17th April 2007

SPARC Enterprise is the new standard SPARC/Solaris server. It provides a next generation merger of the best technologies developed on Fujitsu PRIMEPOWER and Sun Microsystems Inc., Sun Fire. The result is excellence in performance, business continuity, virtualization and connectivity. The technology also surpasses that of Fujitsu mainframes in high-reliability and availability, while the excellent network computing technology developed on Sun Fire ensures truly outstanding configuration performance. Solaris, the standard UNIX OS, enables smooth application upgrades from both PRIMEPOWER and Sun Fire. It also provides the richest variety of solutions from world-leading ISV and IHV, and Fujitsu middleware and storage products.

With SPARC Enterprise your investments are secure and no business process is too large.

High performance processors and high speed data transfer technologies

SPARC Enterprise high performance starts with application power focused SPARC64™ VI processors in high-end and mid-range servers and throughput power focused UltraSPARC T1 processors in entry servers. High speed data transfer is enhanced by widened bus bandwidths and incorporation of the latest ultra-fast I/O interfaces.

Bus bandwidths between system controllers, CPUs, memory and I/O are much wider. In addition the latest I/O interfaces including PCI-Express and Serial Attached SCSI are implemented. SPARC Enterprise M9000, the largest model now delivers a 737GB/s maximum data transfer rate.

Business continuity

SPARC Enterprise provides solid reliability at all levels of the server. This includes data, component and system level functions to provide uninterrupted platform operation.

High-end and mid-range servers protect themselves from data errors at each internal operation. Data checking and error correction mechanisms are implemented in all SPARC64 VI circuits. Any memory errors can also be excluded by enabling the Memory Mirror mechanism. Basic data routes between components, data buses and address buses are also protected from data corruption by data checking and error correction mechanisms embedded in the large scale integrated (LSI) circuits.

Entry servers also have greatly improved RAS (Reliability, Availability and Serviceability) functions including core dynamic downgrading and data error detection and correction in cache memory and registers.

High fault tolerance functions such as redundancy, hot-replacement and downgrade are implemented in major components, such as disks, power supply units, fans and PCI cards. In particular, high-end and mid-range models provide for uninterrupted operation by incorporating redundancy and/or downgrading functions in almost all components. Moreover, the independent system management boards (XSCF, ALOM), separate from business processing, enable swift and exact failure detection and problem resolution, as well as server status, even in power-off situations.

With reliability implemented at all levels of the server, SPARC Enterprise provides the best in uninterrupted business continuity.

Finally if your systems demand the highest in availability, including applications and external storage, PRIMECLUSTER ensures your overall business continuity. It swiftly detects errors and provides the recovery mechanisms that enable consolidated redundancy across servers, storage and the network.

Virtualization consolidates and optimizes server and business operation

SPARC Enterprise optimizes server use by enabling the consolidation of hardware and software resources through partitioning.

High-end and mid-range servers allow you to consolidate multiple business servers, accommodating the equivalent of 24 separate servers in the hardware partitions of SPARC Enterprise M9000. Each partition supports independent and fully separate OS and application operation in each partition. Because each hardware partition is an independent hardware unit, any errors in one partition can never affect the other partitions. CPU resources and memory can be allotted flexibly, meeting the requirements of demanding applications and changing workloads during server operation.

Software partitioning is provided by Solaris Containers a feature of Solaris 10. This allows you to create up to 8191 partitions. It means CPU resources can be allotted at a fine-grain level to individual or groups of processes to ensure that what ever size your requirement, optimum utilization of resources is always maximized.

Graphically based multi sever management interfaces inter-working with the Server System Manager (standard function), and in-built management processors provide you with unified server management across all systems and partitions.

Investment protection: Upgrade to SPARC Enterprise

Applications running on other Solaris OS versions also run on Solaris 10 without any recompilation. You can benefit from the smooth upgrade provided by Solaris servers.

All ISV and IHV products running on Solaris 10 are available with SPARC Enterprise. You are able to choose the applications and equipment you require from a rich portfolio of solutions.

---