A Partnership Designed to Support High-Performance Analytics
Scott VanValkenburgh, Sr. Director Alliance Management, SAS
Scott VanValkenburgh, Sr. Director Alliance Management, SAS
High-performance analytics is increasingly critical to business success. It doesn’t work without an underlying architecture that can support and enhance the analytics to deliver value to the business. HP and SAS enjoy one of the longest hardware/software partnerships in the industry. In over 30 years of partnering, we have worked with more than 20,000 joint customers.
The partnership has been committed to collaborating well before either of us releases a new product. Rigorous planning involves HP providing us:
• An early view on product roadmaps to allow solutions to be tuned for specific new features to be included in HP Converged Infrastructure (CI) offerings.
• Access to HP lab personnel to resolve issues both from an end-user standpoint where SAS is the customer as well as when both companies are serving a mutual customer.
• Sizing assistance and collaboration on customer-specific issues allows a faster time to resolution.
Likewise, we ship over early versions of our latest software to help HP understand how it stresses the infrastructure and how HP can tune offerings, configure components, reduce equipment needs and collaborate with customers to resolve issues faster and with a lower total cost of ownership. The ultimate goal is to gain the most power with the smallest footprint. As new data sources proliferate and the advent of technologies such as Hadoop shift the economics of storing more data, optimizing the infrastructure for analytics has emerged as a crucial consideration for IT departments.
Why Working Together Matters The underlying work that both the companies are doing involves speed and precision. Organizations increasingly tell us they can’t wait days or weeks for information. If it takes months to build a model, then by the time it is deployed, market conditions or customer preferences could have shifted. This focus on providing near-real-time analytics allows organizations to, for instance, create personalized marketing messages, stop fraud before it happens, reduce error-prone processes, and automate manual ones.
In recent years, advances in grid computing and in-memory analytics have solved some of the issues of providing speed and precision without vast new hardware purchases. IT departments can now take advantage of reference architectures and performance metrics for in-memory solutions and grid offerings running on the latest servers and storage systems. The key is to get it all to come together efficiently. Cheaper storage isn’t cheap if configuring it costs too much or if it delays business users from getting the data they need.
A solid partnership between hardware and software organizations leads to smooth transitions toward these new technologies, lower cost and better performance. Here are a couple of examples from our partnership with HP.
Aiding Medical Research A national medical research group wanted to provide privacy-protected health information to researchers across the country so they could better study the connection between genetic differences and the way patients respond to medications, along with tracking outcomes far beyond traditional studies.
The organization’s platform could not manage the increased demand. It turned to HP for a federated data sharing platform and SAS for visual analytics that makes it easier and faster for researchers to do their work. They can explore data using drag‑and‑drop capability and quickly design interactive reports that can be easily distributed via web or mobile devices. In addition, the project is more cost-effective than the older, slower platform they were using.
Helping an Insurance Company Stay Competitive An insurance company with an aging hardware platform and strict cost constraints needed a platform that could help it transition to high-performance analytics. HP and SAS worked together to help the company’s IT organization meet its goals. The work involved moving 150 pieces of software and provisioning the operating system and software application along with the new data.
The net result: The insurer achieved double the performance, reduced licensing costs by 30 percent, and has satisfied users. Most importantly, it can bring new insurance products to market faster.
Spurring Cost-effective Growth When a healthcare services company entered a period of high growth, its IT infrastructure could not keep up. The organization was looking to upgrade quickly and efficiently. The company chose SAS® Grid Computing on HP Converged Infrastructure, a combination that has a fully tested reference architecture design. The HP 3PAR Storage system enabled better throughput and availability for their SAS environment, while their HP ProLiant server blades provided their IT department with the flexible infrastructure needed. The solution was deployed in weeks, increasing business agility and security.
The company’s vice president of data analytics says they are now able to run complex analytics faster and continuously without disrupting IT or the organization.
All of these examples suggest that a partnership needs to go beyond an announcement. Extensive testing and benchmarking, the creation of fully tested reference designs before the first customer signs on, and solutions that decrease cost, increase speed and allow for the rollout of more sophisticated analytics are hallmarks of a sound hardware/software solutions partnership.
