Advertisement
Articles
Advertisement

Cloud R&D – A New Model for Drastic Change

Tue, 02/26/2013 - 8:52am
Randal Kenworthy, consulting solutions director & Kevin Long, director, Cisco Systems

It should come as no surprise that today’s Chief Information Officers (CIOs) and Information Technology (IT) organizations are looking to leverage the new capabilities of cloud technology, as it is transforming the way enterprises manage their infrastructure. To gain a better understanding of how to leverage the cloud, there are a number of questions these executives need to answer, including:

  • Which type of cloud to deploy – Public, Private or Hybrid?[i]
  • How to deploy that type of cloud?
  • How to charge for it?
  • What applications provide the greatest value to this cloud?

The last question poses the biggest opportunity for a company to drastically change its business model. To many Fortune 500 corporations, they have already identified research and development (R&D) as the key area that proves to have the most value.

There are many reasons why R&D is the “sweet spot” for cloud computing. Regardless of the industry, there is always a lot of ramping up and down, short-term testing and modeling for any new developments. If a company uses the cloud during this process, they won’t only reduce development time and costs, they will also be able to more efficiently and securely get a product to market. Figure 1demonstrates the value chain. This article will focus on the first step, design and innovate, which is where R&D occurs.

(Figure 1)
 

Cloud R&D Use Case: Storage Industry Company

Since the storage industry is a very competitive market in which the speed and precision of creating new features and supporting existing ones can make or break an organization, there is a need for a new breed of solutions that enable companies to be more flexible, scalable and allow them to focus on production, rather than testing. The solution must be supported by senior leadership within a company in order to accelerate their speed of adoption, and, ultimately, return on investment.

Through cloud-based, high-end data-storage solutions, companies are finding that entire product development lifecycles are improving. EMC, a world leader in high-end data storage solutions that develops, delivers and supports the IT industry’s infrastructure and virtual infrastructure technologies and solutions. EMC has developed pre-configured Virtual Machines (VMs) and storage simulators that can deploy in minutes, reducing the lengthy time of deploying physical hardware systems and eliminating the need to have physical storage systems in use for testing. By utilizing development VMs and storage simulators, engineers can develop, build, and test without ever leaving the cloud. EMC connected code repositories were connected to the cloud via 10GB Ethernet. By doing this, they were able to achieve high performance access to development VMs. This allowed them to deploy infrastructure to remote sites without having to physically ship equipment or send lab staff to the sites, saving time and money.

The current technology initiative among many R&D organizations is a self-service, virtualized, automated, streamlined cloud utilized to enable thousands of developers, QA engineers, and programmers to perform their daily job functions in a centralized, self-managed environment. This enables corporations to scale out thousands of individual test and development environments while also offering a complex simulation environment.

Benefits and Opportunities

Like EMC found, cost savings with cloud R&D can be achieved in many areas, including hard-dollar capital and operating expenses, engineering and IT productivity and product time-to-market. EMC has projected more than a 30 percent reduction in IT hardware costs associated with deploying virtual development VMs instead of physical workstations. Savings can also be generated by using simulators in place of physical storage systems for testing.

In addition, company savings with cloud R&D comes in the form of reduction in time spent by engineers setting up test environments, which the companies have calculated a more than 20 percent decrease. Lab personnel are able to be more efficient by deploying systems into the cloud instead of having to physically rack and stack hardware. For instance, in 2010, it would take a typical R&D engineer 55 days to procure, install, configure and load a server environment for product testing. The average cost to do this was $5,000. In 2011, that same engineer could accomplish this process in 24 hours at less than half the cost.

It is critical to note that none of these benefits can be achieved without a certain level of support from senior leadership. A general consensus is that most massive deployments stall out at 20 percent when not driven by an executive sponsor. Companies must also share business cases for successful deployment and put a timeline in-place for adoption of new technology to ensure that migration occurs consistently and does not experience lag, as that will account for additional spending and less return on investment. In addition, companies must understand the need to be transparent about the cost of business with old technology and methodologies versus new to ensure employees are on-board with the shift and committed to positive change for the organization as a whole.

Cloud R&D Business Case: Large Software Vendor

In the case of a large software vendor, we found an environment with 800 software solutions being cared for by 800 development teams around the globe. The case for cloud was very strong in this environment. Development teams often work in cycles where they need to rapidly model or test variations, often in the many thousands, to validate designs, or regression test to ensure that new features don’t break old ones. This creates a business case where they need the maximum amount of computing resources available for a very short period of time. The servers we have modeled in this environment sit idle often 90 percent of the time.  For one firm, that was $100Ms in investment that sat unused.  What drives this? If a pooled environment cannot scale adequately to reasonably prevent competition among users for the shared resource they need, the user will often hoard resources for themselves.  At the firm developing 800 software products, IT leadership quipped that they had 800 miniature data centers that moved with the developers. Another software firm CIO commented that he had two data centers on site – one inside the traditional building, and the other spread out beneath the desks of his developers on campus.

Solutions that propose sharing or combining are often met with skepticism because developers may have experienced competition with other development teams. These developers don’t want to wait in line because they care about meeting deadlines and being as productive as possible. 

The cloud enables a new approach and offers a solution to the hoarding problem. The cloud can help aggregate all of the mostly-idle computing resources together, reducing the waste in equipment investment by 75 percent. This means that managers aren’t only handing developers one server, rather, they are giving them access to thousands of servers.  This large-scale consolidation within the company accelerates time-to-market for new products & feature updates and can easily by embraced as the go-to market strategy once the cloud is enabled.

Enabling the cloud for R&D at this particular software vendor has revolutionized how they have budgeted equipment. Instead of buying people their own servers, they are putting people on these platforms. No matter how old the gear is, there is a strong investment opportunity to invest in the cloud for R&D.

Summary

By adopting the cloud for R&D, companies are able to build new models to forecast the impact of changes in business plans, arm their CIO for faster responses and reactions, and help perform what-if models to consider the impact of shaving resources away from classes of applications to drive efficiency into operations. In addition, cost savings in the form of hard-dollar capital and operating expenses, engineering and IT productivity and product time-to-market is achievable with the right solutions in-place.

In order to successfully incorporate these solutions, companies must develop change management processes to help communicate and gain buy-in to leverage the use of cloud R&D.  Many IT industry engineers typically procure, install, manage, update and retire their own IT and server environments, which cost both time and money. Numerous companies have identified a need for a self-service, automated environment with the ability to monitor all metrics in real-time while allowing easy scalability beyond what can be handled today. The goal is to reduce IT costs while increasing product development by speeding-up the testing process.


Cisco’s initiative

Cisco, in collaboration with EMC, a world leader in high-end data storage solutions, has developed change management processes to help communicate and gain buy-in to leverage the use of cloud R&D.

 


[1] Public clouds are accessed over the Internet and commonly offer services anyone can use. They typically lack some business features.

Private clouds are either: multi-tenant, used by multiple business customers, but in a secure environment that maintains isolation between clients; or dedicated (purpose-built) clouds, sitting often on a client's network, behind the client's security perimeter.

Hybrid clouds are any combination of the above, but managed centrally.

 

Advertisement

Share this Story

X
You may login with either your assigned username or your e-mail address.
The password field is case sensitive.
Loading