Development of the Next-Generation CPU

A week later, the atmosphere in the Hardware Development Department was somewhat somber.

Inside Minister Nakamura's office, Nakayama Takuya, Yuji Naka, and Nakamura sat around a conference table piled high with technical documents.

"Executive Director, this is our evaluation report on the 32-bit chip solutions from several major semiconductor manufacturers in Japan over the past week."

Nakamura Head Nakamura adjusted his glasses, placed a summary document in the center of the table, and his expression was not pleasant.

"And the conclusion?" Nakayama Takuya asked bluntly.

"Not ideal." Minister Nakamura stated succinctly, "Hitachi, NC—their products either have overly complex instruction sets unsuitable for game development, or they are designed for specific industrial sectors, lacking versatility. Customizing based on them would result in exorbitant development difficulty and costs, which contradicts your core requirement of 'ease of use.'"

Yuji Naka added from the side, "I've read the development manuals for those chips, and it's a nightmare. If a third party were to use this kind of thing to make a game, it would probably take half a year just to figure out how the chip behaves, and by then it would be too late."

As a one-threaded designer, he is the most qualified to speak on this matter.

During the MD era, he dealt extensively with those quirky underlying hardware logics and knew firsthand what a simple and efficient development environment meant for game developers.

Nakayama Takuya nodded; this outcome was within his expectations.

While Japan's semiconductor industry is strong, it did start late in developing RISC (Reduced Instruction Set Computing) architecture, and its approach was relatively conservative.

"If domestic options don't work, let's look at foreign ones. Are there any alternative plans?" He looked at Minister Nakamura.

"Yes." Nakamura pulled out another document. "We compared all the mainstream RISC chips on the market, and after comprehensive evaluation, the R3000 architecture launched by MIPS United States 1988 is the most suitable choice."

He forwarded the documents to Nakayama Takuya and Yuji Naka.

"The technology is very mature, has been tested in the market for many years, and is extremely stable. Most importantly, its instruction set is quite concise, perfectly meeting our requirements for ease of use. Customization based on this foundation carries the least risk and is the easiest for developers to get started."

Yuji Naka leaned over to take a look at the technical specifications, his eyes lighting up. "That's it, that's the one."

Seeing that the two core technical personnel had reached a consensus, Nakayama Takuya smiled.

Isn't this the same path that Sony PlayStation took in its previous life? Now, not only have they stolen their controller design, but they've also snatched the CPU solution along with it.

"Great minds think alike," Nakayama Takuya, picking up his teacup, blowing gently on it, "Since we both think it's good, then it's settled."

He put down his teacup, tapped his fingers lightly on the table, and made his final decision.

"I recall that MIPS was just acquired by SGI (Silicon Valley Graphics) this year, right?"

"Yes, Executive Director." Director Nakamura nodded in confirmation.

"That makes things easier." Nakayama Takuya's lips curled into a smile. "Have the United States branch contact SGI immediately. Tell them that Sega wants to use the R3000 as a blueprint to customize a custom heart for our next-generation console."

He paused, a hint of amusement flashing in his eyes.

"Oh, and by the way, a quick reminder to SGI: this is a huge, exclusive order. Make sure they offer a competitive price."

"I understand, Executive Director!" Nakamura nodded solemnly.

With the matter settled, Nakayama Takuya stood up and stretched.

He wondered to himself how far Sony had progressed with their next-generation console project.

Hopefully, they haven't set their sights on MIPS yet.

Otherwise, their expressions would be priceless when they excitedly took their proposals to SGI, only to find that SGI was already customizing CPUs for Sega.

Let them rack their brains and see where they can conjure up a suitable CPU.

Just as Nakayama Takuya was about to get up and leave, Department Head Nakamura, who had been engrossed in taking notes in his notebook, suddenly looked up.

"Executive, there's one more question."

"Speak." Nakayama Takuya stopped in his tracks.

Nakamura Head Nakamura adjusted his glasses and pointed seriously to the documents on the table. "We've set the direction for the CPU, but there's still a fundamental issue that needs clarification. Since next-generation consoles will fully adopt CD-ROMs as the game medium, what solution will we use to store players' game records?"

Upon hearing this question, even Yuji Naka, who had been excited about the CPU architecture, calmed down and looked at Nakayama Takuya.

This is indeed an unavoidable hurdle.

In the cartridge era, game saves could be stored directly on the cartridge, which was simple and convenient.

However, CD-ROMs are read-only media and cannot be written to.

Although Sega already had experience with the SEGA-CD external memory card, it was ultimately an accessory, and the experience wasn't perfect.

In an instant, the image of a USB flash drive flashed through Nakayama Takuya's mind.

But he quickly dismissed the idea; that thing wouldn't be available for several more years, and bringing it up now would be like performing magic.

His gaze swept over the two tech experts, and after a moment's thought, he gave a seemingly conservative answer.

"The memory storage solution follows a similar approach to SEGA-CD, using a flash memory chip paired with a memory card with a built-in battery."

Upon hearing this answer, Nakamura and Yuji Naka nodded. This was the most mature and reliable solution at present, and it was within their expectations.

But Nakayama Takuya hadn't finished speaking.

"However, some optimizations are needed." He tapped his finger lightly on the table. "We can't just focus on storing the records; we also need to consider how to store them for a long time and ensure their security."

"First, add a counter to the memory card's chip to specifically record the number of read and write operations."

"A counter?" Minister Nakamura was taken aback; this was a completely new concept.

"That's right." Nakayama Takuya's tone left no room for argument. "Each read/write operation consumes a significant amount of the built-in battery. Although we don't know the exact number, there must be an upper limit. Nakamura, your Hardware Department needs to conduct extreme testing immediately and give me a rough estimate of a safe range for the number of read/write operations. Then, we'll build upon that and allow for sufficient redundancy."

He looked at Yuji Naka: "Then, you need to consider adding a clear prompt to the system software so that players can see it when managing their save data. For example, when a player enters the console's memory management interface, if the read/write cycles of a memory card inserted in the console are about to run out, a warning pops up on the screen requiring manual confirmation from the player, reminding them that the memory card is about to reach the end of its lifespan and to back up their save data to a new card as soon as possible. Or, if the memory card has exceeded its safe read/write cycle limit, it reminds the player that the memory card has reached its lifespan and to back up their save data to a new card as soon as possible, and that the memory card no longer supports game memory writing, and locks the write function of that memory card."