Symbolic Mind is working with Intel to bring its Charm AI model to enterprises that want to train and run artificial intelligence models using CPU-based server infrastructure.
The companies have completed a proof-of-concept using Intel Xeon 6 processors and are now moving the technology toward commercial deployment.
Symbolic Mind said Charm AI can be trained and served entirely on CPU servers, providing an alternative to AI architectures that rely heavily on GPUs.
The company’s Deep Control AI architecture is designed to reduce training and inference costs by allowing both processes to run on commodity CPU-based servers.
Symbolic Mind used Intel VTune Profiler and Intel Advisor during development and optimization of the technology.
Charm AI is also designed to continuously learn as customers add new data while keeping data, models, and intellectual property inside the customer’s own infrastructure.
Symbolic Mind plans to distribute Charm through several channels, including hyperscaler marketplaces, on-premises deployments, and standalone AI server appliances powered by Intel Xeon processors.
The company is particularly focused on regulated industries such as financial services, insurance, and healthcare, where organizations may have strict requirements around data security, infrastructure control, and model ownership.
Symbolic Mind also plans to offer an AI-Ready Rack as a turnkey product that would allow enterprise customers to train and run their own models using a relatively small number of CPU server racks.
The company’s patent-pending Deep Control neuro-symbolic architecture uses linear computational complexity rather than the quadratic attention complexity associated with transformer models.
Founded in 2023, Symbolic Mind has operations in the United States, Israel, India, and Brazil.
The collaboration with Intel is intended to give enterprises another way to deploy AI within existing computing environments while maintaining direct control over data, costs, and operations.
KEY QUOTES:
“We are going to deliver our Charm AI model through multiple channels, including but not limited to hyperscaler marketplaces, on-premises deployments and standalone AI server appliances, powered by Intel Xeon processors, especially for regulated industries such as finance, insurance and healthcare.”
Vamsi Sistla, CEO And Co-Founder Of Symbolic Mind
“Enterprises are looking for practical ways to leverage the power of AI without having to overcomplicate their IT infrastructure deployments. Symbolic Mind’s efforts underscores how organizations can train and run AI entirely on Intel Xeon 6 processors, enabling AI adoption on existing CPU-based infrastructure while maintaining control of their data, costs and operations.”
Lynn Comp, Head Of Xeon Market Readiness At Intel
