With 1.4 Terabytes of processor memory an entire application and its data can be loaded into main processor memory.
The fastest processor for data analytics, and not too shabby on engineering, scientific and other workloads either, is the new Intel Xeon processor E7-8800/4800 v3 product families (Xeon X7v3), according to Intel Corp. (Santa Clara, Calif.) Available with up to 18 cores, the Xeon E7v3 boosts performance while cutting power, putting it ahead of its main competition, IBM’s Power8, according to Edward Goldman, chief technology officer in Intel’s Data Center Group.
“Over the last few years, the cost per server has dropped 40 percent while the market is growing to $17 billion,” Goldman told us. “To meet those customer expectations Intel’s latest high-end E7 version 3 processor improves performance by 25-to-40 percent, has the largest memory space available today — up to 1.4 terabytes per core — and runs on 60 percent less power than an IBM Power8 with 85 percent less total cost of ownership.”
Real-time business intelligence (BI) and analytics has become a top priority as the time-to-market has shrunk from years to months, and the flood of “big data” has created a deluge overwhelming the traditional data center. “White box” or low-end processors do fine for easy tasks like serving up web pages, but for heavy-duty analytic loads all the processor makers — led by Intel and IBM — are looking for ways to harness the biggest clusters of multicore processors they can, not just to keep up, but ideally to predict where the future of their businesses are going so they can plan to be there when it happens, rather than catch-up after the newest trend is.