Open source cloud AI startup Together is the latest developer to receive huge paydays as more and more investors pour money into artificial intelligence. $102.5 million Series A investment including tech giant NVIDIA.
“I think one of the things that excites investors is our focus on research,” a Together spokesperson said. decryption. “We have built a truly leading research team doing deep research on how to optimize both training and inference for these large-scale generative models. And this has been proven in some of the publications we have released (and) in our recently released inference engine. “This is about three times faster than anything else.”
American venture capital firm Kleiner Perkins led an investment of $102.5 million with participation from NVIDIA and Emergence Capital. Kleiner Perkins partners spoke together. Bucky Moore As a condition of the investment, you will join Together’s board of directors.
“(NVIDIA) is a partner and leader in this space,” said Vipul Ved Prakash, co-founder and CEO of Together. bloomberg. “This is the platform for AI today.”
Other companies participating in the Series A investment include NEA, Prosperity 7, Greycroft, 137 Ventures, Lux Capital, Definition Capital, and Long Journey Ventures. Last May, Together announced that the company had raised funding. 20 million dollars According to the report, seed funding venture beat.
“We focus a lot on price/performance, so the services we provide for inference servers are six times cheaper than Open AI, and our training services are four times cheaper than AWS,” Prakash claimed.
A Together spokesperson said the investment will be used to improve the cloud platform for AI applications, invest in research and development, contribute to open source AI, and strengthen infrastructure.
A spokesperson described the company’s privacy-centric approach, saying developers have the flexibility to decide on data storage options, giving them significant control over how their data is shared and used on the platform.
“By default, we do not use data from our customers or users to train our models,” the spokesperson said. “And (customers) can control whether data storage is turned on or off.”
While interviewing him bloombergVed Prakash said Nvidia was the only option to meet Together’s computing needs.
NVIDIA unveiled the upgraded Supercomputing 2023 at its annual Supercomputing 2023 event last November. grace hopper This superchip is named after a pioneering computer scientist and U.S. Navy admiral. NVIDIA has unveiled its quantum processing unit (QPU) development work to drive AI development.
NVIDIA earlier this year reported Quarterly revenue increased 101% compared to the same period in 2022, totaling $13.51 billion in revenue in the second quarter ending July 30.
The spokesperson added that Together AI strongly believes in open source as a key part of the future of generative AI, and that rather than a single dominant model emerging, a variety of proprietary and open source models emerge.
“You have deeper control as you fine-tune the open source model,” they said. “And most of our customers find that they can achieve higher accuracy in their operations by fine-tuning open source models than by using large proprietary models like GPT-4.”
The spokesperson emphasized the importance of using smaller, open-source models, which he said provide several benefits, including accuracy, ease of use, cost-effectiveness and improved performance.
“So for all of these reasons, we think open source is going to be a major driving force in how most organizations implement it. Generating AI“He concluded.
Founded in June 2022, San Francisco-based Together develops open source AI cloud software. The team is comprised of veterans from Apple, ETH Zurich, and scholars from Stanford University, University of California, Berkeley, and MIT.
Edited by Ryan Ozawa.