My mentor is dishonest?

Three days later, there were five days left until school started.

To Lin Yuan's surprise, the two H100s provided by OCR were high-end models, not entry-level ones.

This was because Lin Yuan's algorithm draft was completed on the third day of training.

For AI training, as long as the AI model is well-designed and the computing power is sufficient, you just need to tie a dog next to it.

When Lin Yuan relied on skillfully using the system to solve the AI model building problem and then freeloaded computing equipment from OCR, everything was essentially accomplished. It's like when you hold a gun in your hand, you don't necessarily have to shoot someone to prove its power.

Everything came so naturally.

Lin Yuan originally intended to submit the algorithm draft, along with the simulated test results of this algorithm, to the head of the technical department at the Good Group Zijin Branch.

Before submitting, due to Professor Zhang's special reminder, Lin Yuan first sent the algorithm submission email to his mentor. It was his mentor who personally forwarded it to Good Group.

The email contained two main things: a PDF document outlining the AI model used, recording key nodes during the training process, and presenting the evaluation results; and the core Python code for the AI model.

The specific data used for training and the algorithm entity trained by the AI model could not be included as attachments due to their massive size, over a hundred gigabytes (mainly because the training sample data was too large).

However, a qualified professional only needs to look at the data in that PDF document to understand the value of Lin Yuan's algorithm. As for the hundred-plus gigabytes of data, they are used to prove that everything written in the PDF document is true. This can naturally be handled later.

As for why it had to be submitted through his mentor, Lin Yuan hadn't thought of this at first. The older the ginger, the spicier it is. Presumably, his mentor had seen enough workplace infighting and was afraid Lin Yuan would be taken advantage of.

With his mentor stepping in, naturally, no one would dare to act presumptuously.

— —

Lin Yuan had great confidence in the system; he was certain that the algorithm he submitted would attract Good Group's attention. All he needed to do now was wait.

After that, he could say goodbye to OCR!

He'd delete that application specialist's WeChat, and he'd pretend this whole thing never happened.

However, for some unknown reason, his mentor called Lin Yuan specifically on the day the algorithm was submitted. After they discussed the most important algorithm submission matters, his mentor suddenly brought up the issue of OCR's computing equipment.

“Lin Yuan, what do you plan to do with the computing equipment from OCR?”

“Huh? I want to cut ties with them from now on.”

“Cut ties?”

Hearing his mentor's doubtful tone, Lin Yuan tentatively said, “I don't have any evidence to accuse OCR, otherwise I would definitely report it.”

In fact, even if Lin Yuan had evidence, he wouldn't want to get involved in such a messy situation.

“Report what?”

“So, Professor, you mean... if I can't report it, I should just destroy it? Give it a virus to paralyze it?”

“Think outside the box.”

Lin Yuan couldn't think of how to think outside the box for a moment.

“Think about what I did when I saw those caviar and black truffles at the tea break back then.”

“I wasn't even born then, how would I know what you did?”

“To use your young people's words, can't you imagine?”

“Eat?”

Professor Zhang answered himself, “Eat and take. If it's hard to take, secretly lick a bit.”

“Uh...” Lin Yuan was a bit... 'Is this something an intellectual would do?'

“So, you understand now, right?”

“I... what do I understand? Learn from you, eat and take? And then lick a bit?”

“Exactly.”

“But how do I take it... I... I'll use it for mining?”

Lin Yuan suddenly realized, “That's right, why don't I use it for mining?”

“Exactly!”

After a brief moment of excitement, Lin Yuan immediately realized a problem, “Two H100s are powerful, but at most, they can only mine for a few days. It won't be much money, will it?”

“Cough, cough~~” A cough came from the other end of the phone, and then Professor Zhang continued, “Well, if you have to pay for the electricity yourself, it really won't be much money.”

“I understand, Professor. Someone will pay for my electricity.” Lin Yuan loved his unconventional mentor; Professor Zhang truly taught him by example and didn't hold anything back. Lin Yuan was so moved that tears welled up in his eyes. It must be that his mentor knew he was too poor and, out of consideration for his self-esteem, reminded him of this way to make a fortune.

“You just focus on your mining. As for the algorithm, as long as your algorithm is fine, you won't lose out on what's rightfully yours. I'll handle things with Good Group for you.”

The most expensive part of mining is not the equipment, because no matter how expensive the equipment is, it's a one-time investment, and capital assets themselves hold value. The most expensive part is actually the electricity cost.

Lin Yuan roughly calculated that with the computing power of two high-version H100s, he should be able to earn a thousand yuan or so per day, right?

He had checked the price of a high-version H100, which was around 300,000 yuan. Two of them combined would be 600,000 yuan worth of equipment. So, without considering electricity costs, earning a thousand yuan should be possible. Otherwise, what fool would be willing to take the risk to mine?

— —

And so, Lin Yuan was finally able to embark on his grand mining plan.

What he sought with great effort was found effortlessly. After failing to mine with the system, Lin Yuan finally found a new path with his mentor's reminder.

‘Xia Ye, we have the milk money for our child.’

— —

Lin Yuan hadn't expected that after submitting the algorithm draft, his next task would be: figuring out how to use OCR's computing equipment for mining.

But as happy as he was, problems also arose. Because, damn it, mining isn't something you can just do whenever you want.

Mining is a competition of computing power, and computation cannot exist independently. The CUDA computing cores within a GPU are what enable its computing power; they are like factories that need input to produce. And the input to be processed is data.

The production capacity of a factory depends not only on the operating speed of the machines inside but also on the speed of transportation—how quickly raw materials are transported in and out.

Therefore, among the important performance indicators of a GPU, in addition to the main frequency, there is also the bandwidth on the internal memory bus.

This bandwidth represents the speed at which the GPU's internal factory transports things.

And to ensure sufficient bandwidth, there are certain requirements for the GPU's own video memory. Video memory represents the total amount of data you can cache, which can be simply understood as the capacity of a warehouse in a factory.

It is precisely because of this that mining has a minimum requirement for GPU video memory.

Because the data needed for mining is the “ledger,” and the ledger must be loaded into the video memory.

Lin Yuan was too lazy to care about the data structure of the ledger and its underlying principles; he cared about: once the ledger is loaded into the GPU's video memory, it will be very easy for OCR to scan out its data features. This would then reveal that he is mining.

The ledger of each digital currency has its own data characteristics, because all data in this world has its characteristics; data without characteristics is just a jumble of random codes.

Various security software can tell what you are mining just by looking at the data features.

So, the problem Lin Yuan faced now was: how to mine under OCR's nose.

Even if he only earned a thousand yuan a day, that would be seven thousand yuan in seven days. This money shouldn't be passed up.