IBM Advanced Computer Systems project

The ACS-1 and ACS-360 are two related supercomputers designed by IBM as part of the Advanced Computing Systems project from 1961 to 1969. Although the designs were never finished and no models ever went into production, the project spawned a number of organizational techniques and architectural innovations that have since become incorporated into nearly all high-performance computers in existence today. Many of the ideas resulting from the project directly influenced the development of the IBM RS/6000 and, more recently, have contributed to the Explicitly Parallel Instruction Computing (EPIC) computing paradigm used by Intel and HP in high-performance processors.

After the ACS project folded, the engineers were given to choice to rejoin other divisions of IBM. Many declined as it would require them to return to the east coast from California. Many of these went on to form MASCOR, Multi Access System Corp, in 1970. This was short-lived as they were unable to raise capital. Gene Amdahl took the opportunity to start his own company, building IBM-compatible mainframe computers using the ECL designs worked on for ASC. Amdahl Corporation's 470V/6 were both faster and less expensive than IBM's own high-end designs.

History

The ACS project began in 1961 as Project Y with a goal of “building a machine that was one hundred times faster than Stretch”. Initial work began at the IBM Watson Research Center. Several significant computer pioneers contributed to the project, including John Cocke, Herb Schorr, Frances Allen, Gene Amdahl, and Lynn Conway.

In late 1964, Amdahl had taken a teaching position at Standford University. In January 1965 he was named an IBM Fellow for his work on the System/360, which by this time was one of the most successful computers in history. Around the same time, IBM decided to centralize development of the ASC at a new location in Mountain View, California. As a Fellow, he was entitled to work at any IBM facility of his choosing, and having heard of the move, asked to join ASC. After a short time, in August 1965 he argued that the machine should be based on the System/360 instruction set, which caused ASC management to ostracize him.

Over the next two years, the issue of 360 compatibility was argued back and forth. In January 1967, Ralph L. Palmer asked John Backus, Robert Creasy, and Harwood Kolsky to review the project. Kolsky concluded that this would be too difficult, and pointed out that the ASC was aimed at the CDC 6600 market, not the 360's, so if the customer was interested in compatibility, 6600 would seem more useful. The next month, Amdahl once again argued for 360 compatibility for marketing reasons, and then in December he met personally with Kolsky to demonstrate how this might work. The previous month, in November 1967, Herb Schorr had presented a timeline for the first delivery in 1971 with a development budget for software on the order of $15 million. Amdahl argued that a fast 360 platform would offer much of the performance yet eliminate the need to develop a new operating system and all of the associated programming languages and support, and that that money would be better spent on improving OS/360 in general, which would improve the entire line.

As a result of this ongoing argument, Amdahl was ostracized within the team. As further punishment, they assigned John Earle to work for him. Earle was a brilliant circuit designer, but his interactions with other team members led to him being forced out as well. Earle and Amdahl then designed a new concept using the ASC's circuit design to produce a high-end 360. The two quickly proved that one could use the ASC circuit designs to build a 360 that was slightly faster than the existing models, but cost perhaps 75% as much to build, with only 90,000 gates instead of 270,000 (a gate requires about five transistors using the ECL logic of the era).

In December 1967, Amdahl began calling people within IBM to tell them about the new design, which eventually led to a management overview and "shoot out" between the two approaches. Lynn Conway would later conclude these numbers were unlikely to be anything close to correct, but management accepted them in any event. In May 1968, the decision was made to go with Amdahl's approach, and this resulted in the name change from ACS-1 to ACS-360. At its peak, the ACS-360 project involved over 200 engineers and staff.

The ACS-360 project was canceled in May 1969; Amdahl claims this was primarily due to it upsetting IBM's carefully planned pricing structure. The company as a whole had an understanding that machines above a certain performance level would always lose money, and that introducing a machine that was so fast would require it to be priced in a way that would force their other machines to be reduced in price.

When the project was canceled, many of the engineers were not interested in returning to the main IBM research campus in New York. and wished to remain in California. Some ended up at IBM's hard drive research facility in San Jose, California, while many others left to form a new company, Multi Access System Corp, or MASCOR. This failed to raise capital and folded over a period of only a few months. Amdahl then formed his own company to build the system he had outlined with Earle, introducing it as the Amdahl 470/6 in 1975. Amdahl Corporation would become a major vendor of IBM-compatible systems into the 1980s, when the mainframe market began to shrink.

Many of the innovations resulting from the project would eventually find direct realization in the IBM RS/6000 series of machines (later known as the IBM System p line of workstations and servers), apart from influencing the design of other machines and architectures.

Influence

Although neither the ACS-1 nor the ACS-360 was ever manufactured, the IBM Advanced Computing Systems group responsible for their design developed architectural innovations and pioneered a number of RISC CPU design techniques that would become fundamental to the design of modern computer architectures and systems:

• Aggressive reduction in the number of logic gate levels for pipeline stages to reduce the cycle time
• Tight integration between processor and memory
• Cache memory with streamlined I/O to/from cache
• Compiler optimization techniques
• Virtual-memory operating systems
• Multiple instruction decode and issue (a first)
• Use of a branch target buffer (a first)
• Multithreading implemented in hardware (a first for IBM)
• Dynamic instruction scheduling/out-of-order execution
• Hardware register renaming
• Instruction predication
• Level-sensitive scan design (used by IBM)
• Fixed-head hard disks
• Air-cooled high-speed LSI circuits
• Advanced simulation tools used in the design process

Further reading

• Mark K. Smotherman; Edward H. Sussenguth; Russell J. Robelen (2016). "The IBM ACS Project", IEEE Annals of the History of Computing". 38 (1): 60–74. doi:10.1109/MAHC.2015.50. {{cite journal}}: Cite journal requires |journal= (help)

External links

• IBM Advanced Computing Systems (ACS) — 1961–1969 — Documentation project for the IBM ACS-1 supercomputer maintained by Mark Smotherman