Tech Field Day: Six Feet Up CTO Weighs In on Efficient AI Implementation

Calvin Hendryx-Parker, Six Feet Up’s CTO and AWS Community Hero, was invited to participate in the latest Gestalt IT Tech Field Day Showcase: Reduce Your AI Platform TCO. The discussion centered around making AI applications more cost-efficient for businesses, as Accenture’s Ramtin Davanlou presented findings of the company’s partnership with Intel to the panel.

“We’re trying to give computers ears, eyes, hands and a brain to do things for us,” Ramtin said in his opening remarks.

Ramtin outlined Accenture’s work trying to make implementation of AI applications more cost-efficient in order to boost adoption across use cases. The company does this by matching different compute architectures with software optimizations for a given use case and adding cost to the equation.

As CTO of Six Feet Up, a Python and Cloud expert consulting company that helps organizations build apps faster, innovate with AI, simplify Big Data, and leverage Cloud technology, Calvin’s questions centered around deployment of Accenture’s open source solutions and how they could be leveraged.

Among other things, Calvin noted that Accenture’s offerings have not been supported in the latest versions of Python. He noted that while it is supported in Python 3.8, Python 3.9 was just released and Python 3.10 is not far behind.

“What is the support and timeline looking like for getting these releases so we can get the latest versions of some of these tools?” Calvin asked.

Ramtin responded that Accenture’s clients typically use older versions of Python. However, he said the company is looking to close that gap and bring tools to the market sooner.

In partnership with Intel, Accenture put together a study that showed upgrading to the latest processors and using an optimized machine learning (ML) library. Specifically, Intel’s latest generation Xeon processors along with Intel Extension for Scikit-learn, can be:

• 25 times more cost-effective for some AI workloads,
• nine times more cost-effective for AI training workloads; and
• eight times more cost-cost effective for AI interfacing workloads.

WATCH RAMTIN’S TECH FIELD DAY PRESENTATION: