There are several striking aspects to this article about the University’s Computer Centre, found in a dusty 1970 edition of ANU Reporter.
First is the black and white image of Computer Centre staff testing the interface between a computer and one of its “standard processing devices” – a trusty typewriter.
Computers have come a long way since then. The days of punch cards and metres-long printouts are now, depending on your age, either a nostalgic memory, a (thankfully) distant awkward stage, or, for Gen Ys, a point of complete incredulity.
It’s hard to reconcile today’s palm-hugging smart phones with the six-foot processors of yore. Or to fathom that 1980s systems made up of hundreds of square metres of hardware possessed processing power equivalent to that of a 9-inch Apple iPad 2.
But in some ways, today’s cutting-edge research computers are not so different from their long-decommissioned ancestors.
The evolution of computing at ANU can be traced from the IBM 360/50 computer pictured, through the ANU Supercomputing Facility from 1987 and the Australian Partnership for Advanced Computing from 2000, to the current National Computational Infrastructure (NCI), which was established in 2007.
While the system that currently provides ANU researchers with number crunching power – NCI’s Raijin supercomputer – takes up a similar amount of room to some of the earliest computers, it packs a whole lot more crunch. The original IBM system could perform up to 150,000 calculations per second; Fujitsu’s Raijin easily handles more than one thousand trillion. In other words, Raijin is six billion times faster. And the 256 kilobytes of memory that the 360/50 computer offered is almost laughable in comparison to Raijin’s 160 terabytes.
But perhaps the most striking similarity in this deluge of disparity is the fact that despite the exponential growth seen in computing capacity and capability since 1970, the demand for computational services is still outstripping availability.
Almost half a century on, the titular “computing crisis” is yet to be solved. Last year, NCI received requests for more than twice the amount of compute time available – not an insignificant amount when you consider that Raijin is the largest and most powerful supercomputer not only in Australia, but in the entire Southern Hemisphere.
This research simply wouldn’t have been possible 44 years ago. Or, more accurately, it would have been possible but would have taken decades, or even centuries, to complete. The question is: what will computers be able to do 44 years from now?