AWS Case Study: Mentor Graphics
About Mentor Graphics
Founded in 1981, Oregon-based Mentor Graphics, Inc. is one of the pioneers in electronic design automation (EDA) software. EDA software helps customers from the aerospace, consumer electronics, industrial and defense industries create electronic products faster and more cost-effectively. The company employs about 4,400 people around the world.
Mentor Graphics software powers some of the largest electronics design operations in the world, often requiring powerful computer hardware and special installations. Providing effective software evaluations and sales demos can be an expensive, time-consuming proposition.
For its most complex products, Mentor often sent hardware and engineers directly to the customer site to manage the entire evaluation. That proved extremely costly in terms of staffing, travel and resources.
Mentor also developed separate trial versions of some software products, with limited features and/or piracy-preventing restrictions, since providing the full product put Mentor’s intellectual property at risk. Doing so, however, required Mentor software developers to manage multiple versions of each new release.
The situation was not cost-effective for Mentor Graphics, and not ideal for its customers, either. In the defense and aerospace industries, for example, companies work on a contract basis and must be able to bill each hour to a specific project.
“Loading a new trial application was hard because customers had to do it on their own time,” says Ron Fuller, Mentor Graphics Web Manager. “Often they had to load it on their own machines, which weren’t powerful enough for the trial version. And you can forget putting trial software on their development workstation—that just wasn’t allowed. We needed a way to let our customers get hands-on experience with our tools without disrupting their current workflows.”
Fuller’s team began exploring ways to use the Amazon Web Services (AWS) Cloud to deliver software trials and demos. “Our AWS evaluation versions are built around the idea of showing a specific set of features so that you don’t have to download the full package,” Fuller says. “It was just what we needed.”
Why Amazon Web Services
The company uses AWS to provide its customers with Mentor Graphics Virtual Labs, an on-demand training and evaluation platform that makes it easier and less expensive for customers to evaluate products. “We’re delivering a better experience to our customers, thanks to AWS,” Fuller says. “We launch a customized instance in the cloud for customers, and they can take their time evaluating the software in their environment.”
The company employs a multiple-region strategy, using AWS resources in the US, European Union, and Asia-Pacific to serve customers and employees all over the world. This strategy helps increase the performance of virtual labs for worldwide customers while also providing fast upload times for Mentor Graphics’ globally distributed product development teams. “AWS has given us incredible agility,” Fuller says.
When application engineers (AE) are first building a virtual lab, they use Amazon Simple Storage Service (Amazon S3) to upload early versions of images to Mentor Graphics headquarters. AWS CloudFormation scripts build a working image so that the AEs can start configuring products. “Amazon CloudFormation made it easy for us to connect and provision AWS resources,” Fuller says.
To deploy a virtual lab globally, the company uses Amazon Machine Images (AMI) to cross-copy images into all 7 AWS regions. The image is updated automatically with any fixes made to the master.
“Before AWS, we had to do this using scripts, and it was almost impossible,” Fuller says. “And because of cross-copy, we can move 100 GB root volumes out to the regions almost instantly. That makes it easy for people who are on the road to get the latest version of the product.”
Sales engineers can set up a demo at company headquarters in Wilsonville, Oregon, and replicate it to all regions through virtual private gateways and the Amazon Virtual Private Clouds (Amazon VPC) that connect them. If the network experiences issues, the system fails over to the centrally located domain controller, licensing, and terminal server.
The company uses AWS CloudFormation scripts extensively to manage all the AWS resources in the Mentor Graphics stack. “Amazon CloudFormation builds every aspect of the evaluation,” Fuller says.
Mentor employs Amazon Route 53 to create a DNS entry for every customer, and attaches a volume of Amazon Elastic Block Store (Amazon EBS) to each virtual lab. The EBS volume stores every file the customer uses during the evaluation, so that customers have access to those files without having to keep a single instance open during the entire evaluation period. “It’s like having a huge USB drive,” Fuller says, “and it’s one of the most awesome things about EBS.”
“Thanks to AWS, the Mentor Graphics customer experience is now fast, fluid and simple,” Fuller says. “Using the company’s old solution, customers would download files needed to install the evaluation, set up licensing, and perform other tasks—a process that took up to 15 minutes. Using the AWS Cloud, the process now takes about 15 seconds.”
Internal customers are seeing benefits, as well. Developers who once had to wait 8 hours for a 3GB file to finish uploading can now do the same job in only 30 minutes.
The Virtual Labs system has significantly reduced Mentor’s software evaluation and demo costs. “Using AWS, we’ve seen a huge savings in time, logistics, patches, overrides, networks, variability in the customer’s environment—you name it,” Fuller says.
“Before AWS, for each on-site workshop or demo, we would have to provision 10-12 laptops and spend thousands in travel, shipping and personnel costs to install a classroom at the customer location. With the AWS Cloud, we spin up 12 8-hour instances for about $2 per instance-hour, and the customer gets the full software experience on any computer with a web browser. There’s no comparison.”
Frequent AWS feature releases allow the Mentor web team to continually improve virtual lab features and performance, helping them support the growing number of use cases the company is finding for this system. “AWS is so extensible, we’re able to continually push our services and tools to our internal customers at a rapid pace,” Fuller says.
“The AWS Cloud allows us to offer customer evaluations, training, and demos in ways that would have been cost-prohibitive before,” Fuller concludes.
To learn more about how AWS can help your data needs, visit our Big Data details page: http://aws.amazon.com/big-data/.