From: acquiredfm
The “excess capacity narrative” is one of four distinct origin stories for AWS, and it is ironically the most widely believed by the public, despite being untrue [00:00:00]. This story suggests that Amazon leveraged its idle server capacity to create AWS [00:00:09].
The Excess Capacity Narrative Explained
The myth posits that around 2001-2003, Amazon.com’s retail business, like other retail operations, experienced significant seasonal spikes in traffic and demand during Q4 (the holiday shopping season) [00:00:41]. For the first five years of the business, a rule existed prohibiting new code commits to production in November and December to maintain stability during this peak [00:01:06]. Additionally, the executive team and engineers would work in warehouses during Q4 [00:01:29].
The urban legend claims that because Amazon had to build out its technical infrastructure for peak Q4 demand, it had excess capacity during Q1-Q3 [00:02:04]. The “brilliant realization” was to rent out this unused capacity to other developers, thereby transforming a major expense into a revenue stream [00:02:25].
Why the Myth is Untrue
The excess capacity narrative falls apart for several reasons:
Technical Impossibility in a Pre-Cloud Era
In the pre-cloud technology landscape, it was not feasible to simply rent out unused servers [00:02:42].
- Servers were highly customized and tightly coupled to specific applications [00:02:53].
- Security measures for multi-tenancy did not exist, nor did network hardware understand how to serve multiple distinct tenants [00:03:04].
- Amazon.com’s codebase was literally installed on physical boxes they owned, making it impossible to simply “rent out capacity” [00:03:32].
- The servers were DEC Alpha servers, high-margin, monolithic platforms leased from manufacturers like Digital Equipment Corporation (DEC), following a business model similar to IBM or Oracle [00:03:41].
The Impact of Linux and Open Source
The advent of Linux and open-source operating systems changed the infrastructure landscape by allowing companies to run applications on commodity hardware rather than expensive, bundled platforms [00:04:11]. This shift enabled Amazon to undertake a massive rewrite of Amazon.com to run on Linux and HP servers around 2000, which significantly reduced costs and saved the company from potential collapse during a period of financial strain [00:04:50]. However, this move to open source was about cost savings, not about creating virtualized cloud servers for rent [00:05:16].
Business and Logistical Impossibility
If AWS were based on excess capacity, Amazon would be unable to serve its AWS customers during their own Q4 peak [00:05:26]. For example, a company like Netflix, dependent on AWS for streaming, could not be told they couldn’t operate during Q4 because Amazon needed the servers for its retail business [00:05:34]. Such a model would be inherently unreliable for any serious customer [00:05:40].
Official Dispelling of the Myth
Werner Vogels, who was the AWS CTO and is now the CTO of all of Amazon, explicitly debunked this myth in a Quora post in 2011:
[00:06:17] “The excess capacity story is a myth. It was never a matter of selling excess capacity. Actually, within two months after launch, AWS would have already burned through the excess Amazon.com capacity. Amazon Web Services was always considered a business by itself with the expectation that it could even grow as big as the Amazon.com retail operation.”
Intentional Strategy, Not Accidental Discovery
The excess capacity myth undervalues Amazon’s deliberate strategy and intentionality [00:06:49]. Amazon never viewed technology as a mere cost center; rather, they considered themselves a technology company and saw technology as a continuous investment [00:07:01]. The creation of AWS was an “incredibly intentional strategy” focused on an emerging market that Amazon believed it could create and dominate [00:07:24].