iyr

Might be a Poe's law thing, but I genuinely can't tell if half the youtube comments are serious or sarcasm.

I use two gtx 590s to render in Blender and have never had an all-out system crash, and these things run hot. That said, blender itself has crashed before due to running out of vram (whopping 1.5GB per gpu). It might be your power supply struggling to deliver power for an extended time or itself overheating since the older Fermi architecture of 500 series geforces is quite a power gusler.

shenanigans

I prefer/make digital art, namely because you can't Ctrl+Z "happy accidents" with traditional art. ^_^

I find region locking really obnoxious since it can prevent you from playing imported games. When I had a PS3, there was a chance the latest installment (at the time) of one my favorite game franchises would not be localized to North America. I ordered it off Play-asia and was able to play it (albeit without any idea what was going on) because the PS3 did not have any region protection.

Why do I have this image? Why don't you?

I'm half-convinced it's the reason why my old RAT 7 is dying. Taking my fluffy jacket off in winter then touching the mouse would cause a nasty static shock. Not certain though.

My Modest Collection minus katana. Used to have a few more balisongs (one broke, gave the others away).

Even though 700 big ones would go much further for gaming performance on a geforce or radeon, FirePros don't do too bad for gaming but are often good enough for most games, (speaking from experience with v4800 and v7900). One of the biggest industries that would be using these cards is the gaming industry and artists find it handy to adequately test their models in a DX environment. They're quite versatile.

I have a FirePro. I am the 1% :D

I've been running a v7900 with twin GTX 590's for compute for a while now and I'll just say it now: you CANNOT run a FirePro for 3d tasks on one monitor and a geforce(s) for gaming on another simultaneously, even if secondary monitors are connected to the geforce. You CAN however run Windows on one of the cards and have the other for compute. That is, you can simultaneously game on the geforce while programs utilizing OpenCL do compute tasks on the firepro (individual program compatibility varies). If Windows is initialized on the firepro, it will do all DX and Opengl rendering throughout the os, even for monitors connected via the geforce (video signal is simply passed through the geforce). I've tried doing setups where I'd have the Firepro drive one monitor for 3d modelling and the 590s in SLI driving another for games. It just doesn't work. When I booted up a game on the monitor connected to the 590s, I observed that all of the gpu load was going directly to the firepro while the 590s were doing diddly squat. When I initialized Windows on the 590s the opposite occurred: they'd do well for gaming, but also ended up doing the OpenGL viewport rendering on the monitor connected to the firepro while its gpu remained mostly idle. Therefore, it'd probably be best to connect all of your monitors to one card and reserve the other for compute tasks, but don't expect to be able to game and model at once. OS-wise, it's really easy to setup in Windows 7 (and should be just as easy in Vista or 8 since these os' are based off Windows kernel 6.X with natively supports multi-gpu driver installs). Just install the drivers for both cards as you would with any build.

If I may make a suggestion, can you add one or two RAM queries to your form? RAM can affect OC stability. My 3930k used to be rock-solid stable at 4.6GHz when I was only running 16GB (4x4), but when I upgraded to 64GB I had to lower the multiplier down to 4.4GHz to be stable again.

If we've learned anything about the last time BMW cooked up a case it's that this one will likely be horribly designed favoring aesthetic over function and ridiculously overpriced. But then there will eventually be a revamped successor to it that's more reasonably priced and a little more practical.

Yes and no. I've been running a FirePro as my primary graphics card and two Geforce GTX 590s as compute cards for a long while now and have made some interesting observations. You can connect an additional monitor to your 240, but rendering will only be done on the card the OS is primarily using. That is, rendering would be done directly on your 550ti, then passed through the 240 to your secondary monitor, so your 550ti would be doing all the work for both monitors. Also, the secondary monitor will have some input latency to it since the 550ti would be sending data through the chipset and through the 240 to output to that monitor. Dragging "demanding" windows from one monitor to the other can cause driver crashes or program crashes, though I haven't had this happen too often.

You also might want to wait on Haswell to be released which is due here pretty soon before you decide on the CPU. If you're too impatient to wait, AMD usually has a decent track-record for forward/backwards compatibility with their sockets.

I use Mushkin Blacklines in my desktop and Corsair Vengeance in my server. I have a biased predisposition for Mushkin products because they're a Colorado firm and I'm Coloradan, so buying their stuff supports my local economy. Aside from that, their RAM has actually ran fairly well and stably in my system.

I'm kinda split. I love CentOS because it's rock-solid stable and gets great long-term support, being an equivalent to RHEL; it's a good "set and forget" os. However that excellent stability comes at the price of long wait periods for new features, since they're tested, tested, and retested to be absolutely bug free. I also like it's adaptability to be a great workstation, server, or virtualization platform (or any combination thereof). On the other hand, I also enjoy Sabayon because it's bleeding edge and gets the cool new toys before most of the other kids on the block. However, that comes at the cost of stability since any new features have not had the bugs ironed out (I once installed an update that completely broke x.org's ability to render font). I also really like the Sabayon dev team's mission statement of just wanting to provide an excellent user experience that's not backed by any political or commercial agenda. CentOS and Sabayon are more-or-less opposites, but I love 'em.

AMD now has TressFX (the fancy hair simulator used in the new Tomb Raider) and Nvidia has upped the ante with a fluid engine that uses PhysX. AMD's FirePro cards are also capable of Framelock and Genlock synchronization (with the requirement of an s400 module).

Maybe I'm just crazy, but to me you kinda look like a young Donald Faison in your picture.

"Watcher (Aramaic, עִיר, iyr; Theodotian trans: ir; from the root of Heb. `er, "awake, watchful";[1] Gk. ἐγρήγοροι, trans: egrḗgoroi; Slav transliteration, Grigori,[2] "Watchers", "those who are awake"; Chaldean, "guard", "watcher"[3]) is a term used in connection with biblical angels. Watcher occurs in both plural and singular forms in the Book of Daniel (2nd century BC), where reference is made to their holiness. The apocryphal Books of Enoch (1st and 2nd centuries BC) refer to both good and bad Watchers, with a primary focus on the rebellious ones." -- http://en.wikipedia.org/wiki/Iyr In other words, I lurk too much.

It's just a simple gear I animated myself. Specifically it's an escapement used in mechanical watches. The illustration on the escape wheel itself was drawn by a friend. Spoiler for full-sized version:

All I could think about while reading this post is to take one of these future chips, grinding it to dust, mixing it with thermal grease and then call resulting the mixture hummus. Just me? ok. But seriously though, as a budding junior programmer this definitely seems interesting as current parallel computing languages have a bit of a learning curve to them.

Have any OpenCL or compute benchmarks cropped up? I heard the Powercolor 7990 and Ares II ended up being really bad for compute since no official drivers for dual gpu cards had been released by AMD. Since AMD will be churning out their own 7990, hopefully driver support for OpenCL applications will be improved quite a lot.

Shenanigans

Thanks for the suggestions guys, a friend help out to make a sub network of sorts to get things working as desired, so everything's all good.