Is there some ad or review that makes people buy the Sabertooth and overshoot the motherboard by about $100, as it always seems to pop up...
Optical Drive: Give Pioneer a try. Robust, my DVR-215 makes a smooth rumble compared to high-pitched buzzsaw sounds I've gotten from LGs, Sonys, Lite-Ons and so on.
Case: Fractal Design R4 / NZXT H2 / Corsair Obsidian 550D etc. (link)
Take a gander and see what style and features you like. Don't get anything with too many holes/windows or sans
Power Supply: The SeaSonic G Series SSR-550RM
should fit the bill from Newegg's selection, as money seems to be no object. The 360 W sibling reviewed very well on SPCR.
1. Anything in http://www.silentpcreview.com/Recommended_PSUs
? You don't need 600 W
- that's a real-life test of a maximally overclocked setup with a ridiculous CPU. Your target is more like 500-550 W (80-82% efficiency, get Gold or Platinum with even higher efficiency to make sure), and people have made do with even less (a certain hi-power passive 400 W PSU setup comes to mind).
2. Aftermarket coolers are superior to stock when it comes to silence, especially under load. Consider a higher grade of cooler, since you seem to have the budget (Thermalright Macho maybe?).
3. SSDs are unbelievably quick compared to HDDs. In fact, there is no comparison. I've had a laptop go from booting for 30 s to two blinks (2 seconds). I would get an SSD for OS drive, no contest.
4. If you get a case with SPCR approved fans, you probably won't need extras, except if you want to replace the cooler's fan (the CM budget model you currently have selected needs a swap IIRC). Separate sound card or DAC is my personal preference.
5. You went and chose a big hulking ATX motherboard for a one card setup, where you could easily have made do with mATX or even ITX. Damn right you're going to have to have a full tower now.
PS. Anandtech offers a similar testing setup but with a more comprehensive selection of cards: http://images.anandtech.com/graphs/graph5818/46562.png
(from this article
). How there is a 100 W discrepancy to the X-Bit Labs results I can only guess, but I like to use the X-BL readings as a Worst Case Scenario.