Cloud Computing and SaaS (Software-as-a-Service) have been at the heart of the vision of the Xuropa Platform from the very beginning…on the proverbial napkin back in May, 2007. It’s current incarnation is captured with our Online Lab facility, and our development progress is metered by the adoption rate of key concepts by the electronic design industry.
This first post in a series of 4 was spurred by Daya Nadamuni’s article in EETimes.com and many conversations with a huge proponent of Cloud Computing and SaaS for EDA, Harry Gries (aka The ASIC Guy). I thought it was time that I took a crack at articulating how I see this technology evolution and it’s application and adoption within the electronic design industry.
As with any “next big thing”, the definition of Cloud Computing will evolve over time as it’s better understood and the software industry moves the ball forward. However, it’s important to note what Cloud Computing is not. As we saw within EDA and the stretching and pirating of the definition of “ESL” in the early days, it’s important to have a common understanding.
Wikipedia offers a pretty good and comprehensive description and definition of Cloud Computing.
I’ve paraphrased points here for convenience.
What it is not:
- Grid Computing: “The application of several computers to a single problem at the same time.”
- Autonomic Computing: “computer systems capable of self-management, to overcome the rapidly growing complexity of computing systems management, and to reduce the barrier that complexity poses to further growth”
It’s a little tricky, because as is noted in the Wikipedia entry, “many cloud computing deployments are today powered by grids, have autonomic characteristics and are billed like utilities, but cloud computing can be seen as a natural next step from the grid-utility model.”
Cloud Computing Components
I’m a huge fan of models like the OSI Seven Layer Model, and the Wikipedia editors of the Cloud Computing entry have taken a shot at capturing a hierarchy of its components. I’m not sure I completely agree with this model yet, but it’s a good start.
Client: self explanatory. The combined device and browser system.
Application: Software applications that are accessed via the web and leverage cloud infrastructure. Peer-to-peer (Skype); web-based application (Facebook, Xuropa (networking, news, and other applications) ); SaaS (SalesForce, Xuropa (Online Suite, Booth, and Kiosk); software+services (Microsoft Online Services)
Platform: A software application delivery environment. A key differentiator between a Platform and an Application is the degree of openness. In the case of Facebook Apps, the API is open and is written to by third parties who then have their application delivered to users within the Facebook environment. In the case of a Xuropa Online Lab it does not make sense to create an API for software vendors to write their tools to. And so we took the “API” down to the logical level – the Operating System.
Storage: Remote storage that is made accessible via the web and charged as a utility. Pay as you go/use. Eg. Amazon S3 (Simple Storage Service)
Infrastructure: Essentially the hardware resources and utilities to manage and make upper layers of the stack available via the web. Eg. Amazon EC2 (Elastic Compute Cloud).
Another example may be illustrative:
A database accessible via the web can be presented in three different ways to users.
- Raw database that is accessible via the web: Application
- Database with a UI (eg. Salesforce): Application
- Database with a macro API that can be written to and the resultant applet/macro/widget hosted and accessed by other users of the database: Platform
Adoption By Electronic Design: Where are we?
As can been seen from reviewing the above model, aspects of Cloud Computing are being used by the electronic design industry today: webex, Skype, salesforce.com, etc. But these are support services in the context of the EDA industry.
The rest of this series discusses the use of Cloud Computing and SaaS at the heart of the EDA business: delivery and monetization of electronic design software applications via the web.
99% of people in the industry to whom I mention Cloud Computing state with great conviction that it will never happen - “it’s been tried before”. And if they’re anticipating a step-function adoption that will stretch across the entire flow and be able to serve mainstream semiconductor vendors from the start, then they’re probably right. But that’s never the way things happen.
[“All truth passes through three stages. First, it is ridiculed. Second, it is violently opposed. Third, it is accepted as being self-evident.” -- Arthur Schopenhauer]
A common theme that recurs across these conversations is “trust”. And as with any new dislocating technology, this is a major hurdle to mainstream adoption.
So, regardless of all the buzz out there, this is where I think we are today:
So far, not very far.
The next post in this series will address The Drivers of Cloud Computing Adoption. What industry components make cloud computing and SaaS attractive, and what would potentially slow adoption down.
Part three of the series presents a potential roadmap of cloud computing adoption for electronic design.
Connect with me:
This post was written by James Colgan on December 9, 2008