Discussion in 'General Hardware' started by Mussels, Dec 27, 2013.
nice idea, i'll test once my GPU arrives in the mail
Cool. I did all three as I was interested to see what load the GTX 780's put on the PSU at diferent over clocks. What I did find interesting is the Wprime watts went down from 205 to 191 when I had the GPU's at a pretty high over clock...
Buy one there about $20 for a cheap Kill A Watt, then you can see how much your words are really worth..
I hope your mobo is not like my old ASUS MAX mobo which i had a E8400 in it which due to the mobo idled 200w and not the typical 100w..
so what kind of power meter should be used?
oh and im getting unsuspected error "quitting"from Wprime
Finners | Coolermaster silent pro gold 800w | i7 2600k @ 4.4GHz + GTX 680 SLI | 121w idle | 228w | P15942 526W | 30.31 points per watt
an accurate one. run it as admin, never had any problems here.
can someone confirm my math here is correct, been a long time since i learned this electrical stuff:
lets say we do the WPRIME 1024M test, and measure the wattage.
we then go time (that wprime gives as a score) multiplied by wattage and get the joules of energy used (probably round it to kilojoules considering the numbers we'll be dealing with)
we can then rank the systems in order of least KJ used to finish the test.
Add my result
Rampage | Corsair HX750 | i7 3770k @ 4.2GHz + 7970 X-Fire | 64w idle | 144w w prime | P16835 590W | 28.56 points per watt | daily settings x-fire
Rampage | Corsair HX750 | i7 3770k @ 4.2GHz + 7970 | 64w idle | 144w w prime | P10901 378W | 28.83 points per watt | daily settings single card
Love to know how you got 64w idle.. Odd that you have the same idle with a single too
his second card powers down completely, he hasnt disabled ULPS.
his mobo accounts for most of the low wattage, its just a low powered board. he's my brother so i've looked into it before.
hehe, tried it i have underclocked lowered the voltage as low as it can go got the system boot of one SSD with no video card installed although still booted in to windows and still not that low.
EDIT: on about how low they say on that site just to be clear.
as mussels has said the main reason for my lower idle wattage is my mobo, it was one of the key reasons i decided on the mobo in the first place.
also if i have worked out my KJ usage for wprime correctly i come out at 28.9 KJ (200.827 * 144 = 28919.088 (28.9 KJ )).
yeah all so down to the gfx card too ( with 2 fans but no gfx).. Dunno why i could not get mine that low before but i just tried it again and got it down to what that review said, maybe i had a HHD plugged in and did not realize at the time.
Part of the reason i got mine as i got sick of the Maximus running 200w idle which was all so due to motherboard.
EDIT: it's the 8w difference with the cards.. Sorry i was just curious how you got it so low..
Just a note, power consumption in 11 will vary with the card used. For example, budget to low midrange will show more power use in the combined test than in any others while high end cards will be in the first four (likely test 3 iirc)
DF is BUSY
ANTEC 550w BP+ (80+ bronze)
i7 3770K @ 4.2GHz + GTX 760 @ stock
desktop idle = 135w
wprime 1024 watts = 206w
3dmark11 watts = 326w
3dmark PPW = 25.5
comment = daily settings, had all my fans running max though
What do you use to test your watts then Mussels?
On my rig in specs on wprime I maxed at 350watts
i have this with a converter to use aussie plugs (same voltage, just a socket change)
James888 | XFX pro 850w | I7-4770k@ 4.375ghz +7970@ 1050/1850 | 95w idle | 210w wprime | P10923+ 360w | 30.34 PPW | Daily gamer, 213w boinc
I need crossfire to compete with some of the big boys here in benchmarks, but I am happy with gaming performance. My cpu is overclocked with high volts because I want a higher clock. If I went down to 4ghz and lower volts I bet PPW would be far greater.
James888 | Cougar SX 750w | Celeron G1620 @ 2.80GHz +7870 Tahiti @ 975/1500| 58w idle | 65w wprime | P5649 + 260 | 22.59 PPW | Cruncher/Miner, 65w boinc
The cpu is weak but sips power. Look at the idle vs WPRIME. I honestly think the cpu is holding back the gpu.
James888 | Cougar SX 850w | 2500k@ 4.4GHz + Intel HD3000 | 60w idle | 150w wprime | --- | --- | Dedicated cruncher, 130w boinc
A 3dmark score would of been too sad.
I think my FX rig uses almost as much power @ load as a meth lab.
I don't have the equipment to measure thew power usage at the moment, but this thread has piqued my curiosity.
I have some numbers for the rig in my specs. GPU's are at stock speeds and voltages with my 24/7 OC on my 3820.
Aquinus | Seasonic X1000 Platinum | i7 3820 @ 4.5Ghz + 2x 6870 (Crossfire) | 217w | 340w | P8224 (400w avg/ 445w peek) | 18.4 points per watt @ peek | Stock GPU volts + freq.
I should note that the only time it ever hits peek is when it does the combined physics and graphics test. The rest of the time it's between 390 and 400. I used my UPS to measure power draw with only my tower on the battery backup.
WOW. 800 mV under full load? Isn't that something nearly impossible, or am I just unfamiliar with all this undervolting stuff?
Take my 2500k, I have it overclocked to 4.5ghz at 1.35v. He had his xeon clocked at 3.5ghz at .8v.
Not trying to be offensive, but 1.35 V is not as impressive as 0.80 when we're talking about full load.
Haswell ULV can undervolt itself only into 0.730 V (1 core active, 800 MHz), and that's the most energy efficient platform ATM.
1.35 V still sounds pretty awesome for 4.5 GHz.
I have my scores soon. Im borrowing a meter
Separate names with a comma.