Intel and Accenture's Vision for Secure and Sustainable AI Innovation

Ever wondered how leading tech partnerships are shaping the future of Secure and Sustainable AI innovation?

In a recent exclusive interview at Intel Vision, Ramtin Davanlou, the CTO for Intel and Accenture Partnership, at Accenture, shared insights into “secure and sustainable AI”.

Here are the highlights of our discussion:

Accelerating Innovation Through Partnership

Ramtin Davanlou sheds light on the decade long partnership between Intel Corporation and Accenture. Their partnership aims to leverage the latest hardware and software solutions from Intel to optimize client workloads and applications. For instance, Accenture has successfully shifted AI workloads from expensive GPUs to Intel CPUs and provided more economical solution.  Additionally, they utilize Intel's Gaudi accelerator for fine-tuning and high-performance inference and further enhance the efficiency of AI applications.

Fine-Tuning Generative AI through Retrieval Augmented Generation ("RAG")

Fine-tuning plays a crucial role in enhancing the potential of generative AI, particularly in the context of retrieval augmented generation (RAG). Different techniques are employed based on specific use cases to customize model outputs. For example, when generating code for software engineers, technical language and deterministic outputs are prioritized, whereas interactions with patients in a hospital require a different approach.  Fine-tuning enables the adaptation of models to various scenarios and improves the quality of output by incorporating contextual data. By integrating organization-specific data with existing generic data, the model can provide hyper-personalized and contextually relevant responses and enhance the overall user experience.

Prioritizing Security and Trust in AI Applications

Accenture and Intel prioritize security and responsible AI in developing AI applications. They collaborate to integrate technologies like Intel Confidential Computing and ensure data protection during processing in the CPU. This approach safeguards against breaches and maintains data integrity and addresses the critical need for real-time security in AI applications across industries.

Optimizing cloud performance and reducing carbon footprints

Recognizing the compute intensity of AI workloads, Accenture advocate for a combination of compute architectures to optimize performance and cost-effectiveness. This approach involves extensive benchmarking and optimization efforts.  It culminates in the development of Switchboard.  This automated process not only improves energy efficiency and reduces costs but also enhances the quality of responses. Clients can dynamically choose architectures in real-time to maximize return on investment. This forward-thinking approach demonstrates a commitment to sustainability and efficiency in AI deployment.

My conversation with Ramtin gave me a new perspective on what we can achieve leveraging AI and its ecosystem. The partnership between Intel and Accenture exemplifies a commitment to driving innovation, security, and sustainability in the rapidly evolving landscape of AI and computing.

Subscribe to the #CXOSpice Youtube channel and watch the full interview here.

Disclaimer: This article is sponsored by Intel. The opinions expressed in this post are those of the author and do not necessarily reflect the views of BBN Times or Intel.