Chapter 24: Curiousity

The moment he returned home to his house in Tondo, Timothy went straight to his bedroom and lay there staring at the ceiling.His mind was still fixated on the business he wanted to start. The Buy and Sell business was doing good and he had been profiting a good amount of money out of it.

Just like before, it wasn’t enough.

"Hmm...I know I thought about this before but what if I just sold advanced technologies to a billion-dollar company?"

He opened the browser on his Iphone and searched specifically for AI. According to it, there was an AI race going on and the market is in the billions. And what does a company need to train AI?

He searched further.

Page after page, article after article. Reports from Bloomberg, Wired, even Reddit threads filled his screen. Every one of them seemed to repeat the same theme: computing power.

It wasn’t just about data. It wasn’t even about the algorithms. Everyone had access to similar models, similar code. What separated the winners from the losers in the AI race was simple: the hardware to train them.

Timothy sat up, brow furrowing.

"GPUs?" he muttered aloud.

Graphics Processing Units. Originally designed to render video games smoother, now they were the backbone of modern AI. The articles explained how one company—NVIDIA—was practically holding the world by the throat. Their chips weren’t just powerful, they were essential. Training a large language model or an image recognition system required thousands of high-end GPUs working in parallel.

The price tags were obscene. Each top-tier GPU cost thousands of dollars, and companies ordered them by the hundreds, sometimes thousands. One line caught Timothy’s attention like a flashing siren:

"AI companies without access to state-of-the-art GPUs face years of disadvantage compared to competitors who secure them early."

He leaned back, his mind racing.

"So that’s the bottleneck," he whispered. "Not the software, not the hype. The real power is in the chips."

He kept scrolling. His eyes widened at the numbers. NVIDIA’s market cap had exploded past a trillion dollars. Competitors like AMD and Intel were scrambling to catch up. Entire governments were throwing billions into semiconductor fabs, desperate not to be left behind.

Timothy’s pulse quickened.

If the Reconstruction System could create something more advanced than even the best GPUs—something beyond 2024 tech—he wouldn’t just have a business. He’d hold the keys to the next industrial revolution.

But a pit formed in his stomach at the same time.

Selling brain pills was risky. Selling flood-free cars was manageable. But selling hardware that outpaced NVIDIA? That was like painting a target on his back. Governments would come. Corporations would come. Some would offer checks, others would send men in the dark.

Still, he couldn’t shake the thought.

"What if I reconstructed a GPU into something twenty years ahead?" he muttered. "How much would that be worth?"

The idea pulsed in his skull, impossible to ignore.

"Let’s try it. But I need a GPU... perhaps I should go to an internet café and purchase one..." Timothy muttered.

The thought alone felt surreal. The same cafés where he once spent his nights grinding Dota matches or typing papers for school were now his hunting ground for something far bigger.

The next day, Timothy found himself walking down a familiar Manila side street, neon signs flickering above rows of cramped computer shops. The air smelled faintly of dust, fried food, and cigarette smoke. A faded tarp banner read PC Express in big red letters. Inside, shelves groaned under the weight of motherboards, RAM sticks, and a few GPUs locked behind glass.

"Can I help you, boss?" the shopkeeper asked, chewing gum lazily.

Timothy scanned the glass case. Most of the cards were mid-range—GTX 1060s, RTX 2060s, a lone RTX 3060 looking almost proud under the fluorescent light. Nothing like the monsters NVIDIA was selling to corporations, but enough to serve as raw material.

"That one," Timothy said, pointing to the RTX 2060. "Second-hand if you got one."

The shopkeeper raised a brow. "Planning to play Cyberpunk?"

Timothy forced a faint smile. "Something like that."

Minutes later, he walked out with a used 2060 tucked in a box under his arm, its warranty long expired. He paid ₱8,000—barely a dent in his new fortune, but possibly the start of something bigger than the entire dealership.

Back in his Tondo bedroom, he placed the GPU on his desk, still smelling faintly of dust and solder. The fan on the card looked worn, but the System didn’t care about scratches.

He placed his hand on the casing, heart thudding.

"System," he said slowly, "reconstruct this graphics card. Transform it into a state-of-the-art GPU designed for artificial intelligence training and supercomputing. It must remain electrically compatible with today’s computers but contain technology twenty years beyond 2024 standards.

"Implement fabrication at 2-nanometer scale or smaller. Integrate stacked HBM memory directly on-die for maximum bandwidth. Optimize for energy efficiency and cooling using futuristic materials. Deliver exaflop-level processing power per card.

"And finally..." Timothy’s voice steadied, "...generate a blueprint in printed paper form beside it—schematics, diagrams, full engineering documentation—so that I may understand exactly how this unit is designed."

The GPU glowed under his hand, the faint blue shimmer deepening until it filled the room with an almost electric hum. The old casing dissolved, traces of circuitry knitting themselves into impossible complexity. The air buzzed with the faint smell of ozone.

When the light faded, the battered RTX 2060 was gone.

In its place lay something sleek and alien—black metal with sharp, polished edges, its surface etched with faint glowing lines that pulsed like veins. Letters appeared across the shroud, softly illuminated:

NeuralCore NX-1 — Quantum-Optimized AI Processor.

And next to it, fresh sheets of paper slid onto the desk as though pushed by invisible hands. He picked one up, his heart hammering.

Blueprints. Full schematics. Node diagrams, transistor layouts, exotic materials he had never even heard of. Every detail, printed in neat, legible lines.

"This is it!"