News

Announcements

Press release

Product reviews

Awards

	March 30, 2009   |   September 16, 2008    |   August 26, 2008

Unprecedented performance. Real world results
New, Faster, More Powerful IBM System x3850 M2 and x3950 M2 models
For your business to perform, your business applications must perform. As the market leader in high-end x86 servers, IBM understands real world business performance. With the introduction of the IBM System x3950 M2, based on the Intel Xeon 7400 series processor, IBM offers the first x86-64 system to exceed 1 million tpmC in the industry standard TPC-C benchmark . In fact the system exceeds that mark by 20% and owns well over 100 other leadership benchmarks.

But real business performance demands more than just benchmarks. The x3950 M2 is designed to help your business succeed with:

• Flexibility, scalability from 4-socket up to 16 sockets. The x3850 M2 provides an easy upgrade path to allow you to “pay as you grow”
• Mainframe-like reliability with extensive memory protection features and other system availability features, like Predictive Failure Analysis® to help keep your systems running
• Designed for energy efficiency with high-efficiency power supplies, and memory technology that’s capable of up to 37% lower overall power consumption in high DIMM count implementations

Innovation for real business performance
The x3850 M2 and x3950 M2 are built on IBM eX4 technology, supporting the performance, reliability and scalability your critical business applications demand. With its efficient design, both models are ideal for virtualization and server consolidation. With just one server, you can support a large number of applications, with a higher threshold for protecting data. Combining memory scrubbing, IBM’s patented Chipkill™ and Memory ProteXion™ technologies, as well as memory mirroring, the system’s enhanced memory subsystems work together to minimize bottlenecks, promoting fast, efficient and balanced functionality for ERP, database and other demanding business applications

Operating at a lower wattage than competitive systems, the new servers offer power-efficiency features to help you save operational costs. For example, the DDR II memory used in the x3850M2 and x3950 M2 uses up to 37% less power than the Fully Buffered DIMMs used by the competition, with no loss in performance. IBM Systems Director Active Energy Manager™ helps monitor and cap power consumption to improve energy efficiency and help lower costs.

With the ability to scale up sixteen sockets — a maximum of 96 cores using Intel® six-core processors — the x3850 M2 and x3950 M2 are designed to offer easy, flexible expansion to meet a wide range of business application demands. For example, you can populate as few as two CPUs per chassis while still using all memory DIMMs. With six-core processors, this configuration allows for the creation of a 12-core, 32-DIMM server utilizing only two processor sockets. This flexibility offers extraordinary price-

performance balance for applications constrained by socket-based licensing. As demands grow, the system can be scaled to a 48-core, 128-DIMM server utilizing only eight sockets, or all the way up to 16 sockets, 96 cores, and 256 DIMMs (1TB) to meet the need for virtualization, databases, and other applications requiring massive memory capacity, many processors/cores, and tremendous I/O bandwidth and flexibility. Equally important, the x3850 M2 and x3950 M2 provide industry-leading reliability and availability features.

Optimized for reliability, availability and efficiency
The x3850 M2 and x3950 M2 are designed to offer extraordinary reliability and availability for your business. Time is money, and time offline is money lost. IBM understands the demands of business applications and offers innovative features to help you keep your server, and thus your business, running at peak efficiency:

• IBM Memory ProteXion: Designed to provide a deeper level of redundancy that corrects multiple single-chip errors to help keep the memory up and running longer without risking data integrity
• IBM Chipkill memory: Helps correct multiple, single-bit errors using off-the-shelf DIMMs
• Memory mirroring: With hot-swap and hot-add support, memory mirroring is designed to protect data through the ability to write simultaneously to independent redundant memory cards
• Advanced Buffer eXecution (ABX): Provides on-board technology to help resist chip failure and improve availability and reliability, while reducing power consumption and increasing performance
• Memory scrubbing: An automatic daily test of all system memory. It detects and reports memory errors that might be developing before they cause a server outage.
• Hot-add/hot-swap memory: Should a DIMM fail, a memory card can be swapped out without taking the server down. Likewise, memory can be added with the server online.
• Hot-add/hot-swap I/O: Up to 8 PCIe slots (two per chassis) are Active™ PCIe slots. This permits adding and removing adapters without taking the server down.
• Hot-add/hot-swap HDDs: All HDD bays support adding and removing drives with the server online.
• Hot-swap/redundant fans and IBM Calibrated Vectored Cooling™: The fans automatically adjust speeds in response to changing thermal requirements, depending on the “cooling zone,” redundancy, and internal temperatures. This helps to reduce the ambient noise, reduce the wear-and-tear on the fans and reduce the server power draw. If one fan fails, the others increase in speed to make up the diference until the hot-swap fan is replaced.
• Hot-swap/redundant power supplies: If one power supply fails, the other picks up the full load until the first is replaced.
• RAID support: The optional IBM ServeRAID-MR10k controller provides data redundancy via RAID-10/5/50/6/60.
• Predictive Failure Analysis: Reports out-of-range components before they fail, giving servicers time to acquire and replace the component before it can fail. For non-hot-swap components, this means the opportunity for an orderly shutdown, with no data loss. PFA-enabled components include processors, memory, PCIe slots, voltage regulator modules, power supplies, fans, and hard disk drives
• Light path diagnostics: A pop-out-drop panel identifies the type of component failure (memory, processor, etc.). Then when the system cover is opened, an LED identifies the specific component requiring servicing (DIMM 32, for example). Monitored components include processors, memory, power supplies, fans, HDDs, adapters, and voltage regulator modules (VRMs), plus system temperature.