IBM Flex System x280 X6, x480 X6 and x880 X6 Compute Nodes

Highlights

Today’s IT organizations must deliver game-changing cloud, big data and analytics capabilities. But they’re also expected to drive innovation with declining budgets. In addition, mission-critical applications are required to do more as businesses increase mobile and cloud deployments. Despite these challenges, IT solutions must be able to scale performance, manage large data and reliably make information available in real-time.

The new IBM Flex System X6 family of compute nodes incorporates the sixth generation of IBM enterprise X-Architecture® to help deliver better, more efficient business results. X6 platforms are fast, agile and resilient so they can produce significantly faster compute performance than previous-generation systems. The X6 portfolio increases virtualization density, decreases infrastructure costs and complexity, and enables analytics engines to process faster, allowing you to rein in IT sprawl.

The new IBM Flex System X6 Compute Node family consists of three servers from the Flex System portfolio. The Flex System x280 X6 is a two-socket scale-out server optimized for workloads that require large memory footprint and virtualization performance. The Flex System x480 X6 is scalable to a 4-socket system optimized for applications that require more processor cores and memory for larger virtual machines and databases. The Flex System x880 X6, the first scalable 8-socket Intel Xeon processor-based blade server, is ideal for workloads that demand the highest levels of capacity, performance and resiliency, such as analytics and large databases.

x280 X6 Product features

Hardware summary

x480 X6 Product features

Hardware summary

x880 X6 Product features

Hardware summary

1 100 and 300 percent performance improvement is based on preliminary projected results of SPECint*_rate_base2006, SPECfp*_rate_base2006, plus performance gains from eXFlash DIMM storage. SPEC benchmark results will be available at www.spec.org after 5/13/14. 100 percent configuration: x480 X6 Compute Node using four Intel Xeon processor E7-4890 v2 vs. 4-socket server using four previous top-of-the-line E7-4870 (v1) processors. 300 percent configuration: x880 X6 Compute Node using eight Intel Xeon processor E7-4890 v2 vs. 4-socket server using four previous top-of-the-line E7-4870 (v1) processors. Also, a performance increase is expected for the x280 X6 based on preliminary testing.
2 Based on information obtained from www.cisco.com, www.dell.com, www.hp.com, and www.ibm.com, as of April 4, 2014.
3 Up to 12 TB maximum memory vs. 6 TB or less for competitive Intel Xeon processor-based blade servers.
‡ IBM intends to offer eXFlash memory-channel storage for the new IBM Flex System X6 Compute Node Portfolio. This new offering will enable clients to increase database performance, realize higher levels of storage consolidation, and scale virtualization on a single server.
1 100 and 300 percent performance improvement is based on preliminary projected results of SPECint*_rate_base2006, SPECfp*_rate_base2006, plus performance gains from eXFlash DIMM storage. SPEC benchmark results will be available at www.spec.org after 5/13/14. 100 percent configuration: x480 X6 Compute Node using four Intel Xeon processor E7-4890 v2 vs. 4-socket server using four previous top-of-the-line E7-4870 (v1) processors. 300 percent configuration: x880 X6 Compute Node using eight Intel Xeon processor E7-4890 v2 vs. 4-socket server using four previous top-of-the-line E7-4870 (v1) processors. Also, a performance increase is expected for the x280 X6 based on preliminary testing.
2 Based on information obtained from www.cisco.com, www.dell.com, www.hp.com, and www.ibm.com, as of April 4, 2014.
3 Up to 12 TB maximum memory vs. 6 TB or less for competitive Intel Xeon processor-based blade servers.
‡ IBM intends to offer eXFlash memory-channel storage for the new IBM Flex System X6 Compute Node Portfolio. This new offering will enable clients to increase database performance, realize higher levels of storage consolidation, and scale virtualization on a single server.