ThirdAI Raises $6M to Run AI on Any Hardware

ThirdAI Secures $6 Million in Seed Funding to Accelerate Deep Learning

ThirdAI, a Houston-based company focused on accelerating deep learning technology, has successfully raised $6 million in seed funding. The company’s innovative approach aims to reduce reliance on specialized hardware, such as graphics processing units (GPUs).

Investment Details and Future Plans

The investment round was co-led by Neotribe Ventures, Cervin Ventures, and Firebolt Ventures. According to Anshumali Shrivastava, ThirdAI’s co-founder and CEO, the funds will be allocated towards expanding the team and bolstering computing resources.

The Genesis of ThirdAI

Shrivastava, possessing a background in mathematics, has long held an interest in artificial intelligence and machine learning. His focus centered on developing more efficient AI methodologies. He initiated ThirdAI in April, collaborating with graduate students from Rice University, to translate this vision into reality.

A Novel Approach to Deep Learning

ThirdAI’s technology represents “a smarter approach to deep learning.” Its algorithm and software innovations enable standard central processing units (CPUs) to surpass the performance of GPUs when training extensive neural networks.

Historically, companies transitioned away from CPUs due to GPUs’ superior ability to concurrently render high-resolution visuals. However, Shrivastava points out that GPUs often have limited memory, creating a bottleneck during AI development.

Addressing the Limitations of Current Technology

“Our assessment of the deep learning landscape revealed that much of the underlying technology dates back to the 1980s,” Shrivastava explained. “Despite this, approximately 80% of the market relies on GPUs, incurring significant costs in both hardware and specialized engineering expertise, all while awaiting the desired outcomes from AI.”

His team sought to develop a cost-effective alternative to GPUs, anticipating the future trajectory of AI development. Their “sub-linear deep learning engine” algorithm leverages CPUs, eliminating the need for specialized acceleration hardware.

Neotribe’s Perspective on ThirdAI

Swaroop “Kittu” Kolluri, founder and managing partner at Neotribe, acknowledges that this technology is still in its nascent stages. He notes that existing methods are often time-consuming, expensive, and slow, particularly when dealing with memory-intensive tasks like running large language models.

“ThirdAI offers a solution where companies can achieve optimal results without compromising on resources,” Kolluri stated. “This is the primary reason for our investment. It’s not solely about computing power, but also about memory capacity, and ThirdAI democratizes access to this capability, potentially revolutionizing the field.”

The Broader Implications of AI

Kolluri highlights that AI is now capable of tackling complex challenges in fields like healthcare and seismic processing. However, he also raises concerns about the environmental impact of running AI models.

“The energy consumption associated with training deep learning models can be substantial,” Shrivastava emphasized. “The cost can even exceed that of owning multiple vehicles over a lifetime. As we scale AI applications, it’s crucial to address these considerations.”