Fifth Generation Computer Systems facts for kids

The Fifth Generation Computer Systems (FGCS) was a big project started in Japan in 1982. It was a 10-year plan by Japan's Ministry of International Trade and Industry (MITI) to create super advanced computers. These computers were meant to use many processors working at the same time (called massively parallel computing) and a special way of programming called logic programming.

The main goal was to build an "epoch-making computer" that could act like a supercomputer and help develop artificial intelligence (AI). The project was very ambitious and tried to do things that were far ahead of its time. Because of this, it didn't become a big commercial success. However, it did lead to important new ideas in computer science, especially in an area called concurrent logic programming.

The name "fifth generation" was chosen to show how advanced these computers were meant to be. Before this, computers were grouped into four "generations" based on their main technology:

• First generation: Used vacuum tubes.
• Second generation: Used transistors and diodes.
• Third generation: Used integrated circuits (silicon chips).
• Fourth generation: Used microprocessors.

Each generation usually focused on making a single computer processor faster. But the Fifth Generation project believed the future was in using many processors together to get much more power.

Computer Generations

For a long time, people talked about different "generations" of computer hardware. Here's a simple way to think about them:

Hardware Generations

• First Generation (mid-1940s): These computers used vacuum tubes. They were very large and used a lot of electricity. An example is the IBM 650.
• Second Generation (1956): This era brought transistors. Transistors were much smaller than vacuum tubes, used less power, and didn't get as hot. This helped computers become smaller. The IBM 7090 is an example.
• Third Generation (1964): This generation used integrated circuits, also known as silicon chips. These chips could hold many transistors, making computers even more powerful and compact. The IBM 360/91 was an early example.

After the third generation, computers started using Very Large Scale Integrated (VLSI) circuits, which packed even more parts onto chips.

Software Generations

There were also "generations" for computer software and programming languages:

• First Generation: Machine language, which is the basic code computers understand.
• Second Generation: Low-level programming languages like Assembly language, which are a bit easier for humans to read than machine code.
• Third Generation: Structured high-level programming languages like C, COBOL, and FORTRAN. These are much easier for programmers to use.
• Fourth Generation: "Non-procedural" high-level programming languages, often used for specific tasks or in object-oriented programming.

Before the 1970s, Japan mostly built computers by following ideas from the U.S. and Britain. But in the mid-1970s, Japan's MITI decided to start looking into the future of computing on its own. They asked the Japan Information Processing Development Center (JIPDEC) to explore new directions. In 1979, the idea of a "fifth-generation computer" started to be discussed.

MITI had a history of successful projects, like improving the steel industry and creating supertankers. They believed that information technology was the next big thing. However, the Japanese language, especially its written form, was hard for computers to process. This challenge was one reason MITI wanted to push computer technology forward.

Project Goals

The main goal of the Fifth Generation Computer Systems project was to build parallel computers for artificial intelligence (AI) using concurrent logic programming.

The project aimed to create a truly new kind of computer. This computer would have supercomputer-like power and would work with huge databases instead of traditional filesystems. It would use a logic programming language to manage and access data, all while using many processors working together (parallel computing).

The project hoped to build a working model that could perform between 100 million and 1 billion "Logical Inferences Per Second" (LIPS). At that time, typical computers could only do about 100,000 LIPS. The plan was to build this machine over ten years: three years for early research, four years for building different parts, and three years to finish the working system.

In 1982, the Japanese government approved the project. They created the Institute for New Generation Computer Technology (ICOT) with money from various Japanese computer companies.

Logic Programming

The FGCS project wanted to develop "Knowledge Information Processing systems," which basically meant advanced Artificial Intelligence. The main tool chosen for this was logic programming.

Logic programming uses rules of logic to:

• Store information in a computer.
• Give problems to a computer.
• Solve these problems using logical thinking.

Simply put, in logic programming, a computer program is like a set of rules or facts. When you ask the computer a question, it tries to prove the answer using those rules. This process gives you the result.

Logic programming was seen as a way to connect different areas of computer science, like software engineering, databases, computer architecture, and artificial intelligence. It seemed to be the missing link between creating smart systems and building powerful parallel computers.

Project Outcomes

The launch of the FGCS project caused a lot of excitement and even some worry around the world. Many believed that parallel computing was the future. Soon, similar projects started in other countries, like the Strategic Computing Initiative in the US and Alvey in the UK.

The FGCS project ran from 1982 to 1994 and cost about US$320 million. After the project ended, MITI stopped funding such large computer research projects. However, MITI and ICOT did start a new project on neural networks in the 1990s, which some called the "Sixth Generation Project."

Concurrent Logic Programming

A key development during the project was the invention of Concurrent Prolog by Ehud Shapiro in 1982. This new programming language combined logic programming with concurrent programming, allowing different parts of a program to run at the same time.

Shapiro's work inspired the FGCS project to focus on concurrent logic programming as its main software idea. It also led to the creation of Guarded Horn Clauses (GHC) by Ueda, which became the basis for KL1, the main programming language used by the FGCS project.

The FGCS project greatly helped the development of the concurrent logic programming field. It also trained a new group of talented Japanese researchers.

Commercial Challenges

The FGCS Project did not become a commercial success. This was for several reasons:

• The highly specialized parallel computers built by the project, called Parallel Inference Machines (PIM), were eventually outperformed by less specialized, cheaper computers like Sun workstations and Intel x86 machines.
• The idea of using concurrent logic programming as the main link between parallel computers and AI applications didn't work as smoothly as hoped. Different languages were developed, but they all had limitations.
• Regular computer processors (CPUs) became much faster than experts expected in the 1980s. This made the need for complex parallel computing less urgent for many tasks. "Off-the-shelf" computers quickly became more powerful than the specialized FGCS workstations.
• The project also didn't adapt well to new ideas happening outside its walls. During its time, GUIs (graphical user interfaces) became common, the internet allowed databases to be shared across networks, and new ways of finding information in data (data mining) emerged. The FGCS workstations simply couldn't compete in a market where general-purpose computers could do more for less money.

Ahead of Its Time

Even though it wasn't a commercial success, the Fifth-Generation project was revolutionary. It did a lot of basic research that predicted future directions in computing. Many research papers and patents came out of it.

Many of the ideas from the Fifth-Generation project are now being explored again with modern technology. For example, in the early 2000s, single CPU speeds started to hit limits. This led to a new focus on parallel computing, with multi-core processors in everyday computers and massively parallel processing in supercomputers. Even Graphics cards like Nvidia and AMD now use large parallel systems.

However, it's not clear if these new technologies directly used research from the FGCS project. The project's direct impact on the computer industry hasn't been clearly shown.