A semiconductor can be considered a broad term for different types of transistors.
At the time integrated circuit technology was faster and more cost effective compared to other methods.
Today, semiconductor chips are created with a material called silicon. Known as the "glue of technology", hence the name given to the San Francisco Bay Are in Northern California, "Silicon Valley".
What is MOS?
The acronym, MOS, stands for Metal-Oxide Semiconductor. A method of creating transistors with three key layers. These layers consist of a metal conductor, insulating silicon, and finally a semiconductor silicon layer.
What is RAM?
The acronym, RAM, stands for Random Access Memory. A type of memory chip that can be accessed randomly so any byte or piece of memory can be used without accessing other bytes or pieces of memory.
There are also two basic types of RAM during this time…
SRAM (Static RAM) – These chips are dual-transistor memory cells that require a constant supply of power to retain their content. Common applications for SRAM memory chips include system caches and video memory. Each SRAM bit is a flip-flop circuit made of cross-coupled inverters. This type of memory chip can be accessed quickly, but it is temporary and cost more to produce. The memory is volatile, meaning each memory cell hold a binary digit if power is supplied. If power is lost, so is the data stored in the SPRAM unit.
DRAM (Dynamic RAM) – A type of semiconductor memory that is typically used for the data or program code needed by a computer processor to function. DRAM is a common type of random-access memory that is used in daily technology such as personal computers, workstations, and servers.
Advantages of using RAM:
• Works at higher speeds than all other memories of computer
• Consumes less power than other devices. Some examples would include hard disks, CD drives, keyboard, mouse, etc
• Can perform read and write operations or any data
• Due to RAM, the processor can work faster because of its ability to store data inside of itself
Disadvantages of using RAM:
• Lose all their contents when powered off
• More expensive than its storage
• Operates slower than the cache memory in the CPU
• Contains less storage than all other memory devices
• Volatile memory
Regitz Talks about Developing MOS:
"As we looked at both of those programs with that kind of background and some of the things we were trying to accomplish, which I had brought into, we decided that [Jordan] was going to concentrate on bipolar and I was going to concentrate on MOS."
- William (Bill) Regitz
From a memory standpoint there was no MOS available at the time.
Regitz explains “…the job was to look to see how to make memory and to come up with a low-cost solution, something that would be cost competitive…” (page 35 Comp. Museum)
Regitz looked at this from a theoretical view. He used circuits and other gathered research about MOS. He took parameters from processes and then integrated them directly.
He used the Framingham model of MOS circuit. They had to have a foundation before starting to build the test chip.
Bill Regitz was tasked to work on MOS while Bill Jordan worked on bipolar in a collaborative effort to build the 512-bit memory MOS memory device.
The project started out as simple and kept getting more complicated. This was due to more parasites being added and all the complete model and capacitances.
Bill Regitz divided the workload of MOS and RAM projects into two distinct sections…
Phase I: Build Elements of Chip
Dynamic cell needed to be refreshed from time to time because MOS is not like core. Core needs to stay on forever.
MOS has resistance that is across the capacitor and it would leak over time.
There was big controversy between core and semiconductor. With the semiconductor, if you lost power, you must refresh all the time. There was speculation that it would cause him access issues that was in memory. The system point of view was a worry. Thus, it gave way to a different design of parameters. All this was theoretical design tested with some simulation. The cell seemed to operate properly. The process at this point was to build a test chip, then test the reliability, test concept, then build memory itself. For testing, a test chip was simulated version of the test. Bill Regitz wanted to test reliability activities.
“And as we went to different companies, what we wanted them to do was put their own reliability models that they were using to test their processes out onto the chip itself.” - Regitz
Phase II: Test Reliability and Find Producer
“There was nothing physically available in the market at the time. This is definitely leading-edge development technology.” - Bill Regitz
They programmed eight or nine different companies to run tests on the chip. After the tests were completed, they selected a company to produce the chip. They wanted standard chips built by two or three companies and give them license to build and produce.
They proposed the circuit design for chip. It would eventually get approved.
They met with companies who were also in the memory business to test the chip. Those companies were Advanced Micro Devices, Fairchild, AMS, Mortorla, and a small startup called Intel (1969).
Bill Jordan knew these companies and contacts because he worked with them previously in the semiconductor business. All of the companies made the test chip. Regitz and Jordan gave each company around $20,000 to complete the production for the test chip. Not a lot of money, although just enough to show interest in what Honeywell was doing. The duo would end up choosing Intel as the company to work with to produce the 1102 chip.
The newly formed Intel company ended up going with a different design and publicly released the 1103, the first DRAM chip, in 1970. It would become the bestselling semiconductor memory chip in the world by 1972.
Bill believes that Intel will continue to be the leader of suppliers in the semiconductor world. The Intel company left a long-lasting effect on the computer industry thanks to Regitz’s contributions.
