|
| Notices |
Welcome to the DriverHeaven.net forums. You are currently viewing our boards as a guest which gives you limited access to view most discussions and access our other features. By joining our free community you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content and access many other special features. Registration is fast, simple and absolutely free so please, join our community today! If you have any problems with the registration process or your account login, please contact contact us. |
 |
Apr 9, 2008, 02:18 PM
|
#1
|
|
Finally time for a new PC
Join Date: Sep 2005
Location: Quebec City
Posts: 1,237
|
is 650w Psu good enough for a 9800GX2?
EVGA 780i mobo
Q6600 quadcore 2.4ghz cpu
4gb OCZ Ram @ 1006mhz
9800GX2
500Gb Harddrive
Coolermaster 690 Black Case
Razer Lycosa keyboard
Razer Diamondback 3G Infrared mouse
Windows Vista x64 version
Im ordering all this tomorrow and I need to know if the Corsair TX 650w is enough, or should I get the 750w?
|
|
|
Apr 9, 2008, 02:27 PM
|
#2
|
|
Member
Join Date: Mar 2003
Posts: 5,966
|
get a hold of kris23 i think he pretty much knows about that and can help with other things as well.
|
|
|
Apr 9, 2008, 02:43 PM
|
#3
|
|
DriverHeaven's Tomboy
Join Date: Nov 2005
Location: The pink house!
Posts: 738
|
The 650W Corsair will be more than enough for your hardware.
|
|
|
Apr 9, 2008, 06:24 PM
|
#4
|
|
DriverHeaven Extreme Member
Join Date: May 2002
Location: Ohio
Posts: 8,479
|
I have a similar question.
However I know my PSU will handle it BUT here is the deal.
How my case & PSU are configured is I had to use an extender from my motherboard to 4-pin ATX12V connector on the PSU. The reason being is I could not find one for the 8-pin ATX12V connector. The computer runs fine even overclocked w/ no stability issues. Okay but here is the question, and I feel silly having to ask.
Since I'm not using the 8-pin ATX12 connector can I use that for one of the connections on the 9800GX2? Honestly what the heck is the difference in the 8-pin EPS12V connector vs the 8-pin ATX12V connector? I guess I could go look up the white sheets but I'm feeling lazy at the moment.
|
|
|
Apr 9, 2008, 08:58 PM
|
#5
|
|
Member
Join Date: Mar 2003
Posts: 5,966
|
i feel like it's too obvious a way to get the answer for that one, it's best to check with the maker of the PSU.
Quote:
Originally Posted by Necrosis
Since I'm not using the 8-pin ATX12 connector can I use that for one of the connections on the 9800GX2? Honestly what the heck is the difference in the 8-pin EPS12V connector vs the 8-pin ATX12V connector? I guess I could go look up the white sheets but I'm feeling lazy at the moment.
|
|
|
|
Apr 9, 2008, 09:11 PM
|
#6
|
|
DriverHeaven Extreme Member
Join Date: May 2002
Location: Ohio
Posts: 8,479
|
Actually I found the answer just now.
EVGA | Power Supply
It does have to be a specific PCI-E 8 pin not ATX12 or EPS.
|
|
|
Apr 9, 2008, 09:23 PM
|
#7
|
|
Member
Join Date: Mar 2003
Posts: 5,966
|
good info, Necrosis. it's quite clear now.
|
|
|
Apr 9, 2008, 09:38 PM
|
#8
|
|
Finally time for a new PC
Join Date: Sep 2005
Location: Quebec City
Posts: 1,237
|
I have also found this
Welcome to Corsair :: Power Supplies
I did the Q6600 + 9800GX2 + 1 drive & Overclock Cpu and CPU
They recommend a 520watts.
Ive read 2-3 forum with users havint a 650tx to power the 9800 and no one got any problem. So, i will buy this PSU.
|
|
|
Apr 9, 2008, 10:56 PM
|
#9
|
|
DriverHeaven Extreme Member
Join Date: May 2002
Location: Ohio
Posts: 8,479
|
What card are you going w/?
|
|
|
Apr 9, 2008, 11:01 PM
|
#10
|
|
DH's Youngest Mod
Join Date: Dec 2006
Location: Los Angeles
Posts: 3,797
|
9800GX2 for him, i personally think its too small
how about you?
|
|
|
Apr 10, 2008, 03:20 AM
|
#12
|
|
DriverHeaven's Tomboy
Join Date: Nov 2005
Location: The pink house!
Posts: 738
|
Quote:
Originally Posted by Necrosis
Since I'm not using the 8-pin ATX12 connector can I use that for one of the connections on the 9800GX2? Honestly what the heck is the difference in the 8-pin EPS12V connector vs the 8-pin ATX12V connector? I guess I could go look up the white sheets but I'm feeling lazy at the moment.
|
No. The EPS and ATX connectors are different. Actually the voltage-ground cables are reversed, you will probably scorch your card if you PSU is not exceptionally quick at identifying short-circuits.
For a C2Q + 9800GX2 system without many mechanical drives, a 650W unit is an excellent choice. This kind of a system will consume about 300-350W under load, it will have plenty of headroom and it will work at maximum efficiency too.
|
|
|
Apr 10, 2008, 09:39 AM
|
#13
|
|
DriverHeaven Extreme Member
Join Date: May 2002
Location: Ohio
Posts: 8,479
|
Thanks but the connectors are the same pin number & are keyed the same too?
|
|
|
Apr 11, 2008, 08:55 AM
|
#14
|
|
DriverHeaven Newbie
Join Date: Mar 2006
Posts: 1
|
Quote:
Originally Posted by kris23
9800GX2 for him, i personally think its too small
how about you?
|
Hey, Ive just been lurker on this forum for a while but when I saw this had to say something.
What makes you think that the 650w is too small?
The system will draw no near the specified 650watts and even if he added another card and extra hard drives/fans he wouldn't come near the 52amps that the power supply offers.
Buying a bigger power supply would just be silly and excessive.
|
|
|
Apr 11, 2008, 09:22 AM
|
#15
|
|
DH's Youngest Mod
Join Date: Dec 2006
Location: Los Angeles
Posts: 3,797
|
Quote:
Originally Posted by BigPanda
Hey, Ive just been lurker on this forum for a while but when I saw this had to say something.
What makes you think that the 650w is too small?
The system will draw no near the specified 650watts and even if he added another card and extra hard drives/fans he wouldn't come near the 52amps that the power supply offers.
Buying a bigger power supply would just be silly and excessive.
|
do you honestly think a PSU can withstand the high heat constantly coming from the unit at the higher 85%-90%? especially under a 24/7 work environment?
there are very few, that can do that over their rated lifetime....
at least not without good cooling. those are PSUs with the highest qualiity, most cutting edge PSUs that have been designed almost flawlessly
as capacitors age, the unit becomes less and less efficient, throughput i believe also fails. while i probably overestimated his system a bit. its still the same concept. high heat speeds up this process...
if he DID add another card, and overclocked and overvolted everything then this PSU will probably touch that limit and may or may not last... the PSU hasnt been around long enough to clear 50,000 hours so i dont know how it will hold up exactly.
while the corsair is a rather good PSU, there are others out there that cant do their rated output safely even if their company was about to go under....
if you wanted to save money by cutting down of the PSU, would you really want to trust the manufacturer with the system you saved up for such a long time to get, only to find that the PSU gives out right after most of your warranty is up? worst case senario is that the PSU gives up and fails to regulate the volts and ends up dumping far more volts and fries everything before goes out....
i'd just want to spend that extra 30-40 bucks and get a better PSU to minimize that chance, so i make it a habit to recommend a PSU thats rated for twice a system's output or higher if its not too much more expensive, allows for great upgradability and lowers the chance of killing the system from overstressed components.
Antec in particuar have questionable units.
|
|
|
Apr 11, 2008, 02:32 PM
|
#16
|
|
Stingy and Stubborn
Join Date: Mar 2005
Location: New Jersey
Posts: 1,320
|
Quote:
Originally Posted by kris23
do you honestly think a PSU can withstand the high heat constantly coming from the unit at the higher 85%-90%? especially under a 24/7 work environment?
there are very few, that can do that over their rated lifetime....
at least not without good cooling. those are PSUs with the highest qualiity, most cutting edge PSUs that have been designed almost flawlessly
as capacitors age, the unit becomes less and less efficient, throughput i believe also fails. while i probably overestimated his system a bit. its still the same concept. high heat speeds up this process...
if he DID add another card, and overclocked and overvolted everything then this PSU will probably touch that limit and may or may not last... the PSU hasnt been around long enough to clear 50,000 hours so i dont know how it will hold up exactly.
while the corsair is a rather good PSU, there are others out there that cant do their rated output safely even if their company was about to go under....
if you wanted to save money by cutting down of the PSU, would you really want to trust the manufacturer with the system you saved up for such a long time to get, only to find that the PSU gives out right after most of your warranty is up? worst case senario is that the PSU gives up and fails to regulate the volts and ends up dumping far more volts and fries everything before goes out....
i'd just want to spend that extra 30-40 bucks and get a better PSU to minimize that chance, so i make it a habit to recommend a PSU thats rated for twice a system's output or higher if its not too much more expensive, allows for great upgradability and lowers the chance of killing the system from overstressed components.
Antec in particuar have questionable units.
|
I've been running my ASPIRE 520watt PSU for two years now, maybe a little more, and it has ran flawlessly, and continues to even on my current hardware with the fan speed always at 100%.
My last setup was an MSI P6N Platinum, Ballistix DDR2-800, X1800GTO OC'd to 574/558, and C2D E6420 OC'd to 3.2 GHz. Not once did I have a problem with it. On weekdays it was on for 8+ hours a day, and on weekends it was on the entire day.
My current system is ran the same way, and so far absolutely no problems even with the overclock. And my overclock is not limited by the PSU at these speeds currently, the heatsink just needs to be readjusted a little. I'm happy with 3.3 GHz though, so I don't think I'm going to bother with reseating + reapplying thermal compound
I agree that many people WAY over-estimate the amount of watts they actually need from their PSU.
|
|
|
Apr 11, 2008, 04:12 PM
|
#17
|
|
DH's oldest Geek?
Join Date: May 2003
Location: Cincinnati, OH
Posts: 1,443
|
Yeah, most folks do over-estimate, but that's not always a bad thing. Running anything at, or near it's limit is going to shorten it's life time.
If you look at my specs, you can see that the 620HX is a bit of overkill, but it's there because I wanted the Corsair quality, and I knew that I was going to upgrade. I'll be putting i a 3870 x2 shortly, and the 620 is going to be able to handle that quite nicely.
Actually, I think that the PSU was just about the most expensive part of my system. But it's the most important, and it's always worth spending more on a PSU than going cheap, and having the cheap one destroy other parts of your system when it fails. I opted for a lesser vid card, and less memory just so I could afford the PSU.
|
|
|
Apr 11, 2008, 04:23 PM
|
#18
|
|
DH's Youngest Mod
Join Date: Dec 2006
Location: Los Angeles
Posts: 3,797
|
its not simply about watts, watts are a meaningless measure now, its about the internal components and how well they take a beating. a good PSU can probably say 650w but its internals are capable of putting out 750w safely. not recommended to try but that is best case senario with the highest quality SeaSonic, Silverstone, and other similar PSUs. on the other hand, you got PSUs marked for 650w that cant even output 550w safely, an antec trio is one of those.
an example of a good PSU i think are certain Silverstone PSUs with similar internal parts, some could just be better components that were binned for higher temps and such.
seems your ASPIRE unit has been holding up rather well and still handles things decently, 2 years isnt that long, if it didnt run for that long then thats not at all good, older PSUs seem to be rated for 50,000 hours while the newer ones at 100,000 my PSU is about 3 years and has been running 24/7 under high heat conditions. i think it was at least 50C in its case at all times. my PSU has been doing 26,000 hours already and is at about 50% of its lifespan, its no surprise its failing with my heavily volted system even though it is only the same as a stock quad core setup or just above.
have you ever inputed any upped volts? that can take your quad core easily to double its recommended power use. power use isnt very highly increased until volts are increased eg. Amps X Volts = Power. i think a 130w EE Quad core can easily do 260w if volted enough.
i think my PSU is only capable of outputting 300-400w stable now. Bulletheory, i think, wanted this system to run several years with no need to change any components. PSU im assuming is included. since his PSU is a better model rated for 100,000 hours, this may work. but im not guaranteeing anything..... once a PSU goes outside its regulated levels, the milage will vary, the PSU may do something like 700-750 but its uncertain as it was never built for that.
i still stand by the 750w unit as it has all the connectors for good expansion, and it will certainly handle dual GX2s now and probably even a heavily OCed system later.
|
|
|
Apr 11, 2008, 07:16 PM
|
#19
|
|
DriverHeaven's Tomboy
Join Date: Nov 2005
Location: The pink house!
Posts: 738
|
Quote:
Originally Posted by Necrosis
Thanks but the connectors are the same pin number & are keyed the same too?
|
Both connectors have the same number of pins. I know for a fact that the PCI-E connector fits in the 8-pin CPU connector, but I'm not sure if the opposite is true.
Quote:
Originally Posted by kris23
do you honestly think a PSU can withstand the high heat constantly coming from the unit at the higher 85%-90%? especially under a 24/7 work environment?
|
Actually yes, any good quality unit should be able to output at least 90% of its top output rating for 24/7 operation. Peak power output and rated top power output are two different things. The first is the top power output a unit can momentarily achieve. The second is what the unit can continuously deliver and still meet an expected lifetime. Reputable manufacturers usually go by the second.
Quote:
Originally Posted by kris23
its not simply about watts, watts are a meaningless measure now, its about the internal components and how well they take a beating. a good PSU can probably say 650w but its internals are capable of putting out 750w safely. not recommended to try but that is best case senario with the highest quality SeaSonic, Silverstone, and other similar PSUs. on the other hand, you got PSUs marked for 650w that cant even output 550w safely, an antec trio is one of those.
|
Actually that's not true. Only a few very expensive units have transformers which are rated as able to deliver 110%-115% of their the unit's top power output. The transformers are the only part which matters on how much power a unit can output, the rest of the components are only serving the transformers and perform other tasks (regulation, control, etc). Of course you need to install proper parts for a unit to perform well, but it won't ever output more than what the transformer can output; only less.
Plus, most good units would have their overload protection circuit switch them off if the power demand the top power output of the PSU by 5-10% for a few seconds, some even momentarily. Failure to do that would mean that the unit cannot protect its own components properly.
Quote:
Originally Posted by kris23
have you ever inputed any upped volts? that can take your quad core easily to double its recommended power use. power use isnt very highly increased until volts are increased eg. Amps X Volts = Power. i think a 130w EE Quad core can easily do 260w if volted enough.
|
That's not really accurate either. While what you just said is true (raising voltage will increase the power demand of the CPU) it would never double it and the maximum thermal design power is quite different from the actual power consumption. A considerable voltage increase can "only" increase the power consumption of a CPU by 25-40% and no CPU can actually eat 130W directly from the motherboard. There is actually no motherboard which can deliver that much power anyway. A whole basic Core 2 Quad system with a low-end VGA card will not draw more than 120W from the AC plug under load. The QX9650 quad needs about 50-75W to properly operate, perhaps 100W if it gets seriously overclocked.
(Think about it yourself. P = V * I, in an ideal world. So for P to double as I is a constant and can't be changed, you need double V. 2,65V+ for a Core 2 CPU. It is not going to happen as that much voltage would tear both the motherboard and the tiny CPU core apart like a thunder. Literally.)
Just a note from myself; using a too powerful PSU is not a really good idea. Actually very low load is more harmful to modern switching units than heavy load and the efficiency of most PSUs is very low below 20-25% load. Ideally you want a unit running at 40-70% most of the time which will make it most efficient and give plenty of headroom for future upgrades.
P.S. the 650W unit will be absolutely fine in this case.
|
|
|
Apr 11, 2008, 07:23 PM
|
#20
|
|
DH's Youngest Mod
Join Date: Dec 2006
Location: Los Angeles
Posts: 3,797
|
hmm.... looks like i missed alot of points with a PSU, ok i guess my stuff either went bad or its something else.
i know that having a too small PSU is worse in that it decreases efficiency and sometiimes puts more stress on the unit as it has a harder time regulating its output.
i guess i have been missguided by certain reviews where a system total power consumption jumped 60+w when the CPU was OCed
i also know that the thermal design is different from the actual power use, the TDP means the maximum amount of heat in terms of watts right? but on average, i have seen the q6600 actual power to be rather close to its TDP so i just used that as my marker.
i guess i still got more to go, but dont PSU output and general efficiency decrease as certain components age?
Last edited by kris23; Apr 11, 2008 at 07:32 PM.
|
|
|
Apr 11, 2008, 11:10 PM
|
#21
|
|
DriverHeaven Extreme Member
Join Date: May 2002
Location: Ohio
Posts: 8,479
|
Quote:
Originally Posted by Grace
Just a note from myself; using a too powerful PSU is not a really good idea. Actually very low load is more harmful to modern switching units than heavy load and the efficiency of most PSUs is very low below 20-25% load. Ideally you want a unit running at 40-70% most of the time which will make it most efficient and give plenty of headroom for future upgrades.
|
From my line of work a low load can definitely cause issues. I wonder if the above is worse if the PSU has multiple 12VDC supplies.
|
|
|
Apr 11, 2008, 11:21 PM
|
#22
|
|
DH's Youngest Mod
Join Date: Dec 2006
Location: Los Angeles
Posts: 3,797
|
probably but i dont think so, with multiple rails, a PSU regulates each rail separately depending on the amps output, it may still be just as inefficient but i dont think its as bad as a single rail being regulated.
but then again im not so sure on that as i dont know what kind od components are being used to regulate the current and volts.
|
|
|
|