Chapter 118: Not Such A Small Change
The Chinese scientists present quickly guessed the origin of the Raspberry Pi with just a few words, relying solely on their own judgment.
Not only would Lin Ran feel gratified upon learning this, but Dean Qian, who was right there at the scene, was already gratified by the keen smell of all the scientists.
He clapped and said, “You all are indeed experts in the semiconductor and computer field.
A very good guess.
To avoid my words interfering with you all, I have not provided you with more information.
Your guess is already very close to the information left by the comrade who found this equipment for us.
In the final message he transmitted, he also mentioned that this is a computer made of transistors, and he provided detailed explanations for every component on it.
Including the usage of this Raspberry Pi, he wrote it very clearly.”
The experts felt a pang in their hearts upon hearing this.
Final message? As expected, it’s equipment obtained at the cost of life!
Everyone inwardly sighed; as experts, being able to have a safe environment in the rear was already a kind of fortune.
“Alright, everyone, go back and rest first.
There will be a lot of work ahead, and it will be very arduous.
Everyone needs to have a good body; everyone’s burdens are exceptionally heavy.”
Letting them see the Raspberry Pi first this time was also to facilitate subsequent work.
After all, suddenly dragging people from places like Yanjing and Shanghai over thousands of miles with their families couldn’t lead to them working with resentment.
After Wu Xijiu returned to the two-bedroom, one-living-room apartment assigned to him, he originally planned to crash and sleep, but his wife hadn’t slept yet and patted his arm: “Old Wu, what exactly is this sudden trip here for?”
Wu Xijiu waved his hand: “Nothing, sorry, in short, we might not be able to go back for a long time this time.”
Although he felt this work was very meaningful.
Whether it was the things he saw earlier or learning the tense current situation from Dean Qian’s mouth, it made him have no complaints about the current work arrangement.
He even felt a bit fortunate instead, fortunate that he had chosen to return to China back then.
If he hadn’t chosen to return to China, how could he possibly contact something like the “Raspberry Pi,” the most advanced computer, in America?
He didn’t need to think to know that this research project was highly secretive to Chinese people.
Not everyone is Lin Ran, and not everyone can have Lin Ran’s talent and opportunities.
But Wu Xijiu still felt somewhat sorry to his wife.
Even dragging his wife and child to come here with him to work.
The conditions here were much harsher than before in Yanjing.
“It’s fine; if you can really make a contribution to the country, then do it well.”
Similar words occurred in each of their homes.
Reality is like this; since it can’t be changed, just accept it well.
Not only were semiconductor-related experts and engineers gathering here, but experts in aerospace, airplane design, aviation, and missile design were also gathering in Panzhihua.
After obtaining the Raspberry Pi, following the big shot meeting, two priorities were set.
The highest priority was not to replicate the Raspberry Pi, but to develop long-range ballistic missiles that could directly threaten the American homeland.
That’s right, this is the highest priority task.
From China’s launch silos to the American homeland, the range would need at least 8,000 kilometers.
At this time, China had only successfully replicated the Soviet Union’s R-2 short-range ballistic missile, with a range of only 600 kilometers.
In other words, the range needs to be multiplied by ten.
This is the task assigned by Dean Qian.
He also issued a military order at the meeting: if it couldn’t be achieved in five years, he was willing to accept any punishment.
Originally, China’s first intercontinental missile was to wait until the DF-5, starting in 1965, with the first full-range test success in 1980.
In other words, Dean Qian had to shorten the original 15 years to 5 years.
The current technical bottlenecks in China included propulsion technology; that is, liquid fuel rocket engine technology was immature, with limited thrust, making it difficult to support long-range missiles.
The guidance system was backward; whether inertial navigation or gyroscope technology, it was backward, with precision insufficient for intercontinental strikes.
There were major defects in high-strength heat-resistant materials and precision processing capabilities, unable to meet the re-entry into the atmosphere requirements of intercontinental missiles.
In computing power, it relied on mechanical calculators and early electronic computers, severely limiting simulation and optimization of complex designs.
Among them, the Raspberry Pi could solve most problems; for example, fluid mechanics simulation could greatly reduce the design time for aerodynamics and propulsion systems.
Data analysis and improvements after each test firing originally took months; the Raspberry Pi could speed it up to one week.
But the Raspberry Pi is not omnipotent; high-strength heat-resistant materials and precision processing capabilities still need to be supplemented, liquid fuel rocket engine research and development needs to be accelerated, and the guidance system needs optimization.
In short, five years is very short, but considering the current situation, China has no choice but to do so.
After countless strategic seminars, Yanjing had an unprecedented sense of urgency.
What Lin Ran said was that America discovered alien technology and then provided the Raspberry Pi to us.
Who can guarantee they only discovered the Raspberry Pi? Nothing else?
What if there are some Top Gun weapons?
The Soviet Union is right next to China; what if they have some top-tier weapons or other black tech?
What if what America discovered is another world?
They have resources from two worlds to develop; what if the content in “The Man in the High Castle” is real, which is also the information Lin Ran wanted to convey to them?
China doesn’t know who Philip K. Dick is; not to mention now, even in China twenty years later, he is an obscure science fiction writer.
The reason America hasn’t expanded outward now is because they started a war in another world, fighting against Germany and Japan in that world; once they win there and complete integration, what then if they start conquering here?
In short, after everyone interpreted the news Lin Ran conveyed back, they had only one thought: if we don’t have some trump card, we’ll become meat on the chopping block.
It would truly be others as the knife and cutting board, us as the fish meat.
So Dean Qian’s highest priority task is missiles that can directly hit the American homeland.
China’s goal in developing missiles is of course not just America, but also the “good big brother.”
The second priority is to replicate the “Raspberry Pi” and fully develop semiconductor technology.
For this, the entire area was allocated to them, and it will continue to expand in the future.
In addition, China will exhaust every means to collect all academic journals from America, Europe, and the Soviet Union.
Just so that even if the information in them can provide a little help, it’s good.
If last year was just leaving an opening for academic journals to possibly enter, now that opening has become a pipeline.
The Chinese side treats obtaining academic journals as one of the most important tasks for their foreign affairs department, assigning it to comrades in Eastern Europe.
Lin Ran didn’t expect that this time it could have such a good effect.
In the end, it still requires a crisis of survival or extinction.
China’s scientists were first provided by Dean Qian with a simulation software operation manual for each of them, then he gave them a rough explanation.
The remaining whole day was reserved for them to fully verify the Raspberry Pi.
Xia Peisu pointed to the teletypewriter screen: “Everyone, this is the result of the Raspberry Pi running the simulation software.
I inputted a simple transistor amplifier circuit model, and it generated the output waveform in a few seconds. This is countless times faster than our manual calculations and experimental verification.”
The term “software” here came from directly translating “software” from Lin Ran’s book.
Xie Xide pushed up her glasses in surprise: “As Dean Qian said, it only took a few seconds.
Our previous Model 103 took hours for the simplest calculation, yet this little box has such speed. How strong is its computing power exactly?”
Xia Peisu: “I think we need to design a rigorous comparative experiment for this.”
Huang Kun didn’t speak, frowning and pondering, occasionally burying his head in paper to calculate something.
Only after Xia Peisu finished did he speak: “Its floating-point operations might be hundreds of millions? About the same as my imagination.
I didn’t sleep last night after going back and roughly calculated its theoretical computation limit. Then today I verified it again; I believe its computing power is at 2 million operations, possibly up to 400 million at most.
When we arrived earliest, I had Dean Qian run a task with known computation amount first, doing a 100×100 matrix multiplication operation.
The task took less than 2 seconds; for simplicity, I treated it as two seconds.
The computation amount for matrix multiplication is approximately the cube of the number of rows and columns in floating-point operations, so a 100×100 matrix requires about one million floating-point operations.
That means it takes 2 seconds for one million operations; during that time, the Raspberry Pi ran very smoothly. I roughly estimated its clock.
I conjecture its clock frequency is between 1 MHz and 10 MHz, but actually it might be far more than that; I’ve already overestimated, yet I still can’t imagine its limit.
Considering it can handle complex simulation computations, it might have multiple computing units; perhaps all the black squares on it are computing units.
There are a total of 4 black squares here, so its minimum is 2 million operations per second, maximum 400 million operations per second.
Of course, my calculation is very conservative, based on the information we know now; maybe it can reach higher, like 4 billion?”
Xia Peisu nodded: “4 billion I’m a bit skeptical; we need to run some complex tasks to see.”
Huang Kun’s model was based on 1962 technical cognition, including frequency and instruction efficiency, but without modern concepts like pipeline and cache.
Due to lack of internal details of the Raspberry Pi (like 1.5 GHz quad-core processor), his estimate was conservatively low.
But the 4 billion he casually mentioned was already close to the real value.
“I think our focus needs to be on how to use it well, not on analyzing it.
We need to first make something similar.”
Xie Xide said: “This is of course very useful. We now research semiconductor material properties, often needing approximate formulas and manual derivations.
With the Raspberry Pi, it can directly give numerical solutions, with much higher efficiency.”
Wang Shouwu excitedly tapped the table: “Professor Xia, that simulation analysis interests me greatly. When we design transistors, we always have to repeatedly experiment and adjust parameters. Now with this software, we can directly do optimized design.”
After Dean Qian returned and heard everyone’s report, he smiled bitterly: “You all can use it unlimitedly only during this period.
Even during this time, you have to use it sparingly; the usage frequency can’t be too high.
If it breaks, we can’t bear that responsibility.”
If it broke, Dean Qian would want to die; failing to develop the intercontinental missile in five years is a small matter, losing the Raspberry Pi is a big deal.
“Later, everyone needs to apply in advance to use it.
Because not only you all need it, other experts do too; it can play a bigger role in fluid mechanics.
When other experts use it, they pass notes with what they need calculated, and dedicated operators complete the computation and return the results to them.
While you all can operate it directly.”
The scientists present exchanged looks, and Xie Xide replied to Dean Qian as representative:
“Dean, please give us more usage time.
It can play a huge role in many fields, like simulating PN junction and transistor volt-ampere characteristics, calculating carrier concentration, diffusion length, and other parameters. The Raspberry Pi can greatly accelerate this process.
Huang Kun and they can use it to calculate energy band diagrams of semiconductor materials, verifying theoretical models through effective mass approximation, providing basis for material selection.
Lin Lan Ying and they can use the Raspberry Pi to simulate the growth process of single-crystal silicon or germanium, optimizing temperature gradient and purification parameters, reducing experimental trial and error costs.
Including calculating impurity distribution and defect density in crystals, improving material quality.
Design and simulate simple transistor amplifier circuits, calculating gain, bandwidth, and other parameters.
Simulate frequency stability of oscillation circuits, etc.; it can provide considerable help.
Additionally, we discussed that with the Raspberry Pi, I think we can rely on it to develop a world-leading transistor radio.”
Previously, radios produced in China, like “Panda” brand and “Red Light” brand, all used vacuum tubes.
But that doesn’t mean they hadn’t researched transistor radios.
China’s first domestically produced transistor radio “Red Star” was launched by Hongxing in 1963.
Obviously, how could Red Star compare to a transistor radio made with the collective wisdom of so many experts concentrated and helped by the Raspberry Pi?
“Because we want to advance transistor miniaturization, we discussed together: on one hand, reserve technology based on transistor miniaturization to make a sufficiently small radio.
Additionally, the Raspberry Pi’s simulation software can simulate the core circuits of transistor radios, including high-frequency oscillators, intermediate-frequency amplifiers, and detector circuits. We can quickly input transistor parameters and test effects of different designs.
Design a simple superheterodyne receiver circuit, and the Raspberry Pi can give waveform analysis in a few minutes.
Including optimization of existing domestic transistors and simulation of transistor physical characteristics, all can be verified on the small transistor radio.
If we make it well, we might not lose to Texas Instruments in transistor radios.
