Intel or Nvidia? nForce 680i Challenges Intel P965 and 975X

SLI Memory With EPP

In this context, SLI doesn't really refer to the Scalable Link Interface and EPP doesn't stand for an Enhanced Parallel Port. The buzzwords rather say that the nForce 680i SLI chipset supports Nvidia-certified memory, which the system will automatically operate at the best possible settings. To do this, the system reads the memory's Enhanced Performance Profile (EPP), which is an extension to the SPD ROM and adds detailed clock speed and timing information for up to DDR2-1200 speeds. The conventional auto setting will read the modules' SPD to determine safe timings at which these modules will work, but it doesn't optimize for performance in any way like the SLI memory approach. SLI also allows automatic memory overclocking of up to 15%, if desired.

MediaShield Storage Technology

From a features point of view, Nvidia is more than on par with Intel, offering six Serial ATA/300 ports with NCQ support, plus an UltraATA/133 channel and RAID capabilities. RAID, however, is only available across the SATA ports. All of this is called MediaShield technology, which includes a simple wizard to set up your RAID array. The software will also alert the user about failed drive in a visual way, which means that the program will show you the port where the faulty drive is located.

Dual Gigabit Ethernet With Traffic Shaping

Nvidia's nForce is the only consumer chipset with two integrated Gigabit Ethernet controllers. It implements a technology called FirstPacket, which is nothing less than a client-based traffic shaping feature to provide sufficient networking bandwidth for prioritized applications such as VoIP or gaming. Nvidia also supports teaming up both Gigabit ports for redundancy, or to increase network performance between nForce and a compatible system. TCP/IP and checksum creation are performed in hardware, reducing CPU load.

You can find more information about the nForce 680i SLI chipset on the Nvidia website.