Will a 650W PSU suffice for my build? - Page 4

In all honesty I genuinely tried to be nice ... I didn't misunderstand your analogy... I simply found completely flawed and utterly idiotic. Quoting 10'th grade physics manuals.. is as relevant to this discussion as chess theory or maybe Lotka-Volterra... and a testament of your commitment to pertinence.... which obviously I lack. Again, your analogies betray a complete lack of understanding of the topic/issue discussed.... moving from TDP (which is as relevant to actual power consumption as MTBF to life expectancy of fans), to strong forward winds, lack of gas and being late for boarding... and you're accusing me of impertinence!?

Of course I can't know Precisely what every chip draws in every possible scenario... for Obvious reasons.... from "no two chips are identical" (voltage tolerance differences, leakage differences, etc.. for chips that appear virtually the same), to the difference in hardware used for testing and measurements (platform differences -> VRM differences -> waste/efficiency differences) ... to differences in testing conditions and idiosyncracies typical to each user and reviewer.... that's why I'm always working with a margin of error meant to compensate for such differences, assuming the worst.... that's why your attitude is borderline insulting... for the third time in this thread I'm telling you that those are peak values, that the suggested PSU output it's not the minimum necessary but "plenty"..... and you keep repeating the same nonsensical, idiotic analogies that only you find pertinent..

As for what if an asteroid strikes... or maybe he gets abducted by evolved baboons, dressed as Nazi paratroopers, and forced to watch them violently masturbate to the tune of Kampf Dem Kapital.... ... don't you think that if he'd want a benching/HWBot rig he'd know it?.. and maybe mention it?

650w will be more than enough. I ran 2 heavily OC'd GTX 680's and an i5 of my TX 650w v2 for months. People are so paranoid about power supply's, when they say your power supply can output 650w it can output 650w and there absolutely nothing wrong with having a 650w load on it. It's made to supply 650w, really the PSU could supply a fair bit more than that but they have to compensate to make sure the PSU will take it easily. So drawing 650w form a rated 650w unit is not running it at full load at all. Whilst gaming my SLI 680 sytem never drew more than 520w from the wall. Only reason I upgraded as I wanted modular cables

Alright, let's not turn this into a playground squabble.

So what you're saying is that the law of conservation of energy is wrong? Or that it doesn't apply in this case? There's energy involved, and it's changing forms. I'd say it's relevant.
The only thing I'm getting from this is that you're incapable of understanding the point of a basic analogy.
My entire point, that I've tried to make with those examples, is that it's imprudent to put yourself in a situation where every component has to perform exactly as expected, and you have no room to expand. Can you get away with it? Yeah, probably, most of the time. If you want certainty, you need a safety margin. If you don't want to spend the $20 extra, then you can't afford a safety margin. I don't care about peak vs. steady state. If the component is capable (or even might be capable) of drawing it, my PSU damned well better be capable of supplying it.

So, you're saying that the likelihood of someone on OCN wanting to upgrade their computer, or do something new with it, is about as likely as an abduction by evolved nazi baboons? Kay.

Point being though is that you're entirely overestimating what the rig in question will be drawing which is what led you into this debacle. The post right above yours, which I have quoted below, disproves whatever notion you have that a GTX 670 SLI system will be "at the limit" of a 600 watt PSU.


These are with overclocked 680's with 520 watts being noted as a peak value from the wall (from which efficiency must be factored in for the true draw), and note that 670's draw between 20-30 watts less than stock, and because of the limited overclocking/overvolting potential allowed by these cards it makes overclocking fairly moot in terms of increased power consumption. There will be plenty of room for margin of error and other anomalies to occur given as the amount of overhead that will be present is in fact a large safety margin with such a setup. So can we give this a rest, please?

Proof? Not that I don’t believe you, just that I know my PSU’s max draw is around 890W so if it can draw 100W above that it’s pretty exceptional.

so basically I can run a 680 sli setup with my 630W psu? Cool! What an age we live in when a top of the line card can be run with a minimal PSU!!!!! In the far future(1 year or so) I will definitely try this! Also, limited overclocking potential? That explains a lot!

Proof? http://www.hardwaresecrets.com/article/Antec-High-Current-Gamer-750-W-Power-Supply-Review/1141/8

Managed to deliver 987W

I got a CX600 to replace my dying PSU it blew up in about 2 hours and yes it was enough wattage. Luckily it didn't take any components with it (seriously read some of the reviews on Newegg). Corsair doesn't actually make their power supplies they buy from OEMs and re-brand them so you never know what brand your actually getting. So I would recommend going with an OEM like Enermax, Seasonic etc... They cost more then other companies but for what I have heard they make great PSUs. After the CX600, I got a 600 watt Cooler Master (I think they're an OEM, haven't looked into it) it hasn't given me any issues, but I've only been running it for a few days (since Monday). Hope you have better luck than I did took two dead PSUs before I got a working one (CX600 and a Thermaltake). So good luck!

yeah I went with the rosewill 750M capstone series PSU, heard great thing about them.