Chapter 87: 70% Combat Strength
Two-dimensional is flat, and three-dimensional is solid.
Originally, transistors were "etched" onto silicon wafers using photolithography. Changing to three-dimensional would be similar to "building" transistors onto the chip.
The technical difficulty of this is obviously increased many times over.
"The multiple exposure technology I developed earlier seems to be perfectly suitable for the photolithography of 3D transistors. Let’s try it first."
Without complex application processes, without communication, and without resource collection, Tom had just obtained this possibility one second, and the next second, he made a direct decision.
Thus, many Clones who were trying other solutions immediately put down their work and all joined in the attempt of this possibility.
The relevant resources needed for this attempt were immediately in place, without any delay, and the entire process was extremely efficient.
After completing the preliminary design, the experimental multi-exposure equipment immediately began to etch a silicon wafer in the laboratory. Then came a series of inspection procedures.
When the final result appeared before Tom, he widened his eyes with joy.
"It actually works!"
After changing the transistor from two-dimensional to three-dimensional, the short-channel effect indeed disappeared! The performance of the transistor and the stability of the current increased again!
"This is it!"
Less than half a month had passed since this idea emerged, and Tom, based on the experimental data at this moment, directly made a major decision that could change the direction of the entire industry.
Three-dimensional transistors undoubtedly have higher performance, but they also mean a problem.
That is, the technical difficulty is greatly increased compared to before, and accordingly, to use this process, the manufacturing process will also be greatly extended.
From the purification of silicon wafers to the final chip packaging, the entire process adds up to more than 1,700 steps!
Assuming that the yield rate of each of these 1,700 processes is 99%—which seems quite high—in reality, after 1,700 processes, the yield rate of the final product is only 3.8 parts per billion, which is completely unusable.
Only by increasing the yield rate of each process to about 99.99% can the final yield rate be increased to about 80%, which can then be considered practical.
But a 99.99% yield rate for each process... how difficult is that.
But there is no other way, even if it’s difficult. Tom could only follow the strictest design and construction requirements to build this brand new chip factory.
Tom finally chose the factory site in a basin far from any base and railway line, so as to ensure that even the slightest vibration would not be transmitted here;
The factory building of this huge factory adopted the strictest radiation protection measures to strictly prevent interstellar radiation from interfering with the production line;
Every Clone entering the production workshop must wear the strictest dust-proof clothing to prevent any dust from being brought in, and so on and so forth.
The construction of this huge factory, along with the subsequent installation and debugging of equipment, cost Tom a total of one year.
One year may not seem long, but for the highly efficient Tom, it was a rare and long period of time.
The chip factory was finally completed, and it could finally mass-produce advanced chips with a 20-nanometer process.
However, since he had just mastered this technology, the yield rate of the chip factory was temporarily low at this stage, currently only about 3%.
Its total annual production capacity is ten million chips. However, among these, only 300,000 chips meet the requirements, and the remaining 9.7 million chips have various problems, resulting in a decrease in overall performance.
Should these chips be thrown away? Or melted down and remade?
What a joke!
If a certain part of the chip has a defect, just shield that part through algorithms and programs. Can’t the remaining part still be used?
Although the overall performance will decrease, even if it decreases, the comprehensive performance is still higher than even the most advanced 45-nanometer process chips before, so it is still a good thing.
Therefore, Tom used all 300,000 of the highest quality chips from this year’s production to build a new supercomputer, and the remaining 9.7 million chips all flowed into factories and bases, as well as various spaceships and vehicles.
Amidst this continuous production and continuous optimization and adjustment, the overall yield rate is also slowly increasing. It is believed that it will not be long before it can be increased to over 90%.
Now with the supply of more advanced chips, Tom finally built the new supercomputer he had dreamed of.
Compared to the previous generation of supercomputers, this supercomputer uses more chips, a total of 160,000!
Its computing power also reached 130 quadrillion floating-point operations per second, more than 30 times that of the previous generation of supercomputers!
After construction and debugging were completed, Tom eagerly put the combat AI, whose performance was severely limited before, into this supercomputer and resumed combat experiments.
This time, Tom adopted a different combat method than the last time.
His main body also came to the exercise ground, almost without delay, controlling the Alpha fleet. At the same time, he consciously delayed the control of the Beta fleet, giving it a 100-second delay.
But the Beta fleet has a supercomputer spaceship that can receive real-time assistance from the combat AI.
In this way, one side had no combat AI but no delay, and the other side had combat AI but also a delay. Both sides used blank ammunition and simulated weapons to engage in fierce firefights in space.
With the application of the new supercomputer, it was visibly apparent that the Beta fleet showed significant improvements in flexibility, reaction speed, coordination, cooperation ability, and so on.
Various tactics such as forward thrust, formation, salvo, cover, and suppression were performed in turns, while weapons such as bullet barrages, artillery, interstellar missiles, and electromagnetic bombs were used in rotation.
In the end, the Alpha fleet directly controlled by Tom still won this battle.
But the victory was not easy.
Comprehensive evaluation showed that with a delay of 100 seconds, the Beta fleet’s combat power, with delayed control plus combat AI assistance, could reach about 90% of the Alpha fleet’s combat power.
Tom did not stop and conducted another experiment.
This time, he increased the delay of the Beta fleet to 500 seconds.
In the next battle, the Beta fleet showed a clear decline in combat power. The final evaluation showed that its comprehensive combat power was only about 70% of the Alpha fleet’s.
However, if there was no combat AI assistance and he relied entirely on himself to control the fleet with a 500-second delay, its comprehensive combat power would plummet to only about 30%.
The combat AI provided a 40% increase in combat power in this situation!
"A 500-second combat delay corresponds to a combat radius of approximately 75 million kilometers.
From now on, with the assistance of the combat AI, even for spaceships 75 million kilometers away, I can unleash 70% of the fleet’s peak combat power!
That’s enough!"