fanless seasonic for gaming?

[coffeejunkee]

Why do you keep linking to that Hercules psu? Oh I see, it blew up...strange, the packaging looked so promising. Now go find a Seasonic that exploded at half its rated power? And why do you keep linking to that outervision calculator, even after you are told it overstates/is just an indication?

Maybe you should actually link to this: http://www.tomshardware.com/reviews/radeon-r9-290-and-290x,3728-4.html, a 250W card pulling 365W for short amounts of time. Then again, they also show a card that can only pull 75W from a pci-e slot actually use 141W: http://www.tomshardware.com/reviews/geforce-gtx-750-ti-review,3750-20.html.

 

[thedosbox]

Why do you keep linking to that Hercules psu? Oh I see, it blew up...strange, the packaging looked so promising. Now go find a Seasonic that exploded at half its rated power? And why do you keep linking to that outervision calculator, even after you are told it overstates/is just an indication?

Yeah, nobody sensible (i.e. everyone else) is suggesting buying a cheap PSU that claims a high-wattage, hence all the Seasonic mentions.

Maybe you should actually link to this: http://www.tomshardware.com/reviews/radeon-r9-290-and-290x,3728-4.html, a 250W card pulling 365W for short amounts of time. Then again, they also show a card that can only pull 75W from a pci-e slot actually use 141W: http://www.tomshardware.com/reviews/geforce-gtx-750-ti-review,3750-20.html.

Good find. However, note that those spikes last a millisecond. Even if the GTX 770 exhibited similar behaviour, 200W headroom (assuming a 500W PSU) would still be plenty.

 

[_Rick_]

And even the OPP circuits have a certain reaction time, and they're usually tuned with a bit of headroom - 110% long term load shouldn't be an issue on a well made PSU.

 

[_Rick_]

http://www.thermaltake.outervision.com/Power


personally, i recommend the Hercules
http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=324
funny how after this, i get links for the forum sent to me in the mail.

i generally like to take a minimalistic approach to my systems, but, given that this is an ENTHUSIAST' forum, i tend to assume people have - you know - enthusiast level hardware in their rigs.
So right now my rig contains two HDD, a SSD, mouse/webcam/2 chargers/, 5 fans + a pump and led, a 1.3V CPU, 1.8V CIV, soon-to-be-voltmodded 770 (which also has 3 fans on it, funny, uh?), DVD, all of which like to work at full load. (ofc the guy above wants a 780, and who can blame him).

i'd also like my PSU to run at - shall we say - 70% load MAX, since i like to keep it within reasonable efficiency. i actually like to keep it at 50%, but for the sake of argument, we'll keep it at 70%.

Which means that, with a theoretical draw of 450W at the PSU, you'd want a 680W PSU (drawing 490W at the wall).

Yes, i can do basic math. And build systems too.

Efficiency curves are pretty flat, actually:

Run it at 50% or 100%, and you gain 0.5% efficiency, at a huge markup when you buy the PSU.
So your desire to run a PSU at 50% might be valid for cheap ones, but passive PSUs are usually engineered to a higher standard.
And depending on your usage model, you lose a few Watts at low loads, by over-speccing, which over the lifetime of the PSU might just make up what you gained at the high end. At least my PC idles for 95% of the time at around 50-80W load, with an OC'd 2600 and a more-or-less stock 770. The 500W PSU I have installed is vastly overkill, but a few years ago it was one of the few lower end 80+ gold rated units, and I got a great deal on it, otherwise I would have gone with a 400~450W unit....

 

[coffeejunkee]

Good find. However, note that those spikes last a millisecond. Even if the GTX 770 exhibited similar behaviour, 200W headroom (assuming a 500W PSU) would still be plenty.

To be clear, I have my doubts regarding those results.

I think the best way to show that it will work fine is this: http://www.ifixit.com/Teardown/Steam+Machine+Teardown/20473. GTX780 on a 450W psu.

Or this: http://www.apple.com/mac-pro/. 15 cores, 2 high-end workstation cards, 475W psu.

 

[thedosbox]

To be clear, I have my doubts regarding those results.

Yeah, xbitlabs used to take a similar approach but stopped. But even when they were doing it, they never experienced those sorts of results.

I think the best way to show that it will work fine is this: http://www.ifixit.com/Teardown/Steam+Machine+Teardown/20473. GTX780 on a 450W psu.

Or this: http://www.apple.com/mac-pro/. 15 cores, 2 high-end workstation cards, 475W psu.

Nicely done

 

[mindless1]

Go with a fan for a standard tower design where the case is negatively pressurized by more exhaust fan flow than intake fan flow because if the PSU is passive in that situation, the heat it creates is being pulled into the system.

Will that make the system implode? Usually not, but your other system fans will have to spin faster and create more noise to maintain the same case temperature. This is only a generalization, often true but in some cases and with some fans that need a minimal RPM to keep spinning, your results may vary.

 

[BFG10K]

Yeah, but OP has a 780, so we're looking at around 350-375W. At 90% efficiency, this means around 40 Watts thermal output of the PSU.

40W is really not an issue at all. There are passively cooled graphics cards that use a lot more power.

SPCR tested a fanless X-400 at full load in their hotbox (i.e. no ventilation) and they still couldn't get OTP to trip until they pointed a hairdryer directly at the PSU. Nobody will have a case as bad as that scenario.