear Optics Revisited: New Perspectives for Neuromorphic Computing with Silicon Photonics” have moved the needle for image classification. The speeds they’ve achieved on a new chip platform (silicon photonics) using the computational power of neural networks is impressive.
Light meets deep learning: computing fast enough for next-gen AI
Piscataway, NJ | Posted on March 24th, 2023
Nonetheless, pay attention here to the modal auxiliary verb “can.” Just because something can be done, questions remain. Will it be fast enough? Will it have sufficient accuracy? How energy efficient is it? Is the chip large and unwieldly? This research tackles them all.
One of the attributes of AI is that you can use it at the edge of the physical network; in a camera for example. A camera on a drone is an even better example. To enable a drone with AI, you want the on-board AI chip to be powerful, but energy efficient, small and lightweight, and able to do lots of complex math at lightning speed. That way, the drone can alert humans when something untoward is detected (cancer, a saboteur, damage to a train-track).
Meanwhile, in Greece, researchers have built a neuromorphic photonic processor computing at a speed of 50 GHz that is capable to classify images with ~95% accuracy. Let’s break this down, starting with the photonic part.
After Silicon Electronics? Silicon Photonics.
AI processor chips often start life as graphic processing units (GPUs) for high-end video games or tensor processing units (TPUs) which are specifically designed for neural networks, meaning computation mimicking the human brain. (Except that they like linear algebra!) However conventional processors use silicon electronics as the physical platform, which is reaching quantum limitations.
Switching from electrons to photons increases computational ability because the speed of light is so much faster than the speed of electrons. It’s more energy efficient too. The “wires” don’t heat up. The physics of light can be used for matrix-vector multiplication operations, the computational backbone of neural networks.
After Conventional Math? Neuromorphic Computing with Trillions of Operations per Second
Now the neuromorphic part. The Greek research team, along with Celestial AI, developed a novel design for the chip using a crossbar layout. The layout outperforms the state-of-the-art photonic counterparts in terms of scalability, technical versatility, ease of programming and error tolerance. Said differently, by combining the crossbar layout’s architectural benefits with SiGe electro-absorption modulators employed in their first prototype, the researchers project that a purely optical implementation can perform trillions of matrix-vector multiplications per second, without sacrificing the processing accuracy, while consuming very low power.
Compared with six years ago, silicon photonics is in a much better position to pull neural morphic processors from their currently low computational and physical size (footprint) efficiency to less unwieldy. Notice in Figure 1 the placement of IBM’s TrueNorth chip, Intel’s Loihi chip, the HICANN (High Input Count Analog Neural Network) chip from Germany’s Heidelberg University and Stanford U’s neurogrid device. Compare it to the crossbar-layout chips discussed here, which are falling right along silicon photonics roadmap in terms of computation and size efficiency. The synergy of powerful photonics with the novel crossbar architecture can enable next generation neuromorphic computing engines. Let’s change that that modal auxiliary verb to “will.”
####
For more information, please click here
Contacts:
Media Contact
Kristen Mahan
Institute of Electrical and Electronics Engineers
Expert Contact
Kristen Mahan
IEEE
Office: 17322723320
Copyright © Institute of Electrical and Electronics Engineers
If you have a comment, please Contact us.
Issuers of news releases, not 7th Wave, Inc. or Nanotechnology Now, are solely responsible for the accuracy of the content.
Bookmark:
ARTICLE TITLE
News and information
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Robot caterpillar demonstrates new approach to locomotion for soft robotics March 24th, 2023
Semiconductor lattice marries electrons and magnetic moments March 24th, 2023
Bilayer PET/PVDF substrate-reinforced solid polymer electrolyte improves solid-state lithium metal battery performance March 24th, 2023
Possible Futures
New experiment translates quantum information between technologies in an important step for the quantum internet March 24th, 2023
Graphene grows – and we can see it March 24th, 2023
HKUMed invents a novel two-dimensional (2D) ultrasound-responsive antibacterial nano-sheets to effectively address bone tissue infection March 24th, 2023
A universal HCl-assistant powder-to-powder strategy for preparing lead-free perovskites March 24th, 2023
Chip Technology
Graphene grows – and we can see it March 24th, 2023
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Semiconductor lattice marries electrons and magnetic moments March 24th, 2023
Researchers develop innovative tool for measuring electron dynamics in semiconductors: Insights may lead to more energy-efficient chips and electronic devices March 3rd, 2023
Optical computing/Photonic computing
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Photonic Materials: Recent Advances and Emerging Applications February 10th, 2023
New study opens the door to ultrafast 2D devices that use nonequilibrium exciton superdiffusion February 10th, 2023
Manufacturing advances bring material back in vogue January 20th, 2023
Discoveries
New experiment translates quantum information between technologies in an important step for the quantum internet March 24th, 2023
Graphene grows – and we can see it March 24th, 2023
HKUMed invents a novel two-dimensional (2D) ultrasound-responsive antibacterial nano-sheets to effectively address bone tissue infection March 24th, 2023
A universal HCl-assistant powder-to-powder strategy for preparing lead-free perovskites March 24th, 2023
Announcements
Robot caterpillar demonstrates new approach to locomotion for soft robotics March 24th, 2023
Semiconductor lattice marries electrons and magnetic moments March 24th, 2023
Bilayer PET/PVDF substrate-reinforced solid polymer electrolyte improves solid-state lithium metal battery performance March 24th, 2023
Understanding the mechanism of non-uniform formation of diamond film on tools: Paving the way to a dry process with less environmental impact March 24th, 2023
Interviews/Book Reviews/Essays/Reports/Podcasts/Journals/White papers/Posters
HKUMed invents a novel two-dimensional (2D) ultrasound-responsive antibacterial nano-sheets to effectively address bone tissue infection March 24th, 2023
A universal HCl-assistant powder-to-powder strategy for preparing lead-free perovskites March 24th, 2023
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Robot caterpillar demonstrates new approach to locomotion for soft robotics March 24th, 2023
Artificial Intelligence
Stanford researchers develop a new way to identify bacteria in fluids: An innovative adaptation of the technology in an old inkjet printer plus AI-assisted imaging leads to a faster, cheaper way to spot bacteria in blood, wastewater, and more March 3rd, 2023
3D-printed decoder, AI-enabled image compression could enable higher-res displays December 9th, 2022
New chip ramps up AI computing efficiency August 19th, 2022
Artificial Intelligence Centered Cancer Nanomedicine: Diagnostics, Therapeutics and Bioethics June 3rd, 2022
Photonics/Optics/Lasers
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Stanford researchers develop a new way to identify bacteria in fluids: An innovative adaptation of the technology in an old inkjet printer plus AI-assisted imaging leads to a faster, cheaper way to spot bacteria in blood, wastewater, and more March 3rd, 2023
Photonic Materials: Recent Advances and Emerging Applications February 10th, 2023
New study opens the door to ultrafast 2D devices that use nonequilibrium exciton superdiffusion February 10th, 2023
- SEO Powered Content & PR Distribution. Get Amplified Today.
- Platoblockchain. Web3 Metaverse Intelligence. Knowledge Amplified. Access Here.
- Source: http://www.nanotech-now.com/news.cgi?story_id=57317
Light meets deep learning: computing fast enough for next-gen AI
Republished By Plato
Home > Press > Light meets deep learning: computing fast enough for next-gen AI
Authors of publication
Abstract:
Artificial intelligence (AI) models are essential for sophisticated image classification, the most important part of digital analysis. The researchers who recently published “Universal Linear Optics Revisited: New Perspectives for Neuromorphic Computing with Silicon Photonics” have moved the needle for image classification. The speeds they’ve achieved on a new chip platform (silicon photonics) using the computational power of neural networks is impressive.
Light meets deep learning: computing fast enough for next-gen AI
Piscataway, NJ | Posted on March 24th, 2023
Nonetheless, pay attention here to the modal auxiliary verb “can.” Just because something can be done, questions remain. Will it be fast enough? Will it have sufficient accuracy? How energy efficient is it? Is the chip large and unwieldly? This research tackles them all.
One of the attributes of AI is that you can use it at the edge of the physical network; in a camera for example. A camera on a drone is an even better example. To enable a drone with AI, you want the on-board AI chip to be powerful, but energy efficient, small and lightweight, and able to do lots of complex math at lightning speed. That way, the drone can alert humans when something untoward is detected (cancer, a saboteur, damage to a train-track).
Meanwhile, in Greece, researchers have built a neuromorphic photonic processor computing at a speed of 50 GHz that is capable to classify images with ~95% accuracy. Let’s break this down, starting with the photonic part.
After Silicon Electronics? Silicon Photonics.
AI processor chips often start life as graphic processing units (GPUs) for high-end video games or tensor processing units (TPUs) which are specifically designed for neural networks, meaning computation mimicking the human brain. (Except that they like linear algebra!) However conventional processors use silicon electronics as the physical platform, which is reaching quantum limitations.
Switching from electrons to photons increases computational ability because the speed of light is so much faster than the speed of electrons. It’s more energy efficient too. The “wires” don’t heat up. The physics of light can be used for matrix-vector multiplication operations, the computational backbone of neural networks.
After Conventional Math? Neuromorphic Computing with Trillions of Operations per Second
Now the neuromorphic part. The Greek research team, along with Celestial AI, developed a novel design for the chip using a crossbar layout. The layout outperforms the state-of-the-art photonic counterparts in terms of scalability, technical versatility, ease of programming and error tolerance. Said differently, by combining the crossbar layout’s architectural benefits with SiGe electro-absorption modulators employed in their first prototype, the researchers project that a purely optical implementation can perform trillions of matrix-vector multiplications per second, without sacrificing the processing accuracy, while consuming very low power.
Compared with six years ago, silicon photonics is in a much better position to pull neural morphic processors from their currently low computational and physical size (footprint) efficiency to less unwieldy. Notice in Figure 1 the placement of IBM’s TrueNorth chip, Intel’s Loihi chip, the HICANN (High Input Count Analog Neural Network) chip from Germany’s Heidelberg University and Stanford U’s neurogrid device. Compare it to the crossbar-layout chips discussed here, which are falling right along silicon photonics roadmap in terms of computation and size efficiency. The synergy of powerful photonics with the novel crossbar architecture can enable next generation neuromorphic computing engines. Let’s change that that modal auxiliary verb to “will.”
####
For more information, please click here
Contacts:
Media Contact
Kristen Mahan
Institute of Electrical and Electronics Engineers
Expert Contact
Kristen Mahan
IEEE
Office: 17322723320
Copyright © Institute of Electrical and Electronics Engineers
If you have a comment, please Contact us.
Issuers of news releases, not 7th Wave, Inc. or Nanotechnology Now, are solely responsible for the accuracy of the content.
Bookmark:
ARTICLE TITLE
News and information
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Robot caterpillar demonstrates new approach to locomotion for soft robotics March 24th, 2023
Semiconductor lattice marries electrons and magnetic moments March 24th, 2023
Bilayer PET/PVDF substrate-reinforced solid polymer electrolyte improves solid-state lithium metal battery performance March 24th, 2023
Possible Futures
New experiment translates quantum information between technologies in an important step for the quantum internet March 24th, 2023
Graphene grows – and we can see it March 24th, 2023
HKUMed invents a novel two-dimensional (2D) ultrasound-responsive antibacterial nano-sheets to effectively address bone tissue infection March 24th, 2023
A universal HCl-assistant powder-to-powder strategy for preparing lead-free perovskites March 24th, 2023
Chip Technology
Graphene grows – and we can see it March 24th, 2023
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Semiconductor lattice marries electrons and magnetic moments March 24th, 2023
Researchers develop innovative tool for measuring electron dynamics in semiconductors: Insights may lead to more energy-efficient chips and electronic devices March 3rd, 2023
Optical computing/Photonic computing
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Photonic Materials: Recent Advances and Emerging Applications February 10th, 2023
New study opens the door to ultrafast 2D devices that use nonequilibrium exciton superdiffusion February 10th, 2023
Manufacturing advances bring material back in vogue January 20th, 2023
Discoveries
New experiment translates quantum information between technologies in an important step for the quantum internet March 24th, 2023
Graphene grows – and we can see it March 24th, 2023
HKUMed invents a novel two-dimensional (2D) ultrasound-responsive antibacterial nano-sheets to effectively address bone tissue infection March 24th, 2023
A universal HCl-assistant powder-to-powder strategy for preparing lead-free perovskites March 24th, 2023
Announcements
Robot caterpillar demonstrates new approach to locomotion for soft robotics March 24th, 2023
Semiconductor lattice marries electrons and magnetic moments March 24th, 2023
Bilayer PET/PVDF substrate-reinforced solid polymer electrolyte improves solid-state lithium metal battery performance March 24th, 2023
Understanding the mechanism of non-uniform formation of diamond film on tools: Paving the way to a dry process with less environmental impact March 24th, 2023
Interviews/Book Reviews/Essays/Reports/Podcasts/Journals/White papers/Posters
HKUMed invents a novel two-dimensional (2D) ultrasound-responsive antibacterial nano-sheets to effectively address bone tissue infection March 24th, 2023
A universal HCl-assistant powder-to-powder strategy for preparing lead-free perovskites March 24th, 2023
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Robot caterpillar demonstrates new approach to locomotion for soft robotics March 24th, 2023
Artificial Intelligence
Stanford researchers develop a new way to identify bacteria in fluids: An innovative adaptation of the technology in an old inkjet printer plus AI-assisted imaging leads to a faster, cheaper way to spot bacteria in blood, wastewater, and more March 3rd, 2023
3D-printed decoder, AI-enabled image compression could enable higher-res displays December 9th, 2022
New chip ramps up AI computing efficiency August 19th, 2022
Artificial Intelligence Centered Cancer Nanomedicine: Diagnostics, Therapeutics and Bioethics June 3rd, 2022
Photonics/Optics/Lasers
Optical switching at record speeds opens door for ultrafast, light-based electronics and computers: March 24th, 2023
Stanford researchers develop a new way to identify bacteria in fluids: An innovative adaptation of the technology in an old inkjet printer plus AI-assisted imaging leads to a faster, cheaper way to spot bacteria in blood, wastewater, and more March 3rd, 2023
Photonic Materials: Recent Advances and Emerging Applications February 10th, 2023
New study opens the door to ultrafast 2D devices that use nonequilibrium exciton superdiffusion February 10th, 2023
Google Maps will have AI-powered summaries for easier charging station navigation
Get University Level Certified for Next to Nothing – KDnuggets
Citroën enters compact SUV market with new C3 Aircross
2025 Toyota Camry First Drive Review: Hybrid-only overhaul is a hit – Autoblog
Transavia will fly from Dutch airports to Granada, Tirana, Tbilisi and Oslo next winter, but also from Brussels to Gran Canaria
Eesee Implements Blockpass for Compliance in Digital Assets Marketplace
5 Common Uses for Grommets
Master Data vs. Reference Data – DATAVERSITY
Binance Could Come Back to India by Paying $2M Fine: Report