Nvidia-Backed Together AI Secures $102.5 Million in Funding

Open-source cloud AI startup Together has successfully raised $102.5 million in Series A investments, with participation from industry giant NVIDIA. The funding, led by venture capital firm Kleiner Perkins, underscores the growing interest in artificial intelligence. Together's emphasis on research and the optimization of training and inference for large generative models has garnered attention.

Leading Research and Speedy Inference Engine

Together's spokesperson highlighted the appeal to investors, citing the company's focus on research. The development of a leading research team has resulted in publications and a recently launched inference engine that outpaces competitors threefold. Kleiner Perkins, along with NVIDIA and Emergence Capital, led the Series A investment, with Bucky Moore from Kleiner Perkins set to join the Together board.

Strategic Partnerships and Cost Efficiency

"NVIDIA is a partner and a leading company in this space. They are the platform today for AI," said Vipul Ved Prakash, Together's co-founder and CEO. The funding round included participation from NEA, Prosperity 7, Greycroft, and others. Together aims to prioritize price performance, claiming services are more cost-effective compared to competitors, with a focus on enhancing its cloud platform for AI applications.

Investment Allocation and Privacy-Focused Approach

The spokesperson outlined the allocation of funds, emphasizing improvements to the cloud platform, research and development, contributions to open-source AI, and infrastructure enhancement. Together underscores its privacy-focused approach, granting developers control over data usage. By default, no customer data is used for model training, and customers can manage data storage preferences.

Nvidia's Role and Industry Trends

During an interview with Bloomberg, Ved Prakash stated that Nvidia was the sole choice for Together's computing needs. Nvidia, a key player in AI, recently unveiled upgraded Grace Hopper Superchips and announced work on developing quantum processing units (QPUs) during the Supercomputing 2023 event. Together AI sees open-source as crucial for the future of generative AI, favoring fine-tuning of smaller models for improved accuracy, ease of use, cost-effectiveness, and enhanced performance.

Future Prospects and Open-Source Advocacy

Together AI, established in June 2022 in San Francisco, envisions a future where a variety of proprietary and open-source models coexist, providing depth of control for organizations implementing generative AI. The company, comprising veterans from Apple, ETH Zurich, and academics from top universities, believes open source will be a major force in shaping the landscape of generative AI implementation.