1 2 3 4 5 6 7 8 9

- Portability of computers and computer storage is improving dramatically.

- Rapid access to the local South Pole network will be required for imported portable

- Most users prefer and need to work in the compute environment with which they are

- Support for large items such as monitors, printers, and some media (i.e., 1/2 inch
tape drives) will continue to be required.

- The traffic on the local science network will be modest with respect to "ethernet"
throughput (10 Mbps, 50 Kbps untuned transfer rate).

- Bit mapped displays are becoming increasingly important (and affordable).

The following describes the several levels of science requirements including intent, operability, hardware support, software support, and user support. In addition, some consequences of these requirements on implementation are discussed to provide assistance in timely upgrading the South Pole computing capability.

Requirements - Intent:

The South Pole computing system should provide the scientific community:

- Access to data (file transfer)
- Access to colleagues (mail)
- Access to instruments (interactive login from US)
- Access to analysis

These requirements are in approximate order of ease of implementation. The priorities are so close, particularly for the first three, that they should be considered equal.

Access to data is a fundamental requirement. Many experiments planned for winter operations at the South Pole will require almost instant access to data. Prior to the availability of satellite transfer, data had to be returned by mail during the brief summer season. This often resulted in a two year response time between detection of an experiment anomaly and providing a fix (data recorded in Austral winter, data return in Austral summer, discovery by analysis in Austral winter, fix in Austral summer). The ability to transmit a limited amount of data permits at least experiment monitoring. The desire, of course, is to return a substantial number of the measurements close to the time of acquisition.

Access to colleagues with the advent of electronic mail at South Pole has vastly improved the ability to include CONUS staff in diagnosis and repair of problems, as well has permitting investigators to continue some of their other research functions while sequestered at South Pole. The current implementation, which involves a non-standard, labor- and time- intensive file transfer step, is frustrating to the majority of researchers, who have come to rely on the ready access and rapid response of mail over the national networks.

The remote interactive logon has been an unfulfilled requirement for several years for currently active experiments. This involves remotely varying experimental parameters