Through a series of industry events, I spent some time digging deeper into the data center and cloud platform market. Mostly, at a technical/semiconductor and component level, but the deeper I went into the markets that drive growth, the more I realized how much I previously had underestimated the TAM growth ahead for companies which supply components to the data center and the cloud platforms themselves.
At this point, I think even the most bullish TAM growth forecasts may be conservative as the world may be moving faster than analyst forecasts are assuming. Most forecasts peg the data center market at a sizable $200 billion market by 2020. But nearly every company from Intel, Nvidia, Microsoft, Google, AMD, and Amazon are seeing faster than forecasted growth from the data center business each year. This is a signal of a snowball effect that doesn’t seem like it will slow down. The dependence on cloud platforms and thus data centers is growing so quickly thanks to a few trends like machine learning/AI and e-commerce to name a few. There are other markets to discuss, but in this analysis, I want to show how just these two markets alone are likely to drive faster than anticipated growth in data center and cloud.
One of my previous underestimations of cloud/data center TAM growth was not realizing the tremendous demand on the data center of e-commerce. Some companies do this vastly better than others, and they are ones who truly come at this problem with cloud-first expertise. Take Alibaba for example. When it comes to e-commerce scale, Alibaba may be at the top of the pyramid. To date, no company drives more e-commerce transactions in 24 hours than Alibaba on Singles Day. No company drives more e-commerce transactions than Alibaba does in their first hour of Singles Day. In their last Singles Day, 11/11/2018, Alibaba drove $10b in the first hour alone and set a new Singles Day record at $42 billion for the full 24 hour period.
To put that into contrast, Black Friday revenues in total are around ~$10 billion for the 24 hours. While those statistics are interesting, the observation that stands out to me is how Alibaba can generate about the total of US Black Friday 24-hour sales in just one hour and have no issues or delays in transactions. Where many US retailers had their websites go to a crawl or even go down during their online sales event. Even Apple has a record of trouble with their online store during iPhone pre-order days, and in all these cases the transaction volume is lower than what Alibaba does on Singles Day. It just goes to show how much demand there will be for data center/cloud platforms as e-commerce continues to grow and become a primary way consumers shop.
To this last point about the growth TAM. If e-commerce is growing faster than people believe, and it is, and we know that e-commerce drives tremendous demands on the data center, then consider that global e-commerce is still less than 20% of total retail sales on a yearly basis. In the US, e-commerce is around 10% of annual retail sales. In China, it is still just less than 50%, and in the UK it is about 18%. There is significant headroom to grow the data center and cloud platform TAM on the massive growth still to come from e-commerce alone. And like I said, it is happening faster than many forecasts assume.
Outside of e-commerce, machine learning and how that translates into AI is easily the biggest evergreen growth area for cloud platforms and data center needs. There is a machine learning/AI element to retail and e-commerce as well as retailers look to include more personalization and anticipation of purchases for the individual consumers, but this extends much more deeply into the relationship between device specific (edge processing) and cloud processing.
As the explosion of data being created by end users continues the demands on the cloud will only continue to increase. This explosion of data creation is enabled by the increase in capabilities of the devices we use day in and day out. For example, as Apple continues to increase the capabilities of their smartphones, by innovation at the silicon level for things like imaging/computer vision, AR and maybe VR, consuming more rich and interactive games, etc., the demand on backend data center and cloud platforms will only grow. Similarly, every smartphone and PC maker is increasing what is possible for software and services with innovations in hardware that grow the demand in data centers and cloud platforms exponentially.
Give developers more computing/graphics/computer vision capabilities, and they will create more rich and immersive software and services to take advantage of those capabilities. And give consumers the option to consume more of these rich experiences at engagement levels we can not yet predict.
The capabilities to create, and consume data and experiences is increasing faster than people assumed as the competition in hardware increases during a period of flat to declining smartphones. All of this will inevitably create even more demand and growth in data center and cloud platforms.
These two areas alone are enough to be extremely bullish on all companies who play in the data center and cloud markets and not to underestimate the growth ramp ahead which will be significant and likely larger than what is currently forecasted.