“Alexa, how will the weather be in Taipei next week?”
Artificial intelligence, which powers Amazon’s virtual assistant, Netflix, Spotify and other digital services, uses enormous computing resources to plow large amounts of data.
Israel’s New Reality reduces the high hardware costs and energy consumption required for AI processing. It has partnership agreements with IBM, AMD and other computing powerhouses and is ready to change the AI landscape.
“Teaming up with New Reality brings an AI-centric approach that disrupts the table, which is the type of industry collaboration we are looking for,” said Dr Mukesh Khare, Vice President, IBM Cloud. “The partnership with NeuReality is expected to lead to a more streamlined and accessible AI infrastructure that will improve people’s lives.”
New Reality has developed a new system architecture to connect multiple chips used in artificial intelligence. It is designed to facilitate and make AI processing more accessible to existing AI users, new industries and companies.
The technology dramatically reduces the cost of computing and networking chips, as well as power consumption – by a factor of 10 in the short term and 50 in the long term.
Launched in 2023, the game-changing platform will target public safety, e-commerce, healthcare, retail and other high-volume applications of AI.
With its next generation of hardware and software, NeuReality’s architecture reduces the overall cost of ownership and improves performance by minimizing system barriers created by existing architecture.
The company expects initial sales of $ 2.5 million this year based on its model and expects revenue to grow to $ 30 million by 2023. Tractica, a market intelligence company, estimates that the total market for AI chipsets will reach $ 75 billion by 2025.
The $ 20 million Series A funding round is currently underway at New Realty, led by Cardmen Capital and Varanasi Capital. Accredited investors can join the round through OurCrowd, a Jerusalem-based venture investing platform.
Although we do not think about it, AI is everywhere, once human intelligence performs the necessary functions: smart TVs, medical technology, autonomous vehicles, facial recognition, power grids, customer service boats. And it only grows.
Traditionally, the array of chips used in an AI application was connected via a CPU or central processing unit. Any command to the chip must go through the CPU. NeuReality’s technology replaces traditional CPU-centric support with the AI-centric platform, allowing components to communicate directly with AI accelerators – as well as connecting those AI accelerators to each other.
“If you have an engine that runs at 360 miles per hour and you put it in an old Subaru, it’s not built to handle such speed and can not hold a 1,000-horsepower engine,” said Moshe Tanach, CEO of Neurality.
“Why should we use a CPU-centric server to connect AI accelerators to the network? With our technology, data moves directly to the AI accelerator and we can do everything on the hardware, ”Tanach said.
A system board with $ 4,000 worth of AI chips currently needs $ 12,000 worth of computing and networking chips to support it. NeuReality plans to reduce this $ 12,000 price tag to $ 1,400.
The company is partnering with Xilinx, a unit of California-based AMD that designs and develops dynamic processing technology – to deliver its new platform.
Eagle Shameev, senior data center manager for the US company, said AMD’s AI accelerator can detect thousands of images per second. He said the New Reality platform could scale out a portion of the current price tag.
“Say you need 100,000 images per second instead of 5,000, so far you’ve put in so much hardware, you need a bigger computer and it consumes more resources,” Shameev said. “But with the technology of New Reality, we were able to take this function and implement a new component that scales it, so the customer’s total ownership cost is much lower than before.”
NeuReality is raising a round of funding. Accredited investors can participate through OurCrowd, a Jerusalem-based global investment platform. For more details, Click here.