Sure could, even with a Fermi. As a rough estimate, you could say up to 120W for the CPU/mobo/RAM and 10W each for the drives (assuming not super efficient 3.5" ones), that leaves you with around 250W for the video card and any minor things.
But that would run PSU very, very close to it's max which is something I would not recommend in the long run. And we are not even taking into consideration that upgrades would be hard and/or impossible without changing PSU also.
Don't forget that PSU age and loose ability to provide original amount of power with time due to capacitor aging so to make a long story short it's a bad idea
to not leave 20-30% headroom for growth and longevity when matching PSU with rest of the system.
Not to mention that your math is kinda off. Most AMD Phenom II are either 95 or 125W TDP alone, but you say that 120W is enough for CPU and motherboard ??!
And if we go down Intel route ?
Intels new quads are also 95W TDP (or 82W in case of single I5-750S CPU).
It gets even worse if we look at the I7 series as most CPU there are 130W TDP and few with 95W and 2 with 82W TDP.http://ark.intel.com/ProductCollection. ... Segment=DT
I5 Specshttp://ark.intel.com/ProductCollection. ... lyID=28037
As you can see you are mistaken
if you think that this PSU would be good choice for running a modern quad with a Fermi card and giving somebody else advice to do just that is....shall we say irresponsible ?
You don't need a 30% reserve and to use the TDP. The TDP is well over what a CPU will use in normal usage. The only way you'll see TDP is if you leave the stock CPU on and run multiple threads of a CPU burn type program plus add a graphics program to the mix.
Add a non stock heatsink of the >700g variety and ventilate the case properly and your CPU won't hit the TDP unless you overvolt it.
Just because a CPU says 95W doesn't mean it will ever draw 95W take a serious look at the data on http://www.silentpcreview.com/article1056-page2.html
and you see:
AMD X4 630, 95W TDP, 82W at load
AMD X3 440, 95W TDP, 78W at load
AMD X3 435, 95W TDP, 75W at load
AMD X3 720, 95W TDP, 71W at load
AMD X2 255, 95W TDP, 57W at load
AMD X2 555, 95W TDP, 57W at load
AMD X2 250, 95W TDP, 50W at load
Sure you can think worse case and treat all of these CPUs as 95W but you'd be giving way more than a 30% overhead on some of these. (The range is actually 14% to 48% overhead gained by treating all of these as 95W).
Similarly with Hard drives or SSDs your highest power draw is likely to be during the OS boot process when the CPU is mostly idle. When you stress the CPU and video card the drives are likely not going to be busy for more than a minute or two at a time.
So if a HD takes 9W at load and 5W at idle you already have another case where treating it as a max wattage device builds in more than 30% overhead in 99% of the usage case scenarios.
Lawrence Lee (SPCR reviewer) took these components and the CPUs I listed above and couldn't get it to break 132W
# Asus M4A78T-E motherboard - 790GX chipset
# Corsair XMS3 memory 2x2GB, DDR3-1600
# Asus EN9400GT Silent Edition graphics card - 512MB
# Western Digital VelociRaptor hard drive - 300GB
# Seasonic SS-400ET ATX power supply
# Microsoft Windows 7 operating system - Ultimate, 64-bit
# Scythe Kabuto CPU cooler
Even with a quad core CPU he never got even close to 150W let alone anywhere close to 70% of 400W. Even leaving a 30% safety margin you only need to stay below 280W during stress testing.
A GTX 460, 465, or 470 could slide in under that limit after you take the old 9400GT out.
And believe me you could make MUCH BETTER choices in components and leave even more head room.
As you can see treating components by maximum power draw and then asking for 30% reserve on top of that is clearly bad math. I won't call it irresponsible, I'll just call it a waste of money with a dash of fear, uncertainty, doubt (FUD) thrown in.
Measure real world usage and add 30% or use TDPs and don't add 30% but don't do both.