What cloud PLM vendors can learn about AWS speed

What cloud PLM vendors can learn about AWS speed


Amazon Web Services is one of the most popular cloud platform these days. We can say that Amazon became de-facto standard of public cloud. Amazon cloud keeps growing and remains the leader these days according to Synergy Research Group. A year ago article by Stackdriver published information about popularity of different AWS services. According to the article these are top three popular AWS services – EBS (Elastic Block Store), EC2 (Elastic Compute Cloud) and S3 (Simple Storage Service).

Lots of startup companies and established software vendors discovered the power and elastic capabilities of Amazon. Cloud is not only about consumer web application and social networks. The cloud era is coming to enterprise software companies too. Salesforce.com, Workday, Netsuite is the only short list of companies in the enterprise cloud domain. CAD and PLM companies are also taking advantages of the power and popularity of Amazon public cloud. As PLM vendors cloud strategies are getting more mature, I can see a potential for CAD/PLM companies to switch their focus on platform development and not only focus on applications. While there are still lot of not answered questions about future of PLM PaaS, many software vendors are asking about what is the right cloud platform for them. Recent GigaOm article raised interesting discussion about cost of public cloud.

Over the weekend, my attention was caught by VentureBeat article – Amazon Web Services speeds can vary by up to 200X depending on region. Thanks Startupmoon blog for pointing on this publication. Takipi, company focused on debugging of servers, discovered huge difference in speed of AWS applications in different regions and shared this information. Here is my favorite passage:

What that ultimately means is that developers who don’t pay attention to cloud regions, as server debugging company Takipi discovered, can actually cause their apps to run 10 or even 200 times slower than necessary. And Amazon doesn’t disclose any of that data. “We noticed that in many cases there’s 10X (or even more) difference in the performance of services due to AWS/S3 regions and external APIs,” Takipi co-founder Iris Shoor told me via email. “It’s possible to make amazing optimizations just by changing the region,” says Shoor. “For example, choosing Oregon over CA for a company that serves the European market will cut the upload time by half.” One of the main problems when hunting down the cause of slow services, Takipi says, is API latency. And regionality has a huge impact there, even if you’re as geographically optimized as you can possibly be.

One of the advantages of cloud applications is global availability. It looks like optimization of cloud PLM platforms can become a next focus for software vendors. One of the challenges of previous PDM/PLM platforms was to insure global availability and performance level. Latency was an issue for PDM users trying to access CAD data located in European servers from China. It looks like, the problem is just migrating from one software layer to another and requires  new implementation approaches.

What is my conclusion? Cloud is not a silver bullet. Customers are moving from digesting of PLM cloud marketing towards analysis and technical investigation of different PLM cloud platforms. PLM vendors should take a note and focus on technological differentiation of their platforms. In coming PLM cloud competition, cost, performance and efficiency will become the most important factors influence future market dominant positions. Just my thoughts…

Best, Oleg

Picture courtesy of VentureBeat article.


Share This Post

  • The trouble here is net neutrality… or lack thereof. Depending on the entry / exit points of the users, the experience can and will vary wildly, and it won’t even be consistent. With recent rulings support a lack of net neutrality, the cloud proposition has a new set of challenges to overcome.

  • beyondplm

    Ed, you are right- cloud and internet is changing. It brings new requirements to cloud providers, since they will need to live in this world. What struck me is that core problems of efficient global deployment is probably remains the same – latency, network traffic, optimized deployment. What is your take?

  • The way I see it, the transition will be difficult because it’s a pushing a customer shift from a model of explicit definition and stability (private on-premise) to a model of variability and persistent unknowns. And there’s no amount of marketing that can provide the same assurance as a system that you completely control.

  • beyondplm

    Ed, what you are saying is right for large organizations. As much as we move towards smaller organizations and supply chain, it less relevant, in my view. For them, infrastructure cost pressure and absence of central IT is a reality.

  • Of course, totally agree with you, but most of the *current* customers are the big organizations… they are ones to struggle with the transition. For the *new* customers they are expecting cloud, but probably struggle more with sizing up PLM itself.

  • beyondplm

    I think new customers are excited about potential better balance of cost/value. For many of them, subscription is a better proposal from TCO standpoint. As they learn more and start using that, they will be able to formulate their expectation about speed and availability. Most of them today think “cloud = google”.