Siheyuan: I eat melons while doing scientific research

Chapter 1317 Each Has Its Own Use (5k)

Gao Zhendong himself did not give a report, partly because he was lazy, and partly because it would be beneficial for his colleagues in the technical department to have more exposure in front of relevant colleagues across the country.

For the colleagues in the Semiconductor Research Laboratory of the Third Factory's Technology Department, writing the project reports for Chief Engineer Gao is the most comfortable. First, it is substantial enough; second, the completion rate is high enough; and third, very few people ask tricky questions.

Because it's difficult to get answers to questions. Chief Engineer Gao's research is almost always groundbreaking, with significant prospects and implications. Those who attend his reports are considered impressive if they can understand them; as for tricky questions, they're unlikely to dare ask them. Furthermore, Chief Engineer Gao's research projects are so comprehensive that they don't leave room for others to ask such questions.

So when they give reports, they are so confident that even if they are wrong, it sounds like they are right.

However, the results this time are worthy of their momentum. In principle, the thing is not very new, but it is very useful.

"According to Chief Engineer Gao's classification, this thing might be a bit expensive to use on a general-purpose computer."

"Yes, but this thing isn't used on general-purpose computers at all! It's used in conjunction with microcontrollers and similar devices in various control systems."

"With this, we no longer need to worry about modifying the control system program. If we want to modify the program, we can do it ourselves directly, unlike before when we didn't dare to make small changes at all, and we would only dare to ask our colleagues at the semiconductor factory to re-open the mask when the changes had accumulated to a certain extent."

This is the greatest good news for all those working on control systems.

Previously, the programs were directly integrated into dedicated read-only chips in semiconductor factories. This caused a major problem – they dared not modify them, or rather, modifications were useless. Minor repairs were simply not enough to make the semiconductor factory staff redesign the entire mask, which was too wasteful of money.

Therefore, some of my colleagues even hoped that any modification would be a major overhaul, something that would seriously damage the structure, in which case the problem of re-molding might actually be solved.

This method is much easier for comrades now. Want to change it? Erase and rewrite it!

The previous method used oracle bone script, which was extremely inconvenient to modify and rewrite.

Now, this ultraviolet-erasable non-volatile semiconductor memory has become like a pencil with an eraser, capable of writing, erasing, and correcting.

Although pencil marks are easily erased and EPROM is less reliable than direct photolithography, this slight difference is not a problem at all and can even be ignored.

Just like writing with a pencil, as long as it is properly protected and not deliberately erased, the carbon itself will not change much over a long period of time.

The happiest people here are the missile engineers. They now understand why they invited relevant personnel from several missile research institutes as soon as possible. This thing is very convenient for them, especially when they are conducting research and the design is not yet finalized.

Moreover, this method can significantly reduce the cost of missiles. After all, even if a large number of missiles are used, it's only a few hundred or a few thousand, which is still a small number compared to the cost of creating a dedicated photolithography mask.

"This is great. We've been struggling with whether to use magnetic tape, hard disk, or photolithographic read-only memory. The former is too unreliable and can't be used in missiles, while the latter has many limitations. This chip combines the advantages of both and discards all the disadvantages. It's amazing."

"The only problem is whether this thing will be convenient to use." Another comrade was a little worried. At this time, semiconductor technology still had a mysterious aura and was undoubtedly high technology of this era.

The thought of permanently altering the contents of a semiconductor chip, a capability that seems to put it on par with semiconductor manufacturers, inevitably raises concerns about its convenience.

Their concerns were quickly resolved.

"The new erasable chip is easy to operate. Simply irradiate the optical window with specific ultraviolet light for 20 minutes to erase the original contents inside the chip. Writing is also very simple. It can be written under a high voltage of 12~24V."

To Gao Zhendong, this set of operations might not seem simple, after all, he had used USB flash drives, SD cards, and SSD hard drives. But to the comrades in front of him, it was as simple as it could be.

"What? It's that simple? Just shine a light on it and adjust the voltage, and you can erase and write?"

"And the whole process takes less than half an hour, which is incredibly fast!"

Gao Zhendong found it hard to believe. Writing a few thousand or tens of thousands of words in half an hour is considered fast? You guys are too easily satisfied.

But for our comrades, this speed is really fast. In the past, to achieve the same effect, it would take months, and that was under the condition that everything went smoothly.

Alternatively, one can only use less reliable methods such as magnetic tapes or disks, which may work under certain conditions but are completely unusable under other conditions.

"...This chip features a long lifespan, a high number of erase/write cycles, and a large capacity..."

The lithography process that Gao Zhendong is using is, after all, 3 micrometers, so it's not difficult to increase the capacity, although in his eyes, "more" and "larger" are not the same thing at all.

"...The chip can be erased and rewritten at least 100 times, and the capacity is available in two specifications: 4 kilobytes and 32 kilobytes..."

Gao Zhendong didn't go for too many specifications; one small and one large were enough.

The comrades were furious when they heard this.

"100 times? How can we possibly use that all? That's way too much!" In fact, if it were used as a finished read-only memory, 100 times would indeed be a lot. At this point, the comrades made a similar exclamation to Bill Gates' later reaction when faced with 640KB of memory.

It can't be used up, it can't be used up at all.

"100 times! That means we can upgrade the existing missiles directly after improving the algorithms and control programs, bringing performance or functional improvements at a very low cost, and we can upgrade them 100 times! That's fantastic!"

"Let alone upgrades after finalization, even improvements and modifications before finalization are much more convenient with this! In the past, our test missiles couldn't be molded with photolithography chips, so we could only use RAM chips and batteries to support various modifications during the research process, which wasn't very reliable. Now with this, it's much more convenient! I don't believe there's any equipment that can't be fixed even after 100 modifications."

Dirty Three: It's hard to say. I don't know how many times I've done it, but the things I've made are definitely from the past few years.

Not to mention those working on missiles and other equipment with extremely demanding working environments, even those in the 11th and 12th Ministries of Machinery Industry who were involved in electromechanical control were equally excited.

They were all chief engineers, and the two ministries and other relevant comrades showed great respect for Gao Zhendong's invitation.

Gao Zhendong invited them for a simple reason: as the top technical experts in their respective ministries, they could grasp the information about ultraviolet EPROM as soon as possible and promote this technology in their respective industries at the fastest speed.

What made these comrades happy was not just the aforementioned factors, but also the capacity!

Unlike missiles, which operate under extreme conditions and environments but have a single function and purpose, their control systems may not have such high requirements, but they often have many and complex functions. This leads to another problem: large program size!

Even though the various control programs at this time were written in C, a highly efficient compiled language, and even assembly language was used for compilation and optimization in extreme cases, the program size of their industrial equipment control systems was still quite large.

This thing has a maximum capacity of 32 kilobytes, which is enough to hold the control programs for most of their various products.

Don't underestimate this capacity. What you all don't know is that five years later, the main computer of the Citibank Apollo lunar lander had only a little over 70KB of program memory, and it still managed to handle the moon landing very well.

Moreover, to avoid waste, Chief Engineer Gao thoughtfully prepared a chip with only 4 kilobytes, which greatly increased the freedom of use for the comrades.

That's typical of Chief Engineer Gao—thoughtful and thorough in every aspect.

What the comrades didn't know was that on the other side of the telephone line, in a mysterious meeting place, the comrades there were even happier!

Loosely speaking, they were participating in a conference call, only they could listen and not speak. The advent of PBX (Private Branch Exchange) greatly facilitated this special type of meeting; it wasn't that they couldn't hold such meetings before, but they weren't very good at it.

"32KB!!! Hahaha, that means we only need 9 chips to get a storage capacity similar to a double-sided 15cm floppy disk! And even more!" These guys are satellite engineers.

Satellites operate in space, so the amount of RAM chips installed is destined to be limited. Satellites currently in orbit have the same RAM capacity as the Apollo lunar module five years from now—4KB.

Don't underestimate this small capacity; the current leaders in the US and Europe would be green with envy.

The limited RAM places certain demands on the random access capability of the program memory. Under these conditions, magnetic tape is no longer suitable. The satellite's rich functionality results in a large program size, and creating a photolithographic mask for the memory of each unique satellite would be too costly.

In this situation, the comrades had only one option—a 15-centimeter floppy disk.

Satellites in space naturally lack the ability to change disks, so the size of a satellite's program memory is essentially limited to 270kB for the time being.

Even so, it's still quite large. After all, 270kB of program memory is considered incredibly large these days. You haven't seen any spacecraft that landed on the moon, and their program memory was only a little over 70kB.

However, another problem it brings is that this floppy disk drive is quite large and heavy.

The floppy disk and the floppy drive itself are fine, but the floppy disk, which is well protected by the Earth's magnetic field, is more fragile than paper in the radiation-filled space. At least paper does not rely on magnetism to record data and is somewhat more resistant to radiation.

Moreover, aside from radiation, there are also adverse conditions such as acceleration and vibration during the launch process, which are all very fatal to the thin floppy disk. The cost of hardening the floppy disk drive is much greater than the drive itself.

Vibration must be prevented, high temperature must be prevented, low temperature must be prevented, radiation must be prevented...

As a result, the weight and size of this floppy disk drive were off the charts. Of course, it was still much better than the 72kB magnetic core program memory of the Apollo lunar module. 72kB magnetic core memory was something that was still an abstract concept.

The technological capabilities of the United States in 69 were terrifying. They used the Hercules rocket, whose blueprints were later said to have been lost, to fly bricks and complete the moon landing. You have to admire them.

The advent of this erasable and rewritable memory solved all the problems our colleagues had with storing satellite programs.

In terms of capacity, nine chips are equivalent to one floppy disk. Keep in mind that the combined area of ​​nine chips and their peripheral circuitry is still smaller than that of a floppy disk.

As we all know, it's much easier to harden semiconductor chips than floppy disk drives. Vibration protection is achieved by applying some glue, high and low temperature protection is achieved by special selection at the factory, and radiation protection is achieved by simple hardening due to their small size...

The differences are so subtle that you can estimate them without even having the chip in hand.

"Haha, even with a full set of reinforcement measures, the resulting storage assembly won't be any bigger or heavier than an unreinforced native 15cm floppy disk drive! This will save a lot of weight and volume!"

"And it's highly reliable! Semiconductor chips are inherently several orders of magnitude more reliable than floppy disks."

"We still have to thank Comrade Gao Zhendong for bringing us newer and better components, helping our equipment develop in a more advanced direction step by step."

"Comrades, don't forget, besides satellites, there are also rockets, and even missiles..."

"Yes, yes, yes! High reliability and small size, this thing is great!"

“In that case, I suggest sending a report to the National Defense Industry Committee and asking them to contact the Third Factory of the Seventeenth Ministry of Machine Building and reserve all the initial production capacity for us.” This sounds rather awkward. Neither “Seventeenth Ministry of Machine Building” nor “Third Factory of the Seventeenth Ministry of Machine Building” is usually associated with chips, while the latter seems to be missing something.

Another comrade laughed: "There's no need for that. You know nothing about the convenience of semiconductor production. They can produce dozens or even hundreds of chips on a single silicon wafer. To put it simply, given our usage, the output of one of their production lines in a single day would be more than we could consume in a year. Note that it's one production line, not one factory. However, it would be very necessary to ask the National Defense Commission to get us some."

The products here are all national strategic assets. The only drawback of these high-quality products is that they are expensive and scarce, so their chip consumption is indeed negligible.

Deep in the desert, other comrades were also paying attention to this thing.

"Hey, with this chip, we're now fully ready to switch from analog to digital warhead detonation devices."

"Indeed, this saves some volume and weight."

"Yes, the warhead detonation procedure is simple. A single microcontroller and a 4KB memory chip are enough. This is much lighter and smaller than the original analog detonation device, and its reliability and functionality can also be improved."

Don't underestimate this small detail; every inch of space on a warhead is incredibly valuable.

Moreover, digital fuses can bring more possibilities to warheads.

For different targets, the warhead may choose to detonate in various ways, from airbursts at different altitudes to penetrating the ground or entering water. To achieve these options, simulation comes at a cost, and its convenience and reliability are incomparable to digitalization. Digitalization, on the other hand, is much better.

"Sure, sure, bring some over to play with, hahaha..."
-
Compared to the individual joys of their colleagues, the joy of Factory 1274 was unique; their joy was something that no other factory could understand or replicate.

This stuff will definitely end up being produced by Factory 1274! Look at them, their happiness is so simple and unpretentious.

The semiconductor process line at Plant 3 is always busy with Chief Engineer Gao's research and development work, so it's impossible for it to be put into formal production. Meanwhile, the other major plant in this system, Plant 1218, focuses on power electronic chips, which is unrelated to this.

When it comes to launching new products, who else but us!
Of course, Factory 1274 was definitely not the most unusual one on site; the most unusual one was the chief engineer of the 17th Ministry of Machine Building.

However, if we had to say, the various control systems in metallurgy could also be used in relation to the main topic of today's meeting.

He was sitting next to Gao Zhendong, listening to the report with a relaxed expression while chatting quietly with Gao Zhendong.

"Zhendong, how's the steel materials handbook coming along?" He asked this question purely as a casual chat, not to urge Zhendong on. After all, this thing didn't need to be finished. As long as one type was developed, it could be put into production as needed without any delay. In fact, the comrades hoped that it would never be finished.

Gao Zhendong laughed and said, "It's still early. We won't be able to finish this year. There's just too much stuff."

Yes, that's the effect I want. It's good that we can't finish it this year.

The chief engineer couldn't stop smiling when he thought about the endless stream of new material reports that the ministry was receiving.

The chief engineer didn't care how Gao Zhendong's brain produced so many results; a genius is a genius.

"Then look at the two projects at the Beijing Steel Plant after the furnace age renovation plan. In your experience, when can you come up with results?"

He's asking about top-bottom reblowing and negative energy steelmaking now.

"I estimate that top and bottom blowing may be faster, but negative energy steelmaking will probably not be completed this year. The latter requires a lot of steel plant infrastructure and renovation. Apart from the converter and workshop renovation, at the very least, a gas cylinder, regardless of size, must be built first."

The recovered steam can be used in the existing steam pipeline system after simple processing, without much modification. However, it's unacceptable to have nowhere to put the recovered gas. It can't just be burned for nothing; otherwise, what's the point of testing, and we wouldn't be able to obtain any system operating data.

"Yes, that's good. With a schedule, things will be easier for the department," the chief engineer said with a smile.

"Huh? Easy?" Gao Zhendong found this strange.

(End of this chapter)