spacer Home > Support > Knowledge Base > Disc Knowledge Base > FAQs spacer
spacer
spacer Ultra160 Technology Paper spacer
spacer
 

Ultra160 Technology Paper

Ultra160 SCSI, also known as Ultra3 SCSI LVD, is a new SCSI interface standard. SCSI has a 15-year heritage of maintaining full backward compatibility and excellent forward compatibility. Every two years or so, small, evolutionary changes are made to the standard, improving speed, reliability, and manageability. Ultra160 SCSI is the next step. Implementation is straightforward, easy, and low-risk.

Ultra160 SCSI provides double the transfer rate of its predecessor, Ultra2 SCSI, while increasing manageability and reliability. All changes are incremental, meaning that Ultra160 SCSI devices are backward compatible with single-ended Narrow and Wide SCSI devices. The ANSI standards T10 committee defines Ultra160 SCSI in the SPI-3 (SCSI Parallel Interface) specification. The SPI-3 standard evolved from the existing SPI-2 specification.

Database servers, RAIDs, audio and video editing systems, workstations, and high-end desktop systems benefit from:

  • Double the transfer rate of Ultra2 SCSI at up to 160 MB/sec
  • Compatibility with existing Ultra2 SCSI cabling for easy integration
  • Ultra2 (80 MB/sec) and Ultra160 (160 MB/sec) devices can coexist on the same bus without degrading performance of Ultra160 devices
  • Increased reliability through improved error detection and correction
  • Ability to monitor interface performance and determine fastest reliable transfer rate
  • Cable length of up to 25 meters with a single device, 12 meters with two or more devices
  • Supports up to 15 devices

Note. Ultra160 SCSI devices are not compatible with High Voltage Differential (HVD) SCSI devices.

Note. Seagate Ultra160 SCSI disc drives do not support SCSI-1. See below.


Ultra160 SCSI employs proven techniques to achieve increased performance and reliability.

Three features take Ultra160 SCSI performance to the next level:
  • Double Transition Clocking
  • Domain Validation
  • Cyclic Redundancy Checks (CRC).

Double Transition Clocking
Ultra2 SCSI uses a SCSI request/acknowledgement or "clocking" signal to synchronize the timing of devices on the SCSI bus. The clocking signal alternates between "off" and "on" at 80 MHz. Data is transferred each time the clock signal makes the transition from "off" to "on". This is referred to as single transition clocking because the device only synchronizes with the off-to-on transition, ignoring the on-to-off transition.
Ultra160 SCSI improves upon this by employing double transition clocking. Both off-to-on and on-to-off transitions are used, effectively doubling the rate of the data signal without increasing the rate of the clocking signal. For example: The clocking signal on Ultra2 SCSI runs at 80 MHz, while the data signal runs at just 40 MHz. This results in a transfer rate of 80 MB/sec on a 16-bit-wide bus. By using both transitions of the same 80 MHz clocking signal, the data rate can be increased to 80 MHz, resulting in an overall transfer rate of 160 MB/sec on a 16-bit-wide bus.
Single Transition Clocking

Double Transition Clocking


Since the clock rate remains the same, double transition clocking provides double the data transfer rate of single transition clocking without compromising reliability. Keeping the rate lower results in longer clock pulses, decreasing the possibility of timing errors and increasing noise tolerance. The real-world benefit of slower clock rates is a reduction in the number of EMI issues encountered by system designers.

Cyclic Redundancy Checks
Ultra160 SCSI includes the addition of a Cyclic Redundancy Check during data transfer. This is the same proven method of error detection and correction utilized by FDDI, Ethernet, and Fibre Channel. CRC provides an extra measure protection for systems with marginal cable routing or external devices, and is one of the best ways to ensure data protection during hot plugging. Ultra160 SCSI CRC detects all single bit, double bit, odd number, and burst errors. Since data integrity is verified during transfer, undetected error rates are reduced dramatically.

Domain Validation
Domain Validation intelligently tests all elements of storage networks to ensure that the network is operating within specifications. The Ultra160 host adapter tests cables, backplanes, terminators, expanders, bridges, and so on. If reliability is at risk, the host adapter lowers the transfer rate and proceeds without a hitch. This happens in a manner similar to the way modem and fax transmissions compensate for varying equipment and line conditions. In the past, new devices such as host bus adapters (HBAs) and hard drives sometimes did not work smoothly with legacy configurations. Domain Validation helps assure that Ultra160 SCSI devices operate smoothly in existing legacy systems. Testing is transparent to the user and happens automatically without changing controller settings, setting BIOS parameters, or fumbling with manuals.

Seagate Ultra160 SCSI Disc Drives
Seagate's high-performance Cheetah and Barracuda drives, drive models with a suffix of LC, LCV, LWV or LW, offer the latest technology by utilizing the Ultra160 SCSI parallel interface capabilities. See our Disc Encyclopedia for detailed information on specific models.

Summary
Ultra160 SCSI provides system OEMs with a simple and cost effective solution for fast, reliable data delivery. In addition, system designers can take advantage of the flexibility offered by Ultra160, easily allowing them to optimize a platform for a particular market. Easy integration with existing Ultra2 SCSI devices—without sacrificing Ultra160 performance—means existing Ultra 2 SCSI users can upgrade incrementally while realizing immediate performance gains.


Note about Seagate Ultra160 SCSI disc drives and SCSI-1
In order to make them ANSI SPI-3 compliant, SCSI-1 features and formats are no longer supported in our enterprise disc drive products beginning with the Barracuda 18XL, Cheetah 18XL, Cheetah 36LP, and Cheetah 73HH product families.

The ability of a drive to function in the SCSI-1 operating definition requires that the drive's SCSI controller firmware includes support for the Change Definition (40h) command and Mode Parameter page formats as defined by the ANSI SCSI-1 specifications. The Change Definition command is required to change the drive's operating definition from SCSI-2/3 to SCSI-1, or vice versa. In the SCSI-1 operating definition, Mode Parameter pages 00, 01, 02 & 04 are not fully defined, as compared to the SCSI-2/3 (ANSI SPI-3) definitions, and pages 07, 0A, 1A & 1C do not exist. In addition, the methods for specifying Logical Unit Numbers (LUNs) in SCSI-1 have been changed in SCSI-3. The LUN field (Byte 2, Bits 5-7) in the Command Descriptor Blocks (CDBs) has been eliminated. Also, the method for specifying the LUN in the Identify message (Bytes 0-5) has been redefined. These changes have been implemented in the Barracuda 18XL, Cheetah 18XL, Cheetah 36LP and Cheetah 73HH products. Because of these differences, systems that support SCSI-1 only will not be able to function with the previously mentioned enterprise SCSI-2/3 products.

spacer