What does the chip shortage mean for cloud computing?

It’s hard to get products that require processing chips—from trucks to toasters. Compute and storage servers are pinched, too, but the big providers have protections the little guys don't.

What does the chip shortage mean for cloud computing?
Colin Behrens (CC0)

The complex problems behind the current worldwide chip shortage boil down to two basic factors: The pandemic limited chip manufacturing, and demand now outpaces supply.

It will take time for the chip supply to normalize and the effects of the shortage to correct throughout the manufacturing and distribution processes of technology products. Unfortunately, it won’t happen at the speed most experts predict or promise. 

Most of the questions I get from technology reporters and analysts these days relate to the chip shortage, specifically the effect of the chip shortage on the cloud computing market. Here are brief summaries of my observations. 

First, the shortage impacts traditional enterprise data centers more than cloud providers. The good news for cloud providers, or those who use cloud providers, is that they are less sensitive to chip price and availability issues compared to private data center owners. Here’s why:

Cloud providers do a much better job of sharing chip-based resources, given that they leverage virtualized and multitenant systems. The typical data center won’t be as efficient at sharing chip-based resources, no matter if they are virtualized or not. 

Cloud providers can keep prices lower per processing cycle because they take a much longer-term look at pricing and its effects. It’s to their advantage to keep usage prices low since the number of customers they acquire translates directly into long-term reoccurring income. For the standard data center, it’s just sunk costs that will not be fully utilized for many years.

Second, cloud service providers now drive more innovations of chips used in cloud computing systems. In case you didn’t notice, cloud providers are inventing, producing, and leveraging their own chipsets. Because many large cloud providers now control all steps of the chip development process and the chips are optimized for their specific requirements, these providers no longer rely on the chip producers for their innovations or their chip costs and power optimizations. This DIY approach means the chip shortage will have a minimal effect on the large cloud providers’ products and services.

For these DIY providers and their customers, the current shortage will most likely result in more innovation and more provider control of other components that make up cloud services, such as networking equipment, power optimization, and even the power itself that can be supplied using renewable sources such as wind or solar. Most private data center owners can’t profitably operate and innovate at the same levels as the larger cloud providers. 

It’s not that I’m shilling for the big-name cloud computing providers, but they do have the upper hand when you consider the chip shortage from all angles. Yes, they will feel some of the effects of the technology shortages, but the current crunch is much less likely to affect their operations or prices. 

If anything, scarcity will continue to drive independent innovation among cloud providers and others. Many enterprises that aren’t strictly cloud computing–related are already taking more control of in-house chip development and manufacturing within their market segments. If these manufacturer-driven DIY chip ideas and innovations come to fruition, today’s chip-related inventory shortages for everything from RVs and refrigerators to smartphones and wearable Internet of Things devices will become a problem of the past.

Copyright © 2021 IDG Communications, Inc.