AVOA Snapshot: What does the partnership between Qualcomm and Amazon Web Services mean for enterprises and CIOs?

Welcome to the first AVOA Snapshot! Snapshots are quick reactions to news.

Today, Qualcomm announced a ‘multi-generational’ product collaboration partnership between Qualcomm and Amazon Web Services (AWS) worth billions of dollars and includes several tidbits enterprises and CIOs will benefit from.

Let’s jump into the Signal, the Noise versus the Reality and the CIO Directive…

The Signal

Qualcomm issued stock warrants to AWS to purchase up to a $4 billion equity stake (25 million shares) in the company. The agreement includes commercial requirements and AWS’ purchase of up to $60 billion of Qualcomm products and services. The term of the warrants spans 10 years to 2036.

The Noise versus the Reality

The noise is that this is just another partnership between a cloud/ AI provider and a chip manufacturer giving AWS access to more processor options. The reality is that it opens the door for greater options in both processor capabilities along with interconnects and an additional supply chain.

Qualcomm’s experience in the mobile space has given them the edge when it comes to power efficiency. In today’s world where AI chipsets are driving rack power consumption into the 50-100+kW range, power efficiency is becoming more critical.

High-bandwidth, low-latency interconnects between processing clusters is a significant hurdle for AI workloads that Qualcomm has addressed with their optical DSP and high-speed SerDes IP technology.

The partnership with AWS gives Qualcomm some serious street cred in the data center space, something they’ve been eyeing for some time. The partnership also opens the door for AWS to gain more direct access to another chip-manufacturing supply chain including custom silicon.

The CIO Directive

With so much concern around AI pricing, CIOs are gunning for lower-cost, flexible frameworks that give them the ability to bob and weave as demand shifts. The combination of AWS’ existing Graviton, Inferentia and Trainium processors along with Qualcomm’s Dragonfly CPUs and AI accelerators gives CIOs several options at their disposal as they look for alternative approaches.