Chapter 96 Just because you didn't do it first in the world doesn't mean you shouldn't do it

Friday, February 29.

Theoretically, the winter vacation has been over for several days, but a person like Gu Biao with a card, of course, can take personal leave and return to school after the Lantern Festival tomorrow.

He has too many things to do in Qiantang, and he is not there for the whole semester, so he can only rely on the winter vacation to deal with it intensively.

After thoroughly arranging the product line and shipping channels of Tiankun Media, Gu Biao spent the last few days looking for Yang Hao and Lu Yunda to plan the next stage of research and development tasks.

The meeting venue was set up in the exhibition hotel in Red Sun Square, and Gu Biao specially spent foreign exchange to invite everyone to the foreign guest restaurant for dinner, and then rented a conference room to talk about business, in order to make his subordinates' blood boil, so as to pay more attention.

From the trial production of "Bomberman" and "Pac-Man" to the establishment of a new generation of products, there was only a break of less than a week in the middle, or because of the Chinese New Year.

This kind of rhythm is definitely very fast for mainlanders in the 80s, like forcing life. But it also allows people to grow quickly and fully explore their potential.

Moreover, the attractiveness of the commission offered by Gu Biao is also beyond the imagination limit of mainland college students: for every game console sold, the project team will be given a commission of US$10, which will be distributed by all participants according to their contributions.

Gu Biao can control their knowledge and ensure that the other party can't see first-hand information about the outside world when he was a student, so this money is definitely enough.

After finishing a French meal and going to the conference room to make black tea with milk, Gu Biao began to draw flatbread.

"The project I'm going to talk about today has actually been well thought out in my head for a long time. If I have a technical team as mature as a large foreign company, I can even be a 'Pac-Man' and start a project at the same time.

But I don't have a team, so I can only count on your growth to keep up. Now, we have completed a ported and improved version based on the Z80 CPU (Thunderbolt fighter) and three independent R&D versions of the Z80 CPU (Tank Wars, Pac-Man, Bomberman).

I believe that you have already run through the whole process of how to understand a new CPU instruction set, supporting circuit layout, and have also proved your learning ability. Therefore, this time, we are going to take on the challenge of changing the platform for the first time - starting from scratch and making an arcade machine with a new CPU for the ship.

I hope that you will pay full attention to the follow-up projects, because history will prove that the previous business was just to make a little money, but it was just a one-shot deal. What we have to do today is a great preparation to pave the way for the occupation of a new media channel in the future!"

"Brand new CPU, can we do it?" Yang Hao and Lu Yunda were actually a little cowardly,

Based on the idea of a programmer, Yang Hao reminded very pragmatically: "Moreover, is there any gap that the current CPU is absolutely unattainable and insurmountable? I think that the product should not only consider the blind pursuit of advanced, but should be aimed at the performance effect, which is what the boss personally taught us." ”

"Of course, there are reasons why you have to use a new CPU, so I'm going to talk about the general idea. Gu Biao waved his whip undoubtedly,

"Let's talk about the CPU I want to change first, I'm going to use the Motorola 68000 model, this CPU has a frequency of 3MHz, although it seems that the main frequency is not as good as Intel's 8086 4M, but the price is only 60% of Intel.

I also value the 68000's 32-bit registers and 16-bit data bus, as well as its high refresh rate. In this indicator, Intel does not attach great importance to it, and it can even be said that it is not forward-looking, still using 16-bit registers and data buses, and the improved version of 8086, 8088, has been reduced to 8-bit buses in order to save money......"

Ma Yanglu and the three of them were a little dazed when they heard it, but it was still easy to understand the first reason, that is, the cheapness of the Tumotorola goods.

At that time, the price of CPUs in the international market, sorted by several major suppliers, was about the following: Intel was the most expensive, then Motorola, then Zilog (the company that made the Z80), and the cheapest was MOS-6502.

Measured on the basis of MOS 1x, the price of the same performance indicator is about 2 times that of Z80, 3.5 times that of Motorola, and 6 times that of Intel.

Historically, when Nintendo began to engage in home machine FC in August 83, considering that the home machine market must have cheap hardware and expand market share, the cheapest MOS home 6502 was used (this is a series, there are constantly new specifications upgraded, Nintendo only used that in 83 years, and it already has 8M main frequency, which is much stronger than the early 3M).

At that time, the price of the CPU was already outrageously low at $15, which also prompted the final shipment price of the FC in North America to fall below the $100 mark. This low price also finally contributed to Nintendo's seizure of the market after the collapse of Atari and the cumulative shipment of more than 30 million units in North America.

The motive for buying a bargain was understood by all three. As for why the register/bus and memory access speed are more important than the CPU clock speed when arcade machines, they can't understand it for the time being, which requires further unity of thought at the meeting.

"Boss, why are you so pursuing memory access speed and refresh rate?" Lu Yunda, who was in charge of the hardware, had to ask.

"Because the game I want to develop in the future will be a scrolling side-scrolling game, and I want to give users the best screen refresh rate experience, and I don't want any flash. Of course, there are many other indicators that cannot be repeated for a while.

Just remember, our goal is to create the best side-scrolling game in the world and create this category. And then let our peers eat ash behind our butts. In your specific operation, if you find any architectural changes and component selection principles that can better serve this goal, you can always put them forward and discuss with me. What I have just said is only a matter of purpose. ”

Several people were still a little vague, and Gu Biao had to explain what a side-scrolling customs clearance game was.

He used as concise language as possible to explain the problem clearly, and the officials naturally did not want to do it again.

Anyway, Super Mario, Contra, Double Dragon, and all the games in which a small plane flies from left to right and finally kills a boss are all called "side-scrolling clearance games".

"The games we used to make were all 'fixed screen', so everything on the interface didn't have to be constantly calculated and refreshed, and the requirements for the number of data transmission bits and memory access speed of the CPU were low.

And once the horizontal screen scrolls up, recalculating terrain and background changes every second, the Z80 is a bit stretched. It's not that it can't be done, but it can only be used with interlaced displays, which is destined to be very flashy even if it is realized, hurting the eyes. ”

This is not alarmist, but has a historical basis - as mentioned earlier, he was not very interested in medieval games in his previous life, so all his experience came from watching Director Ao of Station B in his previous life.

He didn't know the inside secrets of the ancient game that were not introduced in Director Ao's video. He vaguely remembered what the video had touched.

And in his previous life, he happened to watch an issue of "The Defender" narrated by Director Ao in August 2018, which was the world's first side-scrolling airplane game in his previous life, and it was also the highest-selling arcade machine in the North American market (limited to North America, because Americans especially like to play handjob, especially the material of playing UFO and saving the earth)

Director Ao didn't say that the Defender arcade was released in '81 or what CPU it used, but he only knew that more than a year later, Atari tried to port it to the Atari-2600 home console.

And because the 2600 game console is connected to the TV, all the CPU access data lines are also adapted to the home TV instead of the dedicated display, and the result is that the picture performance is extremely bad.

The thrill of killing alien flying saucers on the original silky smooth arcade plane, when it comes to the home plane, the small plane will flash wildly as soon as the laser plane is launched. The fire button is pressed so fast that it will even cause both the plane and the enemy's flying saucer to disappear for a few seconds, completely losing the gaming experience.

Gu Biao knew this conclusion in his previous life, and now he has combined it with the basic common sense and scientific literacy of a coder after this era, and conducted a preliminary appraisal of various components, and found the problem.

That is, any processor and design circuit that is "enough to cater to the TV display" is definitely not enough when it comes to arcade machines.

Because the TV set in this era has to be set up to "interlaced scanning" in order to cater to the function of "playing movies on the TV", to be precise, "interlaced missing frame scanning". The CPU can also be made enough for the TV, and the end is a crazy flash.

(At that time, the TV was set to display only 25 frames, which was designed to combine the power plant's alternating current to 50 Hz, so it was made to "only display one frame every two refreshes", and 50 was reduced to 25.)

The playback speed of the film is 24 frames per second, which has nothing to do with alternating current, and is completely due to the fact that the film projector/camera was used to rolling at 24 frames per second in the early years (there were only 16 frames per second in the early years), which resulted in the difference between film and television by exactly 1 frame.

84 years ago, the rough practice of TV manufacturers was to recognize that when playing a movie signal, they directly replaced the last frame of each second with a white screen flicker, which is equivalent to the first 0.96 seconds of playing a movie, and the last 0.04 seconds to show you a white screen. Then I felt that the screen was very flashy and hurtful to the eyes.

When this mechanism is combined with "interlacing", the effect is super bad when combined with game consoles. To play games with high refresh rates, you need to use a specially set up arcade monitor, and you need to have more CPU and memory access margins. )

Only by making the CPU "more than enough to match the current TV", and with a specially designed display, can the side-scrolling game experience be silky.

Of course, this problem is limited to a few years.

83 years later, after Nintendo discovered the detour that Atari had taken back then, it began to deepen cooperation with TV manufacturers, and persuaded Sharp to participate in the strategic alliance in 84 years to produce a new generation of progressive scanning, a new generation of high-definition TVs with improved refresh rate and no missing frame processing, which solved the pitfalls encountered by Atari that year. (The "HD" here is relative to pre-'84 televisions, when Sharp did pioneer the new generation of "HD" standards.) )

So the current situation of the industry is that peers are still learning from Atari's "advanced experience" and are shocked by Atari's dominance in the world.

Atari, on the other hand, was still content with the status quo and didn't plan to waste the budget, so others followed suit and didn't consider more expensive luxury designs.

Historically, Nintendo only took out this improvement after the collapse of Atari, realizing that "the money that Atari didn't spend doesn't mean that it shouldn't be spent, and what Atari didn't do doesn't mean it's superfluous."

And Gu Biao realized it now.