The new Xserve works faster, harder, and more efficiently than ever
before. And it all starts with the 64-bit Quad-Core Intel Xeon
Harpertown 5400 series processor. Thanks to this powerful 45-nanometer
Intel Core microarchitecture processor, Xserve now runs at blazing speeds
up to 3.0GHz. Each Intel Xeon processor has its own 1600MHz system bus for
up to 25.6GB/s of bandwidth, so theres even faster access to main memory.
Four processing cores deliver an incredible performance boost across
the board. A massive 12MB of L2 cache per processor keeps instructions and
data close to the processor cores, reducing frontside bus transactions and
memory latency. Intels cache technology allows core pairs to share 6MB of
cache. If one core happens to be idle or needs less cache resources, the
other can fully utilize it. The enhanced SSE4 SIMD engine handles 128-bit
vector computations in a single cycle.
High-bandwidth memory.
Xserve supports up to 32GB of 800MHz DDR2 ECC FB-DIMM (fully buffered
DIMM) memory, providing a 64 percent throughput boost over the
previous-generation Xserve.1 And over most of the competition.
PCI Express 2.0 expansion.
In every Xserve, there are two PCI Express 2.0 expansion slots one
8-lane and one 16-lane with up to 8GB/s of bandwidth for intensive I/O
devices. Thats up to four times the bandwidth of any previous Xserve. One
slot can be configured as a PCI-X slot for compatibility with existing
devices.
Testing conducted by Apple in December 2007 using preproduction 3.0GHz
8-core Xserve units and shipping 3.0GHz quad-core Xserve units. All
systems were configured with 8GB of RAM. Results are based on the STREAM
v. 5.6 benchmark (www.cs.virginia.edu/stream/ref.html)
using OMP support for multiprocessor-compiled builds. Performance tests
are conducted using specific computer systems and reflect the approximate
performance of Xserve.