After getting the source code, Kerry naturally has a way to recompile this huge application. Most of the code of this program is written in C and C++, so in the compilation environment and compiler programs are not very different from other applications. On the first day, Duke figured out what to use for the program to compile, and after returning home, he directly installed a set on Kerry.

The new nuclear reaction simulation system compiled by Kerry will soon be run on the improved virtual Titan ii and the original Titan ii system at the same time, making Duke feel calm.

At this time, Duke began to optimize Kerry's application efficiency, first optimizing the original Titan II system, and then preparing to use Kerry's optimization results before starting to modify it.

This program is really worth optimizing, and it is full of a large number of data computing programs. Because the program is confidential and strict, it is obvious that developers often develop in a conservative way. Although the algorithms made in this way are not prone to errors, there is great room for improvement in efficiency.

So Kerry spent almost no time to come up with a new version with a performance optimization improvement of up to 50%. After multiple verifications on the virtual original Titan II system, Kerry gave a list of all improvements for this new improvement and tested the efficiency of each improvement.

Such thoughtful considerations, if they are spread, they will be enough to make all the hardworking people who are programmers envious.

However, seeing this amazing result, Duke was a little hesitant. With Duke's emotional intelligence at this moment, he would naturally not be so stupid that he would directly modify the result to the new version of the application. Doing so would not enhance the halo of genius on his head, but would directly put on the hat of a monster, so he would never want to walk.

The last time I made a major improvement in the operating system, I actually only improved the scheduling part in the code. It can be said that I have experience in researching parallel algorithms. Although this improvement seems exciting, it is not enough to be regarded as a monster. However, if all the improvements listed by Kerry are shown, it would be too amazing.

These modifications involve quite a lot of algorithms, not one or two, involving at least hundreds of thousands of lines of source code. Duke only studied it for a few months, not only to understand hundreds of thousands of lines of code that he had never seen before, but also to make significant improvements. This performance is really not what humans want.

So Duke would not make such a low-level mistake. But Professor Andrew can explain more perfectly how much improvement it will be. Duke has not communicated with Professor Andrew about this, and I don’t know what the bottom line is in his heart. Considering that the operating system performance increased by fifteen percent last time, everyone was a little excited, and Duke decided to control the proportion of this improvement this time.

It was just that how much it was appropriate to control. Duke was really a headache. He looked carefully at the modification list provided by Kerry and noticed that there were two places with better modifications. The total number of codes was modified by less than 5,000 lines, but the improvement effect reached 11%. Duke was moved and the modifications seemed to be OK. There was not much code involved and the effect was not good. It was better to use this to interpolate.

Thinking of this, Duke began to modify the code on the terminal connected to the nuclear reaction simulation program. In order to reflect the careful effect, it took Duke more than a week to modify these thousands of lines of code.

After the compilation test was passed, Duke asked Kerry to write several test programs for these two pieces of code and pretended to test the effect. Of course, this code that Kerry had strictly verified was already impeccable in terms of algorithms except for Duke's input errors, so there was no pressure to pass these test programs.

However, Duke needs this test result to submit to Professor Andrew. There is no credible verification solution for this critical application. Professor Andrew will never dare to make a random decision. After all, it will be a huge mistake. If the application of the nuclear reaction program loses its accuracy, it will be useless to run too fast.

After completing all this, Duke came to Professor Andrew's office.

"Duke, how about it? Is there any result?" Andrew asked affectionately when he saw Duke coming.

"Professor, there are some results," Duke said. "Based on the code I see now, there may be some problems with the algorithm in the intensive operation part of the nuclear reaction simulation program. I have re-modified it. The files I modified are ***.cpp and xxx.cpp. Please send someone to receive verification."

"Is that so good? I wonder how much efficiency this modification can improve?" Andrew was very happy when he heard that there was progress.

"According to the simulation calculations of the two algorithms before and after, it is estimated that it can increase by about 10%, right?" Duke reported a possible range and looked at Andrew's expression nervously. If Andrew was not satisfied with this number, the work might have to continue. If he was satisfied, he would just retire.

"Ten percent?" Andrew repeated. Although he had high hopes for Duke before, when hope finally became a reality, Professor Andrew seemed a little excited.

It was done. Duke saw Andrew's expression and knew that he could quietly retreat. It was just right to withdraw. On the one hand, the short time could make Professor Andrew and those who pay attention to Duke feel at ease, because according to general reasoning, Duke could not see much in such a short time. On the other hand, Duke really didn't want to improve the performance of this system anymore.

Because maybe this system is a major stumbling block in Duke's future energy plan. After Kerry copied this thing to the virtually improved Titan II system, Duke entered the specific parameters of the natural reactor, and has verified that those parameters are completely free of problems. The simulation experiment results are very perfect.

Duke is planning to use the MIT nuclear laboratory to obtain most of these results, and then leave some tails as his own unique weapon before it is announced after leaving. If Duke improves the performance of this system too high, if he joins top experts such as MIT nuclear experiment to study this topic, Duke cannot guarantee that they will not have a big explosion in character and discover the part of the things he has hidden in advance under an accidental condition.

After all, on this issue, I am in a difficult situation. On the one hand, I have to break through the results to the point where I can break through with "luck" in the future, and on the other hand, I cannot expose the fact that I have already obtained the entire nuclear reaction simulation program.

If the results that must be obtained through this nuclear reaction simulation program cannot be "discovered", Duke will lack a reasonable source of results. It is equivalent to providing a very obvious flaw to those who are interested. This is an extremely dangerous thing for Duke, and Duke must avoid it.

(Chapter 200, more than 500,000 words in 5 months. As an office worker, it is not easy to be honest. I’ll collect it and ask for comfort!

Dear book friends who clicked in from various lists, please click the button "Add Bookmark" on the right;

Or right-click and select "Add Bookshelf Bookmark" in the middle of the pop-up menu to facilitate reading next time;

The character is guaranteed to be more exciting and interesting in the future. You don’t have to spend a penny, so accompany Duke to change the world)
Chapter completed!