"How is SEE's E/LM4023 audio source chip?" The E/LM4023 is a 16-bit general-purpose sound processor introduced by SEE in 1980 and was first used in arcades. In 1981, IBM adopted the E/LM4023 as the standard audio source chip on the PC.

E/LM4023 is characterized by high performance, low price and high cost performance, many famous audio companies in the eighties, such as Sony dual card recorders, have E/LM4023 as a standard chip.

Li Yixuan also rarely started a new stove, so he chose it directly. As for the final speaker, do you need to say, since E/LM4023 is selected, the speaker is also a universal type, and the two trumpet tape recorder speakers are used to design an appearance.

To tell the truth, E/LM4023 and small cassette recorder speakers are all garbage in the eyes of Li Yixuan, a traverser, even the cheap copycat stereos sold on the stalls in later generations are more than 100 times stronger.

But there is no way, Li Yixuan has no better choice, the sound that can be heard a little bit is expensive to death.

From these small details, it can be seen that in the wild era of personal computers, there are too many hardware conditions that restrict personal computer manufacturers, and manufacturers need to constantly make trade-offs when designing computers in order to achieve a balance in all aspects.

At this time, personal computers did not have CAD-aided design gadgets, and the design and development of chips still needed manpower to complete.

Yu Youcheng divided his students into two project groups, one group followed him to develop the CPU; One followed Li Yixuan to develop GPUs; As for the chipset, it can only be done after the CPU is finalized.

After the advent of the Oranges I computer, he tried to develop CAD software, but found that the difficulty was not ordinary, on the surface it was an ordinary software, but it involved quite a lot of things.

Circuit operation requires special circuits, and the relevant instructions solidified in general-purpose processors are scarce, and the operation speed cannot meet the requirements; There is no design program developed specifically for circuit operations......

It's nothing, grind it slowly, and you can grind it out. But the point is that there is no relevant underlying data!

The computer itself is a dead thing, it has no mind. Humans give it an electrical signal, and it calculates according to the internal circuit, and then uses an electrical signal. Without all kinds of circuit measurement data, you just draw a circuit, and it is just a geometric figure composed of a bunch of points and lines, which has no meaning. Only the abundant circuit measured data as a reference comparison, after a variety of electrical calculations, is a complete CAD functional program.

Limited by the computer technology of this era, supercomputers can only be used to perform ultra-large-scale data operations in three-dimensional space, which is extremely expensive.

So far, in addition to projects supported by national defense scientific research and other national financial resources, only large companies and consortia such as petroleum, chemical, aircraft, and automobile can afford to use them. There has never been a precedent for personal computer design, and even Big Blue has yet to try it.

How convenient CAD design is, if Li Yixuan doesn't know, it's just that, after using computer-aided design, use the mouse to combine the relevant circuits and components, and the computer will automatically calculate, tell him whether the circuit design is wrong, and point out where the error is, and you can make modifications on the spot. A super-large-scale chip design will not take two or three months.

Li Yixuan also wanted to buy the underlying data of CAD in another time and space, but unfortunately, these companies did not sell it.

There is still a way for others not to sell Li Yixuan, download a few sets of the most popular CAD software on the market on the Internet, and then crack them one by one to call up the underlying data.

After all, another space-time processor has been engaged in for so many years, and the application of various functional circuits in different professional fields has been very skillful. Which circuit has the best effect, which circuit has the widest application area, and the mutual interference and elimination of various circuits after integration ...... And so on, different CAD software companies have their own unique secrets.

Even without the support of another space-time technology, the underlying data of CAD is gradually enriched and improved.

However, now he is facing a technical bottleneck, that is, the amount of underlying data is too large, and the existing computer hard disk cannot be installed at all, and the CPU cannot run.

In other words, Li Yixuan must make a choice.

Li Yixuan carefully analyzed those CAD software and found a pattern. That is, these CAD software are customer installation packages, and after the customer downloads and installs, it is connected to the CAD company's online server, and then the CAD company provides data services, which is such an operation mode.

It's obviously not possible to change to this time and space.,Now the LAN is not popular.,That's the Internet.。

After careful thinking, Li Yixuan digested the fundamental core of CAD, and then according to the hardware technology level of this era, he was reborn and formulated a CAD software design plan, which first graded the underlying data of the cracked CAD according to the level of the level, from 1 to 10 into 10 levels; Each level is divided into 10 smaller levels.

For a 16-bit computer under development, the minimum level of data is sufficient.

Even so, to load all this data, customers need more than 10 large RAID disk arrays to catch up with a tiny database center.

"Professor Yu uses this, don't think about designing it for two years just by manpower, CPU and GPU 1.5 micron process, 270,000 transistors, the workload is too much." After that, Li Yixuan installed the CAD software he designed himself into the Oranges I computer, and connected the computer to the RAID with the SCSI data cable.

Li Yixuan demonstrated it and asked Yu Youcheng to give it a try.

Yu Youcheng tried for half an hour and said, "Your software is good, but how is it semi-automatic?" A lot of things have to be operated by people, and the efficiency improvement is not too high. ”

"The existing 8-bit computer computing technology level cannot run fully automatic CAD software." Li Yixuan explained.

"Are there any plans to launch it to the market?"

"I'm not going to launch an 8-bit computer, I'm going to launch it when our 16-bit computer goes on the market, it's just a trial software, it's for testing. In order to achieve the right commercial application, this CAD needs to improve many details in addition to turning semi-automatic into fully automatic. For example, the problem of LAN needs to be solved. You see, this computer can only be connected to RAID alone, not to other computers; If you don't solve it, CAD design can only be a castle in the air. ”

"This shouldn't be difficult, the United States has already delivered similar equipment to the market, why not buy it?" Yu Youcheng asked.

"The Oranges I computer doesn't have the relevant interface." Li Yixuan replied.

When Li Yixuan designed the computer, like Apple's Apple II, he didn't think about networking at all.

The first is to save costs by not considering networking; Second, the current LAN hardware equipment is too large, and can only be used in larger computers above the minicomputer, and personal computers cannot be integrated into it.

Third, it involves patent issues.

The concept of LAN was formed in the mid-60s of the 20th century, and began to be put into use in schools in the early 70s, and then a small number of scientific research units also began to try LAN. However, due to the limitations of technology and price costs, only large enterprises, research institutes and universities can apply it.

The large-scale popularization of LAN in the United States only began in the 80s.

Why 80 years? Because the LAN patent expires in 80 years. In the United States, except for pharmaceutical patents, which are 20 years, all other types of patents are 15 years like those in Europe and later republics.

The concept of LAN was proposed by a professor at Stanford in the United States in 65 years, and in the same year he registered the relevant patents and licensed the patents to several large American companies such as IBM, National Semiconductor, DEC, Xerox, and Texas Instruments.

In order to monopolize the LAN market, these companies reached an exclusive agreement with Professor Varshna, the inventor of the LAN, that is, several companies paid a considerable patent licensing fee to the professor every year, and at the same time took out a very objective amount of money every year to support his research and development; But there is one condition, and that is that Professor Varshner can no longer license patents to other companies, especially those outside the United States.

How much is this royalty? Combined, several companies can earn as much as $1 million per year. You know, this is 1 million dollars in the 60s, and the Bresendon system at that time has not yet disintegrated, it is a real gold dollar.

For a professor who teaches in a university, this is a huge amount of money, and many people may not be able to make so much money in their lifetime, not to mention that there is a considerable amount of research funding every year.

This kind of good thing will not be agreed, so the professor happily signed the contract.