diaxion dialogue
Diaxion Dialogue banner

Formula 1 Engines and the Cloud


That may have got you thinking – are we going to find out the secrets to Mercedes, Ferrari, Renault and Honda engines? Are they using the cloud to test their designs? Are they using a supercomputer in the cloud to model their engine before it is built? Unfortunately not but, what I want to show is the increasing pace of development and the ‘arms’ war that is happening in the hyperscale cloud providers. Finally my thoughts on where it is all going.

In the long distant past (say 7/8 years ago – a long time in internet time) about the time AWS came in to being in the form we could recognise today, they were like Lotus in the sixties with Colin Chapman. Colin decided to go against all the current thinking (large front engined cars). He introduced a smaller engine that was a stressed member of the monocoque (i.e. was part of the chassis and was behind the driver) as opposed to the standard way of a space frame with the engine sitting in the frame. This proved a break-through and forced everyone to move to similar designs. Similarly we have a number of incumbent virtualisation technologies (some very successful) being impacted by AWS business model and knowing they have to change.

As we move on through time we reach the ‘Formula Cosworth’ era (early to late 70’s), the time when the majority of cars were using the Ford Cosworth V8 engine with a few significant deniers and the era of the ‘garagistes’. This represents the first stage of cloud computing (2006-2010) and while AWS was the largest, there were large numbers of players (the garagistes) e.g. Rackspace, Centurylink, Saavis, Qwest, GoGrid, Softlayer, Google, Redis, etc. All were providing some form of IaaS but at certain price points. IaaS was the product (anything else was minor), mainly because this was what customers wanted (and could understand) and was easiest for them to provide, generally based on existing virtualisation technology.

Moving further forward, turbo technology comes in and increasing diversity becomes apparent in engines and the number of manufacturers goes up but the number of teams decreases. This was effectively caused by the high cost in engine development and the desire for manufacturers to be running the team with their name on the ‘brand’. Our parallel in cloud is the emergence of major PaaS and SaaS providers (thereby taking a different route) in the period 2008-2012. The key players in this being Microsoft Azure and Salesforce. This, to me, marks the real start of the ‘arms’ race in cloud. We have a number of different manufacturers taking different approaches providing choice to customers and at the same time fuelling innovation and competition. What is starting to happen is the development of additional features and functions from a base IaaS to additional IaaS capability such as memory optimised / cpu optimised etc and newer products like PaaS supporting languages such as Ruby on Rails, .Net CLR, Python. It seemed at this time that almost anyone could put together some infrastructure with a public interface and call it cloud.

The next phase of development would be 90’s to early 2000’s when there was a period of relative stability in engine capability (due to FIA rules), but high cost of running campaigns and designing cars led to a consolidation of teams around the manufacturers and teams aligned to manufacturers. For cloud this started with the emergence of basic standards or the semblance of them, in the form of OpenStack and the number of providers that created their own extension of this ‘standard’. Azure and AWS wreath notable exceptions due the need to differentiate themselves. They also had the advantage of being much larger than anyone else and had a significant customer base that had already bought into their ecosystem. Both have driven the pace of innovation both from a price perspective and feature perspective. For the other vendors, they experienced considerable consolidation with a number of the larger (also ran) vendors being taken over e.g. Saavis, Softlayer, Heroku etc, many were bought by Telco’s.

As we move into the 2010’s we see a further consolidation of teams and the domination of 1 team – Mercedes and resurgence of another – Ferrari. What changed were the rules and the requirement to use ‘hybrid’ (and I use the term loosely) engines (power units). In the cloud we see the dominance of AWS in the market but still having to react to price cuts by Google and Azure. Both AWS and Azure are engaging in a feature war rather than a price war with leapfrogging announcements occurring almost monthly if not more frequently. AWS is moving ‘up the stack’ and adding more PaaS like features that Azure has while continually expanding the existing feature set (as they already had a set of higher level capabilities). The majority of the rest have either focused on niche markets in verticals (like insurance claim processing) or horizontals (like Ruby on Rails apps). Others have focused on the enterprise /government market e.g. CSC, HP and IBM (in reality). However AWS cannot rest on its laurels when new technology looks to disrupt the disrupter. I am talking about Docker and similar technologies.

The following chart show the features release each year by AWS and the speed at which it is climbing. AWS have stated that this year they will exceed 500 and are currently on track to meet the goal. Azure do not break out their ‘features’ in the same way but the pace of their growth is similar based on blog announcements, though probably a number ½ the size but this may be the way Microsoft describe features.

So where is this going? I foresee a number of scenarios:

  • AWS and Azure get bigger and HP, IBM, CSC etc retreat to the enterprise market only, providing secure, optimised, private clouds only to enterprise – highly likely in 2-3 years
  • Container technology makes a number of services that Azure and AWS provide obsolete and reduces their importance – moderately likely in 5 years
  • Google plays a wild card and develops something that everybody wants in the PaaS/SaaS market – low likelihood in 3-5 years
  • BI/Data warehousing capability in the cloud becomes the norm and initially there are many players (often using Azure or AWS underneath) but consolidation occurs – Highly likely in 5 years
  • SAP tries to dominate the BI cloud market but ultimately fails – moderately likely in 5 years


May 2015

Our Blog
Click here to read our latest posts  

Diamond Support

– Capacity & performance management

– Mentoring / knowledge transfer

– Resource Management

– Industry best practice alignment

– Quality Assurance review

Click here to read more  

Questions and queries

We welcome any commentary, questions or queries regarding the above articles or any issues you think Diaxion may be able to assist with.

Click here to contact us  


Contact Us

Sydney 02 9043 9200
Melbourne 03 9036 8900

Return to top

Home | About Us | IT Strategy and Architecture | Data Centre Infrastructure | Case Studies | Diamond Support | Blog | Contact US

Copyright © 2015 Diaxion Pty Ltd Sydney, Australia. All rights reserved.