NVIDIA will be producing two nForce3 Professional variants: the nForce3 Pro 150 (which is available now) and the nForce3 Pro 250, which will be available in the fall. Here are the key specs for the nForce3 Pro 150:
NVIDIA enterprise-class networking technologyNVIDIA IEEE 802.3 MAC (Media Access Controller)
Supports 10/100 BaseT Ethernet/Fast Ethernet
ACR and CNR Interface Support
NVIDIA StreamThru technology
NVIDIA RAID Technology
RAID 0 disk striping for highest system and application performance
RAID 1 disk mirroring support for fault tolerance
RAID 0+1 disk striping and mirroring for highest performance with fault tolerance
Triple Fast Ultra ATA-133 disk drive controllers
Each interface supports two devices for a total of six devices
Third IDE interface can support two IDE-to-SATA bridge master/slave fonifugration devices to support up to 2 SATA ports
Third IDE channel additional supports an overclocked mode to achieve UDMA-150 to match the 150MB/sec of SATA link interface
AGP 8X Interface
Support for up to 6 ports
AC’97 2.1 compliant interface
Supports 2, 4, or 6-channel audio
SPDIF output (Stereo or AC-3 output)
Single chip design
If you recall the Opteron architecture, the memory controller is now integrated on the CPU itself, rather than the North Bridge of the chipset. The integrated memory controller is good for CPU’s performance, as it significantly reduces latency and increases memory bandwidth.
nForce3 Pro block diagram
We’ve also found that an inefficient memory controller can hamper overall system performance. For example, VIA had to rework the memory controller in its KT266 chipset, hence the KT266A was born. By removing the chipset’s memory controller from the equation, AMD has more control over system performance, regardless of chipset.
NVIDIA’s single chip design makes life even easier for motherboard manufacturers, as the simplified board design gives them even more flexibility, but all the credit can’t go to the CPU’s integrated memory controller. Like other nForce products, nForce3 Pro utilizes a 0.15-micron manufacturing process, allowing NVIDIA to cram more features on the chip. This also lowers power consumption and dissipated heat. Another benefit of single chip chipset architectures is reduced latency, increasing platform performance.