AI

Aware announced a strategic partnership with Luminous

Aware's purpose-built AI Data Platform and Luminous proprietary AI accelerator architecture accelerates adoption of highly secure AI across the Enterprises

Aware, the AI Data Platform for contextual intelligence, announced today a strategic partnership with Luminous, which designs and manufactures hardware for demanding generative AI inference applications. Luminous is backed by Gigafund, while Aware’s primary backer is Goldman Sachs. This collaboration harnesses the unparalleled power of Luminous’ hardware to supercharge Aware’s data platform, driving forward the next era of large-scale generative AI adoption.

Regulated businesses, multinational enterprises, and governmental bodies are increasingly seeking to scale generative AI within their infrastructures, while meeting the highest standards of security, privacy, and compliance. However, key challenges impede AI acceleration and adoption across Enterprise customers.

Fine-tuned LLMs are known to repeat or leak training data, which might be proprietary or confidential, across customers. This limits true conversational interaction. However, larger memory capacity, such as the Gen1 Luminous Inference AI card with 1+Terabyte of memory, can easily solve this by offering longer context windows since longer context windows allow a user to query and interact with proprietary data without the need for fine-tuning. Luminous Computing’s proprietary AI accelerator architecture solves these problems with 12x to 25x of memory on each card vs. competing products, accessible at HBM-level bandwidths. Furthermore, such large memory capacity allows users to scale up one-card-at-a-time without the need to tackle complicated distributed systems problems.

Pre-Built LLM’s often come riddled with hallucinations, inaccuracies, and require high operating costs so enterprises are transitioning to targeted machine learning models purpose-built to solve specific business challenges. Aware AI Data Platform has been purpose-built to train, deploy and refresh targeted, cost-effective machine learning models rapidly and responsibly. This addresses a critical need from businesses and government bodies seeking to build AI capabilities, especially in private cloud settings.

Combining Aware’s targeted, purpose-built machine learning models built for digital workplace conversations with Luminous’ AI accelerator architecture solves these challenges.

Key highlights of the partnership include:

Scaling AI Securely: Prioritizing security, Aware’s Generative AI may be deployed in highly secure, local enterprise environments, which is essential for sectors such as financial institutions, defense contractors, and government agencies that mandate airtight data protection. This partnership ensures not only the safeguarding of sensitive information but also the harnessing of generative AI’s full capabilities, catering especially to those entities eager to leverage generative AI within secure parameters.

Scaling AI Cost-Effectively: Compared to the use of Large Language Models (LLMs) in the cloud, the Aware platform powered by Luminous’ hardware for demanding generative AI inference applications provide significant cost efficiencies. Businesses can now run expansive AI operations without the financial strain of cloud-based LLMs.

“By integrating Luminous’s state-of-the-art Generative AI Inference chip into our platform, we’re set to redefine the boundaries of what’s possible in generative AI,” said Jeff Schumann, CEO of Aware. “This partnership underscores our commitment to offer solutions that are not only powerful but also secure, reliable, and cost-efficient.”

Marcus Gomez, the CEO of Luminous added, “Our collaboration with Aware will undoubtedly reshape the AI industry and revolutionize how businesses operate. Their expertise, coupled with our advanced chips, sets the stage for a new age of large-scale generative AI, transforming businesses across the board.”

Previous ArticleNext Article

Leave a Reply

Your email address will not be published. Required fields are marked *