Multi-Core Support for Source Unveiled

Scanner wrote..

...

Can you say the same for the newer engines still being written?

Nope.



And the scalability as well. HL2 runs on all kinds of computers, ranging anywhere from the 4+ year old computers (That have to use lowest settings) to the new computers that can use HL2 x64 on their dual cores, getting 100FPS constantly outdoor on the coast.



Other engines? Notsomuch.

FightingChance wrote..

"If we're right with the approach we've taken - which is to iterate and build on top of Source - we can get there a couple of years ahead of where they could be."



Or, Unreal Engine 3 and Cry Engine 2 can already be farther ahead, and engines in development right now can be even further, without having to worry about codebase written years ago.

Notice the first three words of the Valve quote: "If we're right". They obviously don't know if they'll be able to take advantage of their head start or not. Only time will tell.



Also, you're quick to trash the HL2 engine's codebase, but I wonder if mod developers would do the same? Perhaps Valve thinks they have a winning hand partly because of all the mods for the HL2 engine, and the coders' familiarity with its APIs. Were Valve to trash it all and start anew, developers would have to as well--and they might just decide to do it in UT2007 or something that's out right then and there. Right now, those coders can feel confident in their working code, the ubiquity of the HL2 engine on users' computers, and users' computers' capability to run the HL2 engine well. Can you say the same for the newer engines still being written?

Looking good.

Yes, Valve should just trash the whole source engine, and create a new one in order to top those engines. It wont be so bad, we will only have to wait another 5 years for portal and TF2.

FightingChance wrote..

Or, Unreal Engine 3 and Cry Engine 2 can already be farther ahead, and engines in development right now can be even further, without having to worry about codebase written years ago.

well said.

"If we're right with the approach we've taken - which is to iterate and build on top of Source - we can get there a couple of years ahead of where they could be."



Or, Unreal Engine 3 and Cry Engine 2 can already be farther ahead, and engines in development right now can be even further, without having to worry about codebase written years ago.

I like red, but I'm more interested in colors.



*blink*



What?

I like the rain effects, but I'm more interested on how much Valve can take advantage of multi-core processors.

Sweet, that $2000 I spent on my computer will finally go to use.

Very good news.. The CPU has always been a bottleneck in Source games.. Taking advantage of multi-core will give a big performance boost..

I'm interested in how this will improve the complexity of Bot AI, server-side. Hopefully once they can offload some of that state machine code to helper threads, the AIs can become far more complex. I guess someone still has to write all that logic, though.

Sounds good.

Oh man. That rain would add for some really nice dramatic, scary places. =)