The Next Wave Of AI? How Hailo’s Edge AI Chips May Help Transform The Industry
Detroit, Michigan
|
March 28, 2024 08:45 AM Eastern Daylight Time
By Meg Flippin, Benzinga
Artificial intelligence is no longer restricted to big data centers, crunching complex data sets to help businesses increase productivity and improve processes. It is becoming ubiquitous as more use cases for this transformative technology emerge. One only has to look to Microsoft Corp.’s (NASDAQ: MSFT) Copilot and OpenAI’s ChatGPT for evidence. Relying on a large language model, the chatbots enable users to research and write content in seconds. That’s just one example of how AI is becoming more mainstream.
“We’re in the IBM mainframe era of AI, with the giant data centers representing a few giant computers that dominated the scene during the mainframe era,” George Gilder, the well-known investment advisor who has often been ahead of trends and predicted the world would shift to a decentralized model, said during a webinar for the community and the Gilder Private Reserve subscribers. “I think AI is going to be in every smartphone and every pocket, and it's going to be distributed through the Internet of Things. It's going to be ubiquitous. It's really a new IO for the new technology platform of the age.”
Hailo Sits On The Edge
That shift to decentralized AI where it no longer lives in the cloud or isolated in a data center, but rather on the edge, is where Hailo, the Israeli semiconductor startup, is operating. It sees a big opportunity for newcomers like itself to take advantage of the natural progression of AI.
“AI is learning from examples, contrary to plastic computers which are just being given a very specific set of rules. It’s kind of intuition versus logic,” Orr Danon, CEO at Hailo, said during an interview portion of the webinar with John Schroeder of the Gilder Private Reserve. “It’s a huge pivot in the kind of technology you need to implement this. It’s an opportunity for a new player to come and emerge and take a significant market share. And that's exactly what we are doing in the company.”
It is a big and growing market opportunity. Danon said the global chip market is about $600 billion today and is projected to cross the $1 trillion per annum mark at the end of the decade. Hailo makes AI vision processors and AI accelerator chips designed to accelerate embedded deep learning applications on edge devices. The company’s chips can be embedded in a variety of devices including autonomous vehicles, personal computers, smart cameras, robotics, industrial machinery, healthcare devices, drones and home appliances.
Hailo Does It Cheaper
While big chip companies like Nvidia Corp. (NASDAQ: NVDA) are focusing on making AI chips for data centers and cloud computing, Hailo is seeing growing demand for AI chips that sit on the edge, processing the data locally for use with everything from in a variety of devices including autonomous vehicles, personal computers, smart cameras, robotics, industrial machinery, healthcare devices and drones. Hailo says its chips are cheaper, more efficient and only require a few watts for power consumption. In comparison Nvidia's chip requires up to hundreds of watts, Hailo said.
“The apps that you run are local. Of course, you use data that is connected to the cloud, but the majority of data crunching is being done locally. And that's exactly what we believe in,” said Danon. “This is very beneficial in terms of power consumption of the communication requirements. If you're doing anything that requires a response quickly, like a car moving and needing to understand what's going on around it, you have to be local.” That means someday these edge AI chips will be able to make everyday devices even smarter whether it is a high-definition camera or a vacuum cleaner.
Cloud Will Always Have Its Role
That’s not to say there won’t be a need for the cloud when it comes to AI, it’s rather that they will complement each other and evolve. For example, cloud and big data centers will always be needed to train the AI, but when it comes to processing or performing the tasks the AI was trained to do, Hailo believes it will happen locally or on the edge. Danon said since the company launched in 2017, Hailo has seen the majority of AI deployments occurring outside of data centers – and that’s also happening with large language models like ChatGPT. “Over the past few years, we've seen the first wave of AI, which is based on fundamentally machine learning models trained by massive amounts of data,” said Danon. “But now we're seeing the second wave emerge. And that is, generative AI, which we are seeing in things like ChatGPT.”
AI is in its infancy, but it’s quickly moving outside of the big data center and into the devices we use every day. That presents a big opportunity for chip makers catering to this area of the market including Hailo. Interested in getting in on a startup before the second wave of AI takes off? Click here.
Featured photo courtesy of Hailo.
OurCrowd was started in 2013, driven by the idea that the business of building startups grows bigger and better when the global ‘crowd’ gains access to VC-level investment opportunities.Today, OurCrowd is a global venture and alternative investing platform that empowers institutions and individuals to invest and engage in emerging companies. OurCrowd vets and selects companies, invests its capital, and provides its global network with unparalleled access to co-invest and contribute connections, talent and deal flow. OurCrowd builds value for its portfolio companies throughout their lifecycles, providing mentorship, recruiting industry advisors, navigating follow-on rounds and creating growth opportunities through its network of multinational partnerships.
This post contains sponsored content. This content is for informational purposes only and is not intended to be investing advice.
Contact Details
Lisa Graston
Company Website
