Now that the requirements have been identified, we can finally officially launch the project.
However, the first step wasn't taken by the development department, but by the newly established NE division.
We need a new engine.
Just as Need for Speed requires a dedicated engine, Battlefield naturally needs its own engine as well.
"An engine specifically designed for Battlefield?"
"That's right, I don't need an all-purpose engine that can run any game, Carmack. What we need is ultimate performance."
In terms of general applications, the Quake engine, which is still being iterated, is sufficient for now.
To realize the ideal work, these existing engines are simply not enough.
"That's so cool! How about we call it... Battlefield Engine?"
Carmack agreed almost without hesitation.
He was an extreme, even obsessive, engine enthusiast and technologist.
Lin Lixin's idea hit him right on the mark: to take the optimization strategies discussed in the brainstorming session to their extreme.
They might actually create the world's first truly large-scale battlefield game!
"Hmm... let's call him Frostbite."
"Huh? Frostbite? Why?"
"Because it's so cool!"
"... f**k! You're right! Lin, you're right!"
Development of the Frostbite Engine has officially begun.
Lin Lixin, on the other hand, found himself with some free time.
It's currently the Christmas season, and there's still some time before the Gulf War actually begins.
He had ample time to plan what would happen next.
Operation Desert Shield continues to advance. The atmosphere of war seems to be gradually intensifying.
By this point, everyone had a general idea in mind.
War may really be on the horizon.
Maybe tomorrow? Maybe next month?
No one knows the exact date, but one thing is certain: war will break out.
"Rumble, rumble..."
The television set emitted the heavy sound of tracks crushing metal plates.
An M1 Abrams main battle tank slowly drove off the tail of Lockheed C-5 Galaxy transport plane.
The soldiers standing to the side looked as insignificant as shrimps beneath these two steel behemoths.
"It is now 9:37 AM local time. This is a live report from CNN. You are watching a glimpse into what is arguably the largest overseas military deployment in United States. history, or even in human history."
Wearing noise-canceling headphones, the reporter diligently reported the news amidst the strong winds.
"This transport plane, which is nearly 75 meters long, can carry the United States's greatest heavy equipment directly across the ocean to the other side of the world." The camera zooms in on the M1 Abrams, still roaring with the screeching sound of its gas turbine engines.
"Click-clack-click-clack..."
The steel behemoth, weighing tens of tons, slowly drove out of the cabin and onto the wide runway.
Lin Lixin watched the television program with great interest.
This is all firsthand information!
I need to get the original footage from CNN later.
As a Lockheed shareholder, not a Chrysler shareholder, obtaining the exterior design data was simply unrealistic.
Even Lockheed wouldn't entrust him with such valuable and crucial data.
Let alone a major shareholder, even the chairman of the board couldn't be one.
To achieve the most accurate recreation possible, we can only rely on these photo assets for modeling.
The camera zooms out.
Beyond the spacious airport stretched a vast, golden desert, seemingly endless. "Phew—hey? Why are you watching the news? CNN?"
Sam pushed open the door and entered. Before he could even catch his breath, he noticed the screen on the television.
"Didn't I tell you? I'm making a shooting game based on the Gulf War, so I at least need to know what's going on over there, right?"
" Battlefield."
Sam nodded knowingly; he had already read Lin Lixin's proposal.
"It looks like a game specifically for internet cafes. With such a large-scale online battle, it must require ADSL, right?"
He's been busy with network issues at Nexus these past few days, so he's quite sensitive to this kind of thing.
Lin Lixin said something nonsensical, like, "Someone named Sam should connect the whole world!"
But he was happy with it.
After all, the more popular it is online, the more it proves Gamenova's current prosperity.
Lin spends his days running around doing development and negotiating partnerships, and sometimes he has to do these miscellaneous tasks himself.
"I'm afraid so." Lin Lixin nodded.
The network issue is entirely independent of rendering and is a different, more complex problem altogether.
According to the current plan, it's 16 vs. 16. If it were done according to the design of " Quake ", it would be an absolute disaster.
For large-scale data exchange involving 32 people, approximately how much bandwidth is needed?
Let's do a simple calculation. Assuming the server synchronizes 20 times per second, that's 20 ticks.
This value represents the maximum compression possible while still ensuring a smooth gaming experience.
The higher the tick count, the more rigorous the game's calculations.
This is also the main reason why many professional matches use a much higher tick count than casual games.
With 20 ticks, each player's state synchronization alone consumes tens of bytes.
20 times per second, that's thousands of bytes per second.
This number needs to be multiplied by the number of people in the game, which is 32.
This already reaches bandwidths of tens of kilobytes per second. Traditional broadband connections completely fail at this point; only ADSL can possibly meet this requirement.
but...
Tens of kb/s is the speed of one person.
Although it contains data for all 32 players, there is a small problem.
The server needs to send the same data packet to each of the 32 people!
That's... almost a 1MB/s bandwidth requirement!
Keep in mind that even an 8Mbps broadband connection directly connected to the server room only has a bandwidth limit of 900kb/s.
The current mainstream commercial ADSL bandwidth needs to be halved and then folded back down from this figure.
100kb is already considered high-speed broadband.
The most frightening thing is that this bandwidth requirement only includes state synchronization, not extra things like bullets, vehicles, and special effects.
Even this simple calculation is enough to make one feel despair.
Is it really possible for such small water pipes to support 16v16? Sam clearly knows this too.
"Connecting within the internet cafe's local area network should be fine, but setting up an internet connection would be difficult."
"It's hard to say."
Lin Lixin's lips curled up.
"This problem would have existed last week, but it's different now."
The Frostbite Engine is currently under full-scale development.
Unlike other engines, its biggest advantage is its field of view management.
Or rather... the management of "interests".
Looking at later generations, large-scale battlefield games are actually nothing new.
Not to mention 32 players competing on the same stage, PUBG has even achieved 100 players in a single game.
Games like "DNF" and " Fantasy Westward Journey " have long since achieved the ability to have tons of players on screen simultaneously.
Besides the increased bandwidth of the new era, we also benefit from an important concept: interest management.
" Sam, soldiers on the battlefield don't need to actually 'see' the entire battlefield; this isn't an RTS game."
"Uh... how should I put it?"
"The larger the battlefield, the smaller the bandwidth consumption... is actually. If NE can really create an ultra-massive map, bandwidth won't be an issue."
Lin Lixin explained,
" Quake is a typical arena-style game where several players compete in a confined space, but Battlefield is not."
Interest management, in essence, is the core concept of modern multiplayer game network synchronization technology.
Simply put, it means only sending players the necessary content.
He casually pulled out a piece of scrap paper and drew a dot in the lower left corner.
"Look, Sam. If our characters are here, and the enemy is..."
As Lin Lixin spoke, he drew another cross in the upper right corner.
"The enemy is on the other side of the map, and there's a fierce firefight breaking out here right now. Do you think... we really need to know what's going on there?" The answer is obvious.
No, it's not necessary.
Knowing the outcome of the battle is enough.
Who died and who lived—that's the information that truly matters to this player.
As for the entire battle?
Player A displayed astonishing combat prowess, perfectly dodging numerous bullets with skillful maneuvering, and then used a grenade to directly charge and trade armored vehicles with Player B.
Please, I can't render it anyway, why are you telling me this?
I know that players A and B are both dead, isn't that enough?
"Ah... is it filtering? It only actively pulls and syncs data after entering the area of interest?"
"Yes, but this is just a simple model. How to design the actual AIO interest domain will depend on the subsequent algorithm design."
Lin Lixin tucked the platinum pen back into his chest.
The model has a Level of Degree (LOD), and the network world also has a LOD. The lower the interest of a unit, the lower its tick count can be.
Battlefield is the perfect testing ground for interest management.
If they can master this technology, they can then try to create even more extravagant works.
"Wait, won't that increase the host's performance overhead? The server needs to calculate too much."
"That's why we need a dedicated server."
"A dedicated workstation is used to run the server?"
"Yes."
This isn't the first time Gamenova has done something like this.
In the early days of video games, playing online was actually quite simple.
The host uses their own client to start a server, and others can connect through some method.
The advantages are simple development, no need for additional maintenance, low cost, no extra expenses, and out-of-the-box usability.
But the problem is also obvious: the host has a magic weapon! The host has zero latency, so even if their aim is a little bad, they can still beat everyone else.
At the same time, the host leaving the room will directly end the game.
In addition, the host needs to simulate an entire server while running the game, which can only be done with a high-spec computer.
Gamenova found a solution to this problem during the era of "Screeps".
We rent our own server room; all you need to do is connect.