New Employees With an Annual Salary of 1 Trillion Won - Chapter 181

—————
This chapter was translated by Lunox Team. To support us and help keep this series going, visit our website: LunoxScans.com
—————

Chapter 181. Starting Line (1)

Professor Hinton had returned to his accommodation first.

But Jenson Huang and I remained in the lab room to continue our remaining conversation.

To be more precise, when a new application for GPUs emerged, Jenson Huang wanted to cancel all his schedules and continue the conversation.

“Can GPUs really be used for artificial intelligence development?”

“We’ve already confirmed it through the rendering system. It’s definitely usable for artificial neural network development. Of course, we’d need to create and integrate an intermediate system capable of general-purpose parallel computing.”

“I can’t believe Tiger Fund possesses such system development capabilities.”

It was natural for him to find it hard to believe.

Tiger Fund was a financial company, after all.

But Tiger Fund Korea Branch was different.

It was closer to a massive IT company than a financial firm, possessing semiconductors, games, internet platforms, and more.

It was almost the only place that had its own IT ecosystem.

“We’ve deployed over 200 developers just to create the compatibility system. And we’ve broken dozens of GPUs during development.”

“If such a system can be created by breaking just dozens of units, I’d conduct such experiments countless times! That’s truly amazing.”

“But there’s still a problem remaining. To fully utilize GPUs for artificial intelligence development, current GPU designs have limitations.”

“You’re saying we need GPUs dedicated to artificial intelligence.”

Jenson Huang’s voice dropped a tone lower.

The current NVIDIA didn’t have the resources to design new dedicated GPUs.

The CEO had to personally go down to Yongsan Electronics Market to do sales, so you could understand their desperation without needing to say more.

“Tiger Fund plans to cover all development costs.”

“Really? …Then that changes everything. But even if we make AI-dedicated GPUs, would there be a market to use them?”

“There might not be demand right now. But the moment artificial intelligence becomes commercialized, demand will explode.”

“How many years would that take?”

I couldn’t guarantee the timeframe.

But we all knew.

That AI commercialization wouldn’t happen in the short term.

“I’m thinking of it as a long-term project.”

“That’s the problem. NVIDIA doesn’t have the resources to endure a long-term project.”

“I’ll help you with that part.”

“Are you talking about investment? We’ve already sold many shares to Tiger Fund. There aren’t any more shares left to sell.”

Tiger Fund was already one of NVIDIA’s major shareholders.

After the IT bubble burst, we swept up falling IT company stocks.

We actively invested in equity shares of companies we judged promising.

“I’m thinking of helping in a different way. You’ll directly feel how serious Tiger Fund is about artificial intelligence development.”

“Could you tell me specifically what kind of help?”

“I’m thinking of directly creating a market for NVIDIA.”

“A market?”

There were several ways to handle the overflowing GPU inventory.

I had already prepared the most efficient method to use large quantities of GPUs while minimizing losses.

“I’m thinking of opening a PC Bang chain nationwide. At least 50 locations. Massive stores with 200 PCs each.”

“You’re saying you’ll equip all those PCs with NVIDIA GPUs?”

“That’s right. Just that alone would handle over 10,000 GPUs in a short time.”

A PC Bang craze was sweeping nationwide.

So even opening 50 large new stores wouldn’t result in big losses.

Rather, it was a solid business item that could quickly recover initial investment.

Of course, it wasn’t a scale I’d operate directly.

I planned to proceed with the chain business through Redline or other subsidiary companies.

“Certainly, if over 10,000 GPUs are sold, almost all inventory in Korea would be consumed.

But… that alone wouldn’t be enough.”

“I heard worldwide inventory exceeds 300,000 units.”

“Embarrassingly, that’s true. Of course, I don’t plan to process everything short-term. Still, if demand for 10,000 units is created in Korea alone, it would be a huge help.”

GPUs were overflowing.

Unless more people needed GPUs, there was absolutely no way to dispose of the inventory.

“So I thought of one more method. A way to increase the necessity of GPUs.”

“Is there such a method?”

“Tiger Fund has game companies under it. And there are several Korean game companies we have cooperative relationships with.”

The best way to help NVIDIA was separate.

We could just buy all the inventory with Tiger Fund’s funds.

But doing so would kill NVIDIA’s self-sustainability.

We needed to create an ecosystem where they could survive on their own.

“Are you telling us to collaborate with Korean game companies?”

“To increase GPU sales, there need to be more high-spec games that require GPUs. And such games need to become popular.”

“Would that be possible short-term?”

“Korean game companies have already prepared high-spec games. There are quite many works about to be released.”

Tiger Fund already had close relationships with several Korean game companies.

We provided revenue models that could increase sales several times over at once.

With the funds secured that way, Korean game companies were developing high-spec graphics engines.

“What form of collaboration should we have with them?”

“It’s simple. Just ensure compatibility. If games run more smoothly on NVIDIA GPUs than competitor GPUs, gamers will naturally choose NVIDIA.”

Compatibility issues ultimately required time and effort.

If game companies and GPU companies cooperated closely from before release, compatibility would naturally improve.

“Do you mean Tiger Fund will become the connecting link with Korean game companies?”

“That’s right. I’ll directly connect you with Korea’s largest game companies. If needed, I’ll also take on the role of intermediate coordinator.”

“If you could do that, it would be a tremendous help!”

Jenson Huang clenched his fists and raised his voice.

He looked like someone who had finally found an exit in endless darkness.

But reaching that exit was still far away.

“At the same time, we need to solve fundamental problems.”

“What problems are you referring to?”

“Severe heat and noise issues compared to competitors. Gamers are hesitating about NVIDIA GPUs because of those concerns.”

“…We’re doing our best to solve those problems, but it’s not easy to fix them short-term.”

Heat and noise were chronic problems for all electronic devices.

Fortunately, Nexpin had several related patents, and we had already accumulated sufficient know-how during small HDD development.

“If we apply the cooling structure and vibration suppression technology used in small HDDs to GPUs, we can create much more stable designs than now, even if not perfect.”

“You possess such technology too?”

“Of course. Though HDDs and GPUs have different structures, the principles are the same. If we deploy all our personnel, we can greatly shorten the application period.”

Having patents didn’t mean problems could be fixed immediately.

But if I looked into it directly, the story would be different.

I could quickly identify the cause of problems and clearly envision how to fix them.

That was my area of expertise.

Numerous patents were born from solving such problems.

“With such help, I don’t know how to express my gratitude.”

“Tiger Fund considers NVIDIA not just an investment target, but a true partner. I hope our sincerity comes through.”

“I feel it sufficiently. I think I finally understand how deeply Tiger Fund has been considering NVIDIA.”

NVIDIA was still too small to be called a large corporation.

Since Tiger Fund put so much effort into such a company, Jenson Huang couldn’t help but be moved.

I had reasons for going this far too.

NVIDIA absolutely had to take the central axis of the massive structure I envisioned.

I had simulated whether other GPU companies could substitute, but the moment the central axis changed, the entire structure collapsed.

For my sake too,

Nvidia had to grow rapidly without fail.

That’s why all our current efforts were necessary.

“Tiger Fund will wholeheartedly help Nvidia’s growth. So please feel free to tell us anytime if there’s anything lacking.”

“I’m truly grateful that you say that, Branch Manager.”

“We’ll begin GPU modification work starting tomorrow. And once all the modification directions are decided, we can readjust all of Nvidia’s GPUs at the small factory that Tiger Fund owns.”

“By small factory… do you mean the factory that produces iPods?”

I nodded.

Finding a factory capable of modifying 300,000 GPUs was never easy.

Of course, reusing the factory where they were originally produced would be most ideal, but the procedures were complex and would take too much time.

“The iPod Factory has high precision. It also has sufficient experience with heat treatment and noise control. If we adjust the facilities to GPU-dedicated processes, we can complete all modifications within one to two months at the earliest.”

“Thank you so much! If we can do that, I’ll do anything!”

Jenson Huang repeatedly expressed his gratitude.

We shared a firm handshake and stood up from our seats for a new future.

***

Professor Hinton’s contemplation didn’t last very long.

Whether he had been lost in thought all night, he came looking for me early in the morning with a haggard appearance.

I invited him to the Rollbook Office and opened up everything, showing him first how prepared we were.

“Rollbook’s scale is larger than I thought. I didn’t know the Development Team alone had over 300 people.”

“This is just the scale of our Korea personnel. If you add the United States and global staff numbers, it’s more than triple.”

“Then will Rollbook’s Development Team personnel also participate in artificial intelligence development?”

“That’s the plan. Of course, Professor Hinton’s team must continue the existing research, but the current scale is too small.”

Professor Hinton’s team had fewer than 10 people.

Since artificial intelligence as a field of study was treated as dead research, it was difficult to find research personnel, and naturally they couldn’t receive investment, so they couldn’t expand their scale.

“Can Tiger Fund also provide support for artificial intelligence research funds?”

“Of course. Tiger Fund will bear the full amount.”

“Could I know the exact amount, by any chance?”

Professor Hinton asked with slight hesitation.

He seemed like the type who couldn’t easily bring up money talk as a scholar.

I could also guess how difficult it had been due to lack of research funding.

So I wanted to make it clear that he would never have to worry about money again.

“We’ll provide 50 million dollars as initial research costs.”

“Is it possible to invest that much?”

“50 million dollars is just for initial costs. We only set it at 50 million dollars because the specific plan isn’t complete yet. Once the plan is finished, investment of over 500 million dollars per year is also possible.”

Professor Hinton couldn’t close his mouth.

It was an amount that exceeded typical university research funding.

50 million dollars was similar to major pharmaceutical company R&D amounts or national research funding levels.

“Where are you telling me to use that much money?”

“First, we need to expand the scale. And we even have plans to build an artificial intelligence-dedicated data center.”

“I don’t know if I should receive such support. And Korea and Canada are so far apart that I’m worried whether proper communication will be possible.”

Actually, I was worried about that part too.

I wanted to frequently check on artificial intelligence research, but if the research proceeded in Canada, it would be difficult to grasp the situation in detail.

“So may I make one proposal?”

“What kind of proposal?”

“I want to newly establish an artificial intelligence-dedicated Research Institute in Korea.”

“Are you saying to move to Korea?”

“That’s right. I’ll guarantee double your current Annual Salary. I’ll promise the same level of conditions to all researchers as well.”

Professor Hinton fell into contemplation once again.

But he knew well that there was no other choice for artificial intelligence development.

“Whew, alright. I’ll trust Mister Lee and start over from the beginning in Korea.”

“I look forward to working with you!”

Professor Hinton also took my hand in the end.

Now we could begin full-scale artificial intelligence development in Korea.

—————
This chapter was translated by Lunox Team. To support us and help keep this series going, visit our website: LunoxScans.com
—————