Advice for new system for Panda3D Development

I’m building a new system here in about a month or so, and I’d like to solicit some opinions from the developers here on hardware. The system I’d like to build will be for games and game development, specifically for Panda3D.

All of my current systems are a bit outdated. I can barely do shaders on my “high-end” machine. I’d like to really kick the tires and do some serious work with Panda, but I’m at a loss as to what would be even decent mid-range or high-end hardware that would give me good support under Panda.

What would you recommend in terms of graphics card, sound card, processor, RAM, etc.? Bear in mind that the last time I built a computer for myself was about 2.5 years ago.

There is nothing panda3d specific just get the best parts for your budget.

my computer is all in one about 7 years old and i still can play most of modern games and of course, i still use it for game development.

panda runs fine on one of my now 8 year old notebooks. the notebook has an S3 grafics accelerator with 8mB memory(no shaders at all :smiley:) . the card even runs with a barely-working poor-performance driver. rest of the notebook is a 850mhz cpu and 128 mb ram.
it takes ages to load EGG files. but if once i converted the files to BAM format. it loads ok, and it runs just fine. i can even run roaming ralph with good fps in fullscreen mode.

panda itself will work in pretty much every hardware out there.a 3d-graficcard is very recommended,thought. if your content doesnt overload the machines capabilities you’r just fine :slight_smile:.
the older the hardware the more excessive the optimisation. means efficient modelling, scene management etc you will need.

actually question related to this…no issues should happen if some1 was to use the new Intel i7 and the Nvidia GTX 280?

why should it? in worst case your hardware provides stuff you cant use with panda. like the very latest shader model or so since panda might not support it yet. but aside from that it should run perfectly fine.

My 4+ year old system fried over the holiday so I spec’d out a new system. There were 2 basic criteria:

  1. Under $1000.00US
  2. Use my exisiting EIDE and SATA drives.

I’ve built AMD and INTEL systems for years, so I did a lot or research before settling on this configuration:

AMD Phenom x4 9950 2.6GHz “Black Edition”
Motherboard: Gigabyte GA-MA790GP-DS4H
RAM Memory: Kingston HyperX 4GB (2x2GB) 240Pin DDR2 800 (PC6400) x2 for 8GB on board
Video Card: Asus ATI 4870 Dark Knight 1GB DDR5 256bit
Enclosure: xClio Wind Tunnel Full tower
Power: Corsair CMPSU-750TX (750W 60A)
DVD: Samsung 22x DVD r/RW

The whole system came in at $900.00.

A few notes:
The motherboard selection was predicated on being able to upgrade to the new Phenom II processors to released sometime this year and for overclocking.

If you consider this CPU, be sure to get the “Black Edition or BE” of the processor. It has the multiplier unlocked and can be overclocked safely and easily. The Phenom x4 9950 BE is stock at 2.6GHz but has been shown to be stable and runs cool on the stock CPU cooler when overclocked to 3.2GHz+. Plenty of room for more “free” performance tweaks…

The ASUS 4870 video is “just freakin’ awesome” for cranking out 3D benchmarks. 'nuff said.

hm… this thread is getting sorta off-topic :stuck_out_tongue:

Actually, this is kinda what I was looking for. I was looking for something spec-wise from the various members of the community as to what hardware they were running to develop on with Panda.

It’s kinda nice to see that there is a whole spectrum of different hardware configurations and Panda still runs smoothly on all of them.

Oh, my other criteria was: to play Crysis with every graphics option turned on and still run at 60fps. I figured if the system could do that without breaking a sweat then it could handle anything I threw at it with Panda.

well indeed, panda wont have a problem running. BUT you can easily break performance even on a highest-end system with drawing 1000 cubes, seriously (due to some of you hardwares limitations) . if you do it wrong you wont even get close to good fps. on the other hand you can many thousands with good fps if you do it correct.
but if you have an eye on pstats and dont break performance on purpose, then panda most likely can handle it. … or should i say … your hardware can handle it? :wink:

My old PC with nvidia GeForce 5200, 1Ghz cpu, and 512mb ram ran panda fine. Pixel shaders were a bit slow though.

I’ve recently purchased a Dell XPS laptop for somewhat over 1100 euros, with an nvidia 8600M, 2x 2.5 Ghz cpu, 4GB ram, I’ve never complained. Runs everything fine, I can even run most non-shader samples in Panda’s new 1.6 software mode at great speed.
Really, you don’t need much to run Panda. If your computer can run the latest games, it can run Panda for sure, Panda isn’t so hungry.

I’ve found myself that developing on older cards works actually better than on a brand-new card. Because: if you develop on a low-ram machine with GF 5200 you’re much more centered on optimizing it (or at least supporting it). If you develop on a GF 9x00 and get reasonable FPS, older cards than that usually will have less.
This way, your game will support older cards better.

My current system is a little weak on RAM (512MB) with a 1.73GHz Intel processor, and an ATI Radeon 9600. It does okay, but with some of the more advanced features like shaders it tends to chug a bit. Doom 3 has some crazy stutter and lag when running on it.

And I did tend to find myself forced into optimizing as much as I could within the tighter constraints. It’s a good lesson everybody should learn.

I’m still looking over some of the possibilities for now. This machine is being donated to charity here as soon as I get my new one up and running.

The system I posted will do that and more… 8)

Here’s an interesting little youtube vid: providing it’s not too much off-topic… :wink: The 4870 surpasses the nVidia cards in both the anti-aliasing and anisotropic modes by a substantial margin while retaining fps…

Something else to consider using the ASUS 4870 Black Knight, you can easily OC the video cards clock “on the fly” with it’s GamerOSD feature…