Chapter 82

Let's not delve into how Chen, the big boss, was in a frenzy. After Li Xuan found that his brother had matured a lot, he felt at ease handing over the follow-up matters to him.

The Oriental Electronics Laboratory and the Oriental Semiconductor Laboratory of Central University have officially commenced operations and have started accepting project applications. Of course, in addition to accepting applications, the laboratories will also release a series of targeted research projects based on the commission from Oriental Electronics.

The first research project commissioned by Oriental Electronics is about the design of the MOS6502 processor. Oriental Electronics has just reached an agreement with American MOS Technology, paying a licensing fee to obtain the deep authorization for all patents of the 6502 processor. This allows them to develop a CPU more suitable for their needs based on the MOS6502 CPU.

The commission from Oriental Electronics hopes to further optimize the design based on the original 6502 CPU core, making the new processor more suitable for game operation. This project was selected by Li Xuan at first glance, designed by Dr. Cai Kang, the man with a middle part hairstyle whom Li Xuan met at Professor Zhao's weekend salon. It became the first project to kick off in the two laboratories.

The 6502 processor is the CPU used in the first generation of Oriental Electronics' arcade boards, but it is not a processor specifically designed for games. Dr. Cai Kang's new design retains the 6502 CPU core and integrates a series of peripheral components suitable for game operation, including an interrupt controller, a memory management unit, a timer, an 8-bit parallel I/O port, and a programmable sound generator.

The new processor has a 64KB logical addressing space and a 2MB physical addressing space. Since the 6502 is an 8-bit processor, the width of the internal registers is also 8 bits, while the address bus is wider. To access the entire address space, Dr. Cai specially designed a segmented addressing mode, using a memory management unit to divide the memory into multiple 8KB segments.

Integrating a programmable sound generator directly into the CPU can eliminate the need for a separate sound chip in the game motherboard design, further reducing hardware costs.

Dr. Cai Kang's design has many commendable points, which is why Li Xuan selected it at first glance. Through this design, Li Xuan also understood Dr. Cai's professional capabilities. Oriental Electronics would not miss such talent, and the relevant departments of the company would conduct a detailed investigation of Dr. Cai's situation and develop a detailed recruitment plan based on his needs.

Oriental Electronics currently has almost no research and development capabilities. After the initial screening by the two laboratories, all the submitted projects need to be personally reviewed by Li Xuan. He urgently needs to expand the company's R&D team.

Hong Kong is short of electronic talent, and the existing excellent talents are mainly concentrated in three institutions: Central University, the University of Hong Kong, and the Polytechnic College. Through industry-university-research cooperation, Oriental Electronics can almost sweep them all. The company has already begun to look beyond Hong Kong, focusing on Taiwan, Singapore, the UK, and the US.

Due to its small size, Hong Kong's electronics industry has only rapidly risen in the past four to five years. Many early Hong Kong students with aspirations in the electronics industry had to seek development elsewhere. However, those who grew up eating Cantonese cuisine and speaking Cantonese may not adapt well to the Mandarin-speaking environments in Taiwan and Singapore or the English-speaking environments in the US and the UK.

If a Hong Kong company can offer the same treatment and development opportunities, it is believed that Hong Kong engineers working abroad would be willing to return to a city they are more familiar with and find more comfortable. The Human Resources Department and the Public Relations Department of Oriental Electronics are systematically carrying out these tasks.

Recently, Li Xuan has devoted most of his energy to a RISC (Reduced Instruction Set Computing) development project group he is collaborating on with Professor Zhao Weiming, the director of the Oriental Semiconductor Laboratory. Before returning to Hong Kong to teach in the early 1970s, Professor Zhao Weiming participated in the development of the RCA1802, known as the ancestor of RISC processors, at RCA in the US. He is not unfamiliar with the RISC concept, which was proposed just a few years ago.

RISC (Reduced Instruction Set Computing) is in contrast to CISC (Complex Instruction Set Computing). After the birth of electronic computers and decades of development, their instruction sets have become increasingly large, and the design of CPUs to implement these instructions has become increasingly complex.

However, statistical data shows that most of the operations that consume CPU and other computer resources are a small set of commonly used simple instructions, while the vast majority of complex instructions are rarely used. This results in low utilization of the complex structures designed in the CPU.

Faced with this situation, computer experts began to study the necessary simplification of instruction sets, making each instruction simpler and more uniform. They aimed to achieve complex computer operations by executing a series of continuous, simple, and densely encoded instructions, thus replacing the rarely used complex instructions in the instruction set to simplify the design complexity of the processor.

In his previous life, Li Xuan was a designer of embedded system low-level programs, and the instruction set architecture he was most familiar with was ARM, one of the most popular RISC architectures in the 21st century.

The number of devices using the ARM chip architecture is 25 times that of Intel. Ninety-nine percent of smartphones and tablets worldwide use the ARM architecture. About 4.3 billion people touch a device with an ARM chip every day, accounting for 60% of the global population.

However, the instruction set architecture developed by British Acorn Computers had a very difficult and precarious early life. When Acorn initially developed computers, they sought technical information for the 80286 from Intel but were refused. They then decided to develop a new, low-cost processor.

In 1985, the ARM1 was released, with a very simple structure of only 25,000 transistors. The following year, the improved ARM2 was released. The ARM2 was a 32-bit processor with 30,000 transistors and an 8MHz clock speed. Its computational power surpassed the 80286, which had 130,000 transistors and a maximum clock frequency of 20MHz.

However, few people paid attention to this new and destined-to-be-famous chip. All industry attention was focused on the 80386 processor, released on October 17, 1985. Although the ARM2's performance far exceeded the 16-bit 286, it was still meager compared to the 32-bit 386 processor and did not cause any ripples at Intel.

By the time Acorn released the ARM3, Intel's 80486 had already hit the market, and the performance gap remained significant. Acorn, deep in financial crisis and technical bottlenecks, finally saw a true opportunity in 1993. TI (Texas Instruments) was trying to persuade an unknown Finnish communications company, Nokia, to enter the emerging mobile communications market.

TI was already a leader in the DSP field but was unfamiliar with CPU business. Among the few companies it could control, it ultimately chose Acorn.

It was with TI's timely and strong support that Acorn successfully developed the ARM7, the most important processor core in the company's history, in 1993. The ARM7 had very low power consumption, making it suitable for handheld applications, and its manufacturing cost was also low, allowing for rapid market expansion.

The ARM7 fully met Acorn's initial design philosophy of "low-cost, low-power, and high-performance." This philosophy also aligned with the needs of 21st-century smartphones, but it was initially imposed on ARM by Intel.

To avoid the unshakable dominance of Intel's 80x86 in the CISC field, Acorn had to choose RISC technology, which could achieve a relatively high-performance processor chip with fewer chip resources and fewer development personnel.