What are you looking for ?
Advertise with us
RAIDON

History (1992): Key Dates on Mainframe’s DASD Technology

From 1954 to 1992 and beyond, by StorageTek

1954
Until now, core memories (within the CPU) had been sufficient to meet computer users needs for storing and processing data. It was becoming more apparent, however, that additional storage devices were needed.

Even in the middle 1950s, users were beginning to demand higher levels of performance, and faster access, than magnetic tape could provide. These devices became standalone units, that is, outside of, and separate from the CPU or central processing unit.

In 1954, IBM Corp. received its first magnetic disk patent and employed it on its first commercial disk storage system, the 350 Disk Storage, announced in 1956 as part of IBM’s RAMAC 305 (Random Access Method of Accounting and Control) business computing system. The RAMAC 350 Disk Storage occupied an area that filled a 9-by-12 foot room and stored 5MB of data in a stack of 50 fixed-disk platters, each 2 feet in diameter. The capacity of the RAMAC would occupy one square inch of the surface of a platter on the 3380K-type device announced by IBM in September 1987. The data transfer rate on the RAMAC was 10Kb/s. The price/MB, a number by which all following DASD would be measured, was $7,800.

1962
The first disk drive to use 14-inch diameter media was introduced in late 1962. It also was the first device to use a removable disk pack.

1964-1969
The 1964 through 1969 period was a time of high activity as DASD came into its own as the key element of storage. The IBM System/360, introduced in April 1964, provided the base for the largest and broadest line of computer equipment ever introduced. The IBM 2314, introduced in 1965, was the mainstay disk unit for the System/360 family and maintained its premier position into the early 1970s. Beginning in 1968, other disk manufacturers began to compete in the end-user market with IBM-compatible disk drives. These manufacturers were known as PCMs.

1970-1974
This period was noted for a significant number of improvements in price, performance and capacity. The new System/370 family of processors paved the way for the rapid growth of online environments and dramatic increases in storage requirements conditions, which encouraged the emerging compatible manufacturer industry.

Storage Technology, a PCM, introduced the DASD subsystem feature called dual port that permitted any two actuators in the DASD subsystem to transfer data simultaneously. The state-of-the art disk drive in the early 1970s, the 100MB 3330-1, offered roughly the same capacity that many desktop personal computers have today. The data transfer rate on the 3330-1 was 806Kb/s, the cost was down to $238 per megabyte.

1975-1979
This is the beginning of modern DASD. Innovations include fixed and sealed media, fixed heads, thin-film heads, fixed-block architecture and significant improvements in the areal density of recording. The PCMs had a significant impact during this time.

1980-1990
In June 1980, IBM introduced the 3380. It was destined to be the de facto standard device architecture for the entire decade. On the original device there were two HDAs (Head Disk Assemblies), each with two arms, to provide a capacity of 2.52GB per unit. By 1987, with the announcement of the 3380K model, capacity had tripled and the list price per megabyte was down to just under $14. Successful cached DASD control units appeared in 1981.

1990-1992 and beyond
History has indicated that we can expect a new DASD family about every 5 years. Recent technology developments have shortened the cycle. The decade of the 90s should accelerate that trend. Areal densities are increasing while disk diameters are decreasing.

The IBM announcement and shipment of the 10.8-inch-diameter 3390 in 1989 anticipated many of the changes of the new decades, other manufacturers followed suit and the era of the 14-inch disk is drawing to a close. User requirements for 100% availability and “never-fail” DASD – or as near to never as possible – are driving the next advance in online storage for both mid-range and high-performance processor environments. RAID and advanced array technologies will provide the fault-tolerance online processing requires.

This article is an abstract of news published on the former paper version of Computer Data Storage Newsletter on issue ≠49, published on February 1992.

Articles_bottom
ExaGrid
AIC
ATTOtarget="_blank"
OPEN-E