Related Articles

NVIDIA Storage Technology

The nForce 250 series first introduced NVRAID. NVRAID allowed for several compelling features including cross controller RAID and spare disk allocation. With cross controller RAID, a SATA and IDE drive could be used together to define an array. Spare disk allocation had hard disks installed in the system kick in the event of a disk failure.

The nForce 4 series adds a few improvements to the 250 MCP - SATA 3 Gb/s support, Disk Failure Identification, a Dual Controller Architecture, and NCQ.

SATA's bandwidth is calculated by this formula from SerialATA.org

1500MHz embedded clock
x 1 bit per clock
x 80% for 8b10b encoding
/ 8 bits per byte
= 150 Mbytes/sec

SATA 3Gb/s offers double the theoretical throughput up from the current 1.5Gb/s. SATA 3Gb/s should be strictly a embedded clock increase as the SATA specification states that there is only a single bit transferred per clock. Note that SATA 3Gb/s is NOT the same as the SATA II specification. The SATA II specifications calls for more than just a speed increase. NVIDIA's documents claim "more efficient bus utilization, faster data backup and retrieval, and faster recovery from signal drop/failure (asynchronous signal recovery)."

Disk Failure Identification

This is a pretty neat feature that should have been a common sense issue but has not been addressed by any company until now. When a disk fails in a RAID array it is often difficult to figure out which disk is the culprit. With Disk Failure Identification, a picture of the motherboard is brought up and the port where the disk failure occurred is highlighted. This is a SATA only feature unfortunately so IDE RAID arrays will still need to rely on the old guess and check method.

Location of failed drive is highlighted

Dual Controller Architecture

Remember all those tweak guides that encouraged users to turn on DMA (Direct Memory Access)? This allowed for hard drives to transfer information directly from the drive controller directly to the memory without having to go through the system bus. NVIDIA has added an additional path to and from the memory, doubling the bandwidth available to the disk controllers. This should be useful for users with large disk arrays that require simultaneous access.

NCQ

One area that SCSI had over both parallel and SATA drives was Native Command Queuing (NCQ). Traditionally hard disks on the consumer desktop side process disk requests in a linear fashion. This can potentially be a very bad thing and to understand why, there has to be a basic understanding of the physical structure of a hard disk. Hard disks are made up of platters or disks, much like a compact disk. Each platter is divided into tracks which are concentric circles, tracks are divided into sectors. Each platter is read by one or more heads. Seeking data is fastest when the data resides on the same track. Moving between tracks is time consuming. Consider the case where there are three pieces of data, one on the outermost track, one on the inner most track and one on the outmost track. In a traditional hard disk, the data on the outer track would be read first, then the data on in the inner track second, and finally the third piece of data on the outer track is read. This is not efficient and the time it takes to move the head is the seek time. If the head movement can be minimized, the seek time will decrease accordingly. This is where NCQ comes in - NCQ can rearrange the order of instructions so instead of moving from the outer track to the inner track, both pieces of data may be read from the outer track first before tackling the inner track.

NCQ requires both controller and hard disk support, so unfortunately NCQ will not work on most SATA drives today but most disk manufacturers will support NCQ in the very near future.