February 2011: It is important to separate the facts from marketing hyperbole and misleading comparisons. Our competition’s marketing machine has been in full gear lately making bold claims about our servers. We are confident most customers will see these tactics as overly aggressive and not very factual marketing but we welcome the opportunity to set the record straight.

Please consider the following facts:

Fact 1: HP Superdome 2 provides a clear investment path forward for customers with multiple future generations of processors publicly defined by Intel®, all supported in a modular chassis with built-in advanced mission critical resiliency.

Purchasing large enterprise class servers is a major investment and customers expect to be able to leverage that investment over many years with continual processor upgrades. The HP roadmap for Superdome 2 shows a clear public roadmap of future processor upgrades that will all be supported in this new modular chassis. The M-series platform roadmap for any future upgrades is not nearly so visible and clearly defined.

Fact 2: For high-end servers, mission critical availability is key and delivering uninterrupted services to customers is critical. The M8000/M9000 relies on a combination of hardware redundancy and rebooting to recover from key hardware faults1.

By their very nature enterprise customers demand robustness and resiliency for mission critical workloads. The cost of downtime continues to grow putting increased pressure on mission critical environments. HP Integrity servers with HP-UX deliver a best-in-class mission critical infrastructure that enterprise customers have relied on for many years. With Superdome 2, HP engineers have elevated system resiliency further with a system design that is focused on fully insulating customers from that inevitable fault that will occur:

  • Increased reliability with power up once: New redundant and fault tolerant features provide a more robust mission critical environment
    • Fault tolerant fabric using redundant paths, link level retry, link width reduction, and end-to-end retry, is shared by multiple partitions avoiding single points of failure within a system or across multiple systems.
    • Automatic rerouting of data on redundant paths when persistent link failures are detected
    • Redundant clocks with automatic fail over
  • Self-healing capabilities implemented in the system level firmware (via the Superdome Onboard Administrator) protect the system — in conjunction with, but independent of, the operating system.
 

Fact 3: Compared to a comparable HP Integrity server, the M8000/M9000 SPARC Enterprise Servers require more data center real estate and provide less capability to manage power and cooling requirements in the datacenter2.

Superdome 2 incorporates a progressive power supply control scheme where supply efficiency varies depending on load; this control scheme strives to keep the optimal number of supplies in operation that yield the highest overall efficiency. Control of the fans themselves is a significant advancement over prior generation Superdome. Not only are the fans controlled to compensate for load and environmental changes, the algorithm used are a non-linear mapping of fan speed to temperature such that the airflow more closely matches the demand of the system. Energy costs in the datacenter represent an increasingly challenge for customers. Experts estimate that for every $1.00(USD) spent on new hardware, an additional $0.50(USD) is spent on power and cooling3. And the infrastructure needed to support a tier IV datacenter is approximately $25,000/kW4. Therefore, energy efficiency should be a critical element in selecting your infrastructure provider.

Fact 4: On a core-to-core, or processor-to-processor or system-to-system basis, HP Integrity systems can deliver better performance when compared to Oracle M-series SPARC Enterprise servers:

Results as of February, 22, 2011. See: www.spec.org SPEC, SPECint and SPECfp are registered trademarks of the Standard Performance Evaluation Corporation.

Fact 5: Vendor support for mission critical systems is a key — Oracle’s support quality is being questioned in the market.

“Interviews with almost a dozen end users and consultants, along with recent user surveys, have found that Oracle Support isn’t what it used to be. Service requests languish for days — or sometimes forever — unless there’s a follow-up call or you’ve got inside connections with Oracle. In other cases, users say Oracle Support will respond, but the advice isn’t helpful”

Source: “Has the phrase Oracle Support become an oxymoron?” February 7, 2011

“42% of Oracle customers are dissatisfied with the quality of the company’s support”

Source: “Are Your Customers Happy with Oracle” December 28, 2010

Fact 6: The Sun component of Oracle has struggled to compete in the high-end server market and lost significant market share in the past years5.

For more see: http://h30507.www3.hp.com/t5/Mission-Critical-Computing-Blog/Whose-clunker-do-you-want-cash-for/ba-p/88035

1White paper on M-series SPARC Enterprise Servers:
HIGH AVAILABILITY IN THE DATACENTER SUN™ SPARC® ENTERPRISE™ SERVERS
http://www.filibeto.org/sun/lib/hardware/enterprise_Mx000/SPARCEnt-HA-Final.pdf

2Power and cooling feature comparison based on HP white paper “HP Superdome 2: the Ultimate Mission-critical Platform”:
http://www8.hp.com/h20195/V2/GetDocument.aspx?docname=4AA1-7762ENW&cc=us&lc=en and Oracle white paper: “Oracle’s Sun SPARC Enterprise M3000, M4000, M5000, M8000, and M9000 Server Architecture Flexible, Mainframe-Class Compute Power”:
http://www.oracle.com/technetwork/articles/systems-hardware-architecture/m-seriesarchitecture-163844.pdf and physical dimensions system specs comparison: M9000 specs:
http://www.oracle.com/us/products/servers-storage/039222.pdf,
Superdome 2 specs: http://h18000.www1.hp.com/products/quickspecs/11717_div/11717_div.pdf

3IDC: Solutions for the Datacenter’s Thermal Challenges, January 2007

4Uptime Institute Whitepaper “Cost Model: Dollars per kW plus Dollars per Square Foot of Computer Floor for redundant UPS capacity”.

5IDC worldwide server tracker, 3CQ10 high-end RISC+EPIC market share, see the chart at: http://h30507.www3.hp.com/t5/Mission-Critical-Computing-Blog/Whose-clunker-do-you-want-cash-for/ba-p/88035

Intel is a trademark of Intel Corporation in the U.S. and other countries.

Don’t believe the FUD from a competitor that is losing ground. HP has transformed the definition of mission critical by converging the modularity of blades with the mission critical resiliency Integrity is renowned for. HP uniquely delivers all the components of a Converged Infrastructure with HP servers, storage, networks, and software and services. Pre-integrated and tested solutions and HP Services expertise help you deploy confidently and quickly so you can achieve the cost savings that a blade infrastructure delivers for your mission critical workloads.

To learn more about HP Integrity servers please see: www.hp.com/go/integrity