$30 million in investments has been raised to help it take on giants like Intel and Nvidia by a computer chip start-up designing hardware for artificial intelligence (AI) applications recently.
Aimed to be used in applications from driverless cars to cloud computing, the U.K.-based Graphcore is aiming to ship its chips next year for that purpose. Cloud computing in driverless cars is an important process and involves large amounts of data being used to teach AI systems and is one that will require machine-learning technology.
In addition to the amount of energy consumed by the new processors, when it comes to how quickly the AI can learn, its processors will be up to 100 times better than current solutions in the market, Graphcore said.
Currently, machine-learning applications are run with the help of Graphic Processing Units (GPUs). However these are not suitable for the future, says Nigel Toon, the chief executive of Graphcore.
"GPUs have been built to run programs that completely describe the algorithm. Machine learning is different. You are trying to teach the system using data and that requires a different style of compute," Toon said in a television interview recently.
An Intelligent Processing Unit (IPU) system is being created to be marketed in 2017 by Graphcore.
In addition to venture capital firms Amadeus Capital Partners, C4 Ventures, Draper Esprit, Foundation Capital and Pitango Venture Capital, the funding round was led by Robert Bosch Venture Capital and included Samsung.
Toon claims that the technology will be able to help reduce data center costs and improve efficiencies and Graphcore has been working on its technology for the past two years. A social media company typically tends to have the most users at certain times of the day and this was the example that Toon gave to illustrate his point.
The downtime to train AI systems can be used by servers which would have Graphcore's IPU, and then deploy that update the following day in cases when user levels are low.
According to marketing intelligence firm Tractica, with spending on hardware for deep learning projects set to grow from $436 million in 2015 to $41.5 billion by 2024, the potential market for this technology is huge.
But incumbents Nvidia and Intel dominate the market at the present moment. Processors specifically designed for running artificial intelligence applications were unveiled by both companies earlier this year that are said to be the next generation processors.
The company said that it is already talking to its strategic investors about the next wave of applications even though the startup is initially aiming to sell to firms that are looking to train AI systems.
"We have Bosch as a strategic investor, we have Samsung as an investor and Bosch is interested in autonomous vehicles and the next generation of transportation, while Samsung is interested in missing word edge of network devices," Toon said.
"We are working with partners on some of these other applications."
(Source:www.cnbc.com)
Aimed to be used in applications from driverless cars to cloud computing, the U.K.-based Graphcore is aiming to ship its chips next year for that purpose. Cloud computing in driverless cars is an important process and involves large amounts of data being used to teach AI systems and is one that will require machine-learning technology.
In addition to the amount of energy consumed by the new processors, when it comes to how quickly the AI can learn, its processors will be up to 100 times better than current solutions in the market, Graphcore said.
Currently, machine-learning applications are run with the help of Graphic Processing Units (GPUs). However these are not suitable for the future, says Nigel Toon, the chief executive of Graphcore.
"GPUs have been built to run programs that completely describe the algorithm. Machine learning is different. You are trying to teach the system using data and that requires a different style of compute," Toon said in a television interview recently.
An Intelligent Processing Unit (IPU) system is being created to be marketed in 2017 by Graphcore.
In addition to venture capital firms Amadeus Capital Partners, C4 Ventures, Draper Esprit, Foundation Capital and Pitango Venture Capital, the funding round was led by Robert Bosch Venture Capital and included Samsung.
Toon claims that the technology will be able to help reduce data center costs and improve efficiencies and Graphcore has been working on its technology for the past two years. A social media company typically tends to have the most users at certain times of the day and this was the example that Toon gave to illustrate his point.
The downtime to train AI systems can be used by servers which would have Graphcore's IPU, and then deploy that update the following day in cases when user levels are low.
According to marketing intelligence firm Tractica, with spending on hardware for deep learning projects set to grow from $436 million in 2015 to $41.5 billion by 2024, the potential market for this technology is huge.
But incumbents Nvidia and Intel dominate the market at the present moment. Processors specifically designed for running artificial intelligence applications were unveiled by both companies earlier this year that are said to be the next generation processors.
The company said that it is already talking to its strategic investors about the next wave of applications even though the startup is initially aiming to sell to firms that are looking to train AI systems.
"We have Bosch as a strategic investor, we have Samsung as an investor and Bosch is interested in autonomous vehicles and the next generation of transportation, while Samsung is interested in missing word edge of network devices," Toon said.
"We are working with partners on some of these other applications."
(Source:www.cnbc.com)
Homepage
Send to a friend
Printable version
Share
