Getting close to power envelope

[W3dge]

So I want to put together a system in an Ncase M1 with:

Intel Core i5-6600K

XFX R9 390

Silverstone SFX12V 500W

My power draw ends up being 442W

Is this far away enough away from my 500W limit for some CPU overclocking (no overvolting) and slight GPU overclocking?

[i_build_nanosuits]

i wouldnt worry too much about it, this PSU is high quality...if the machine turn off on itself under heavy load then dial things back a little.

[The_Stig]

Yes. byalexandr and stefan have builds with a i3/R9 290 on 400W and i7/ GTX 980 SLI on 550W

 

You should have plenty of power

[turkey3_scratch]

Stress test a 390, about 300W. Overclocked, even more. 6600K overclocked, could easily be over 100W by a bit. Rest of system, probably you are looking at about 475W of absolute peak under torture. How many amps are on the 12V rail? Where is a review stating this thing can actually deliver its rated wattage? I see neither, and those are two critical bits of information that are necessary for these judgements. Can you give a link to the exact PSU? Silverstone's PSUs are all wonky in the naming and you have to be technical with the model numbers.

[byalexandr]

Yeah you're more than fine. My 290 draws around the same amount of power as the 390, and I'm running it on a 450W PSU no problem. I had it running on a 400W PSU but that was pushing it pretty close, but 500W is plenty.

[turkey3_scratch]

8 minutes ago, byalexandr said:

Yeah you're more than fine. My 290 draws around the same amount of power as the 390, and I'm running it on a 450W PSU no problem. I had it running on a 400W PSU but that was pushing it pretty close, but 500W is plenty.

But "400W" doesn't really matter. Any 400W power supply theoretically can deliver 1000W of current. The two important factors are the thresholds at which the protections shut it down, or the threshold at which it burns, hopefully the former being earlier than the latter. That's why I'm trying to figure out by finding any info on this PSU, which merits a hefty zero.

[byalexandr]

Just now, turkey3_scratch said:

But "400W" doesn't really matter. Any 400W power supply theoretically can deliver 1000W of current. The two important factors are the thresholds at which the protections shut it down, or the threshold at which it burns, hopefully the former being earlier than the latter. That's why I'm trying to figure out by finding any info on this PSU, which merits a hefty zero.

No, it means it's rated for 400W continuous power at a certain temperature, it doesn't mean they put a threshold at which point it shuts down or burns. They can't theoretically deliver 1000W of current because of the limitations on the hardware inside, it's just not rated to provide that much power and overloading it can cause damage. The 500W SFX is plenty, it's rated for 500W continuous at pretty high temperatures, and the hardware inside is pretty good quality.

 

Ugh, I'm sorry but the way you see power delivery is really screwed up...

 

[turkey3_scratch]

Just now, byalexandr said:

No, it means it's rated for 400W continuous power at a certain temperature, it doesn't mean they put a threshold at which point it shuts down or burns. They can't theoretically deliver 1000W of current because of the limitations on the hardware inside, it's just not rated to provide that much power and overloading it can cause damage. The 500W SFX is plenty, it's rated for 500W continuous at pretty high temperatures, and the hardware inside is pretty good quality.

 

Ugh, I'm sorry but the way you see power delivery is really screwed up...

 

No, 400W is just rated wattage, which many units fail to meet or happen to surpass. Just numbers and letters. Electrical current is simply electric charges moving along a revolving path. Theoretically, these charges can move at such a rate to merit 1000W of current in any power supply. Yes, the PSU will either burn or shut down from OCP or OPP, but that's why I used the word theoretically. In the integrated controller, an OPP limit or OCP limit is set, and that's what I'm looking for. It's usually a higher limit than the rated wattage, though OCP works with a specific rail and OPP works with the wall reading, but what I was saying is the point at which the protections kick in is important. That is really the limit. Of course, many companies set the limits too high, and then they happen to burn beforehand, so that can also be the limit. Professional reviewers, some like Gabriel Torres, check these things.

[byalexandr]

Just now, turkey3_scratch said:

No, 400W is just rated wattage, which many units fail to meet or happen to surpass. Just numbers and letters. Electrical current is simply electric charges moving along a revolving path. Theoretically, these charges can move at such a rate to merit 1000W of current in any power supply. Yes, the PSU will either burn or shut down from OCP or OPP, but that's why I used the word theoretically. In the integrated controller, an OPP limit or OCP limit is set, and that's what I'm looking for. It's usually a higher limit than the rated wattage, though OCP works with a specific rail and OPP works with the wall reading, but what I was saying is the point at which the protections kick in is important. That is really the limit. Of course, many companies set the limits too high, and then they happen to burn beforehand, so that can also be the limit. Professional reviewers, some like Gabriel Torres, check these things.

Well if it was theoretically then you would still take temperature and load ratings into account. We don't live in a theoretical world so the temperature and load ratings are what cause them to deliver or not deliver the rated amount of power. It's not even worth mentioning that they can 'theoretically deliver 1000W', when theoretically they still do what they're rated for because of the material and quality. You say theoretically but you completely exclude the physical limits of the capacitors, transformers, etc. that are actually what mostly make up the wattage rating.

 

You are right about the rail shutdown point though, they do set it too high.

[turkey3_scratch]

Just now, byalexandr said:

Well if it was theoretically then you would still take temperature and load ratings into account. We don't live in a theoretical world so the temperature and load ratings are what cause them to deliver or not deliver the rated amount of power. It's not even worth mentioning that they can 'theoretically deliver 1000W', when theoretically they still do what they're rated for because of the material and quality. You say theoretically but you completely exclude the physical limits of the capacitors, transformers, etc. that are actually what mostly make up the wattage rating.

 

You are right about the rail shutdown point though, they do set it too high.

Yes, that's my point, the transformers, transistors, etc. all play a role as to the actual limit, but I don't know the limits for those, so how can we know if the PSU is sufficient for the guy's needs? It probably is, but I'm just saying, it'd be nice to find some info online supporting that about this PSU. We would hope those limits exceed the rated 500W wattage. I think we're basically agreeing on the same thing here but in different ways 

[byalexandr]

Just now, turkey3_scratch said:

Yes, that's my point, the transformers, transistors, etc. all play a role as to the actual limit, but I don't know the limits for those, so how can we know if the PSU is sufficient for the guy's needs? It probably is, but I'm just saying, it'd be nice to find some info online supporting that about this PSU. We would hope those limits exceed the rated 500W wattage. I think we're basically agreeing on the same thing here but in different ways 

Yeah I'm probably just reading it in reading it in a way that irritates me for some reason. I'm stuck on Calculus anyways so I suppose that's what's making me irritated.

 

The PSU is more than fine though, the hardware inside is of high quality, there could be better but there are a lot of PSUs that are much, much worse (CX series for instance, and the 400W unit I had before) and the rail shutdown points are high enough that he won't reach them. The quality varies between individual units so Silverstone can't do much about it without providing like an average rating or something.

[W3dge]

Here is a link to the proposed PSU:

http://m.newegg.com/Product/index?itemnumber=N82E16817256116&nm_mc=AFC-C8Junction&cm_mmc=AFC-C8Junction-PCPartPicker,%20LLC-_-na-_-na-_-na&AID=10446076&PID=3938566

Unfortunately PCPP dosent have every SFX PSU listed

[W3dge]

Though a serious undertaking, I will probably end up sleeving my own cables to help airflow and aesthetic.

Also, I plan on using a UPS, if that throws a wrench in the works