Small Computer System Interface (SCSI) online Quiz

Fast SCSI uses synchronous data transfer to achieve speeds up to 10 MB/s. Standard SCSI defaults to asynchronous transfer, but the Fast SCSI specification relies on synchronous timing protocols to coordinate rapid transmission cycles without waiting for acknowledgment handshakes. Asynchronous transfer is too slow for Fast SCSI requirements.

SCSI-1 and SCSI-2 rely on simple parity protection (one parity bit per byte) on the parallel bus for error detection. CRC was introduced in later SCSI standards like Ultra160.

SCSI-3 employs Cyclic Redundancy Check (CRC) for robust error detection of data transferred on the bus, replacing the simpler single-bit parity used in SCSI-1 and SCSI-2. CRC can detect multiple bit errors and provides better data integrity protection. Parity was the legacy method, while CRC is the modern standard.

The SCSI Architecture Model (SAM) for distributed communication has exactly three layers: Application Layer (highest), Transport Layer (middle), and Interconnect Layer (lowest, physical). Since options A, B, and C are all valid SCSI layers, none of them is 'not a layer.' The double-negative construction makes this tricky, but 'None of the Above' correctly indicates that all listed options ARE actual layers.

SCSI-1 defined the core architectural layers: Application Layer, Command Layer, and Protocol Layer. The Programming Interface Layer was introduced in later SCSI specifications (SCSI-2 and beyond) to provide a standardized interface between applications and the SCSI device driver. This layer did not exist in the original SCSI-1 standard.

SCSI-1 specified a narrow (8-bit) bus that could support up to 8 devices total, including the SCSI controller/host adapter. This means you could connect 7 peripheral devices plus the host adapter. The 8-device limit was a key constraint of the original SCSI-1 standard, later expanded to 16 devices with Wide SCSI.