Lenovo’s newest #1 TPC-H benchmark highlights the need for speed when it comes to uncovering crucial business insight

As children, we played “Connect the Dots.” Armed with giant crayons, we waited anxiously for the click of the stopwatch. Then we raced to connect one dot with the next and the next until someone was first to reveal a recognizable picture. The fastest young artist received a prize.

When I think of Lenovo’s latest #1 TPC-H non-clustered 10,000 GB benchmark result recorded on Lenovo System x3950 X6* as of April 6, I think of those games of Connect the Dots.

The TPC-H benchmark measures “business oriented ad-hoc queries and concurrent data modifications.” ** Burleson Consulting explains what that means, stating that users “make repeated requests to the online database, with one query answer stimulating another query. Because the purpose of ad hoc query is to allow free-form queries to decision information, response time is critical.”*** With Lenovo servers, business leaders can get the kinds of lightning fast responses they need. The Lenovo System x X6 server family provides a highly-scalable stable platform for virtualization, database, CRM/ERP/business logic and analytics environments.

Think of ad hoc business queries as dots. The company that can most quickly connect those dots to reveal crucial business insights stands the best chance of winning -- gaining competitive advantage, or redirecting business strategies to match changing trends, or resolving the kinds of operational problems that can lead to lost profits.

Let’s contemplate some hypothetical examples.

In a manufacturing environment, a business leader runs a series of ad hoc queries. The first query involves one particular piece of equipment that is standard on all of the company’s sixty manufacturing lines. The man is anxious to know the total number of sets of ball bearings that have been replaced in the last two years in this kind of machine. That query leads the man to wonder about the number of times that manufacturing lines have come to unscheduled stops as a result of these ball bearing failures, so another query is run. That answer leads to the question, “In each of the sixty individual manufacturing plants, is the number of line shut downs due to this problem roughly equivalent? That query reveals the fact that two plants show significantly lower shut down rates caused by this problem than all of the others.

Perplexed, the manager asks his colleagues what might account for the lower failure rates in the two plants. One coworker remembers a pilot program in which some of the plants have integrated infrared security technology with the manufacturing lines. That technology can be used to detect a rise in equipment temperature that signifies failing bearings before they actual cause the machinery to shut down. This enables replacement of the parts during scheduled maintenance windows – before they fail.

Based on this new finding, the man runs a query to ascertain which plants have implemented the new integrated security solutions. He determines that only the two plants in question have installed the new infrared solutions. Because unscheduled manufacturing line shut downs are extremely costly, especially when multiplied across an enterprise, he uses this data to instigate a study to determine if installing integrated security solutions in all of the corporations’ manufacturing lines might significantly decrease profit losses due to unscheduled line stops.

Another example involves a business professional at an insurance company. He runs an ad hoc query to see how many specialized medical procedures of a certain type have been submitted for insurance reimbursement in a particular city. Then he runs another query to compare that number with a city that is 3000 miles away. Two more queries of the same sort follow comparing those two answers with results in two more smaller cities that lie somewhere in between. Three of the figures fall within the same general range based on the respective populations. However, the fourth is exceedingly large by comparison.

Armed with that information, the individual runs another query, this time, associating healthcare providers in the outlier city with the number of times they have performed this type of specialized procedure. Interestingly, this last query reveals one particular provider who seems to have performed an inordinately-high number of these operations as compared to other physicians working in similar hospital environments in that city. He opens an investigation.to determine the reasons for the outlier.

An imaginary global retail enterprise can serve as third example. Let’s say a movie star wore a particularly lovely peacock blue gown to an awards ceremony two weeks ago. Several ready-to-wear designers instantly produced a very similar gown in royal blue and also in black and made it available to this major global retailer. The retailer immediately displayed the dress in stores and the marketing team determined that sales are skyrocketing. An ad hoc query is initiated to determine the best-selling sizes. (Some team members have speculated that size 8 and size 10 dresses in black must have been most popular and have likely accounted for the bulk of the revenue thus far.) The queries reveal that their guesses are incorrect.

With ad hoc queries about stores in different geographical locations, the group determines that many stores in Spain have sold out of size 4 in blue. Stores in another European city have sold out of all size 8, 10 and 12 dresses in black. Another query reveals that a specific region in the US has sold out of only size 12 gowns in both colors. Stores in yet another location no longer have size 2 blue gowns. Quickly determining that popular sizes are tied to specific regions and particular demographics enables the team to appropriately replenish stock in all of locations and continue the record sales of these dresses while the trend is prevalent.

Connecting the dots: In each of these examples, speed is crucial. Getting the answers to critical business questions and taking necessary actions immediately can make the difference between industry leadership and failure. Look to Lenovo’s X6 family of enterprise servers for speed, exceptional availability, flexibility, scalability and system stability.



To learn more about Lenovo servers, consult your Lenovo sales representative, or visit http://www.lenovo.com

Read San Disk’s blog entitled “The Industry’s Fastest: Lenovo and Fusion ioMemory Eliminate Performance Barriers”




*Benchmark results current as of 4/23/15, http://www.tpc.org/3312.

**TPC Benchmark, TPC-C, TPC-E, TPC-H, TPC-DS and TPC-VMS are trademarks of the Transaction Processing Performance Council.

**DSS and Expert Systems, Burleson Consulting, http://www.dba-oracle.com/tp_DSS_ad_hoc_queries.htm, (April 22,2015).

Original artwork by Kathy Holoman