TPU singlepcb 295 review is up! 20-55W higher power consumption

http://www.techpowerup.com/reviews/I...latinum/1.html

compared to the dual pcb 295:
http://www.techpowerup.com/reviews/Z...TX_295/33.html

the good:
same performance as dual pcb 295
lower noise in idle (same as GTX285)
same idle temps
~12C lower load temps

the bad:
same price
same noise under load
same overclocking results

the ugly:
20-55W higher power consumption!!! :stick:
no soft voltage adjustments :mad:

FAIL! :rolleyes:

Quote:

---

Originally Posted by saaya

20-55W higher power consumption!!! :stick:
FAIL! :rolleyes:

---

Haha, thanks for the post. Well done Nvidia! :p:

Ugh, not very nice at all. You'd think they'd have lowered the power consumption.

Quote:

---

Originally Posted by N19h7m4r3

Ugh, not very nice at all. You'd think they'd have lowered the power consumption.

---

Yeah, doesn't seem reasonable at all... Wonder what's drawing extra power.

Quote:

---

Originally Posted by zalbard

Yeah, doesn't seem reasonable at all... Wonder what's drawing extra power.

---

According to the review:

Quote:

---

Earlier in the review we mentioned that NVIDIA has completely revamped the voltage regulation circuitry of their design. Instead of a Volterra based solution, one from Analog is used. Apparently this change results in a considerably higher power consumption compared to the first iteration of the GTX 295.

---

Or maybe they just needed a HSF that sucks more power to make it run cooler :rofl:

Sorry but this is BS like every GFX review on that server ...

Guru3D says:

System in IDLE = 239 Watts

Dual PCB: System with GPU in FULL Stress = 406 Watts
Single PCB: System with GPU in FULL Stress = 411 Watts

By my own measuring, 10W more in FULL Load on Single-PCB version ..., in Idle +/- the same ... like on Guru3D

So because Guru3D says... maybe Guru3D is BS :p:

Heh, NVIDIA has finally caught up to ATI in terms of PCB design!

As for their choice of analog vs digital voltage regulation, don't the digital solutions produce a lot more heat? Now that I think about it, it does seem backwards (usually greater power consumption means greater heat), but I've always thought this for some reason.

.... by own measuring is? .... too BS?

NO Volterra means no softmods ??

Quote:

---

Originally Posted by eva2000

NO Volterra means no softmods ??

---

yah no soft mods

Quote:

---

Originally Posted by OBR

Sorry but this is BS like every GFX review on that server ...

Guru3D says:

System in IDLE = 239 Watts

Dual PCB: System with GPU in FULL Stress = 406 Watts
Single PCB: System with GPU in FULL Stress = 411 Watts

By my own measuring, 10W more in FULL Load on Single-PCB version ..., in Idle +/- the same ... like on Guru3D

---

sorry but when did guru3d get their sample? 2 weeks before computex...
and when did you get your sample?

techpowerup got a retail card... that might make a diference...

and i talked to w1zzard at computex and he modded the test platform to meassure current draw through the pciE slot, so lets see if he has time to meassure the actual current draw through the slot and power connectors which is more reliable than meassuring the entire psu's power consumption

Quote:

---

Originally Posted by eva2000

NO Volterra means no softmods ??

---

:( forgot to mention that...



EDIT: how does guru3d meassure power consumption?

Quote:

---

The test methodology is simple: We have a device constantly monitoring the power draw from the PC. After we have run all our tests and benchmarks we look at the recorded maximum peak; and that's the bulls-eye you need to observe as the power peak is extremely important

---

sorry but that doesnt sound too reliable... at all...
sounds like they use a watt-o-meter to meassure the power draw from the outlet.
those are very unreliable, and peak alone doesnt mean much... what matters is power consumed per time...

Quote:

---

Originally Posted by Jamesrt2004

yah no soft mods

---

dam no fun then for those volt mod/soldering challenged folks :(

That just means more fun for the rest of us when they get desperate for those few extra MHz! :rofl:

Quote:

---

Originally Posted by labs23

That's a considerably number. Aren't single PCBs supposed to be a bit better in terms of power consumption? If thats a result of a legit review then what does 1 or 2 pcb 295 differ in? -since performance is almost the same.

Poor if we'll get nothing as answer. I really hope that this is not the case though.

*A review on a different brand can help here.

---

Edit

http://www.pcgameshardware.com/. Take a look at the table.

Quote:

---

The new Geforce GTX 295 makes almost everything better than the old one: It is not as loud , doesn't need as much energy and doesn't get as warm.

---

Runs cooler but draws more power? where's the energy going?

Quote:

---

Originally Posted by xoqolatl

Runs cooler but draws more power? where's the energy going?

---

Magic :stick:

Quote:

---

Originally Posted by xoqolatl

Runs cooler but draws more power? where's the energy going?

---

to nvidias secret power harvesting labs, for use against the ATI decepticons

Quote:

---

Originally Posted by OBR

Sorry but this is BS like every GFX review on that server ...

Guru3D says:

System in IDLE = 239 Watts

Dual PCB: System with GPU in FULL Stress = 406 Watts
Single PCB: System with GPU in FULL Stress = 411 Watts

By my own measuring, 10W more in FULL Load on Single-PCB version ..., in Idle +/- the same ... like on Guru3D

---

just out of curiosity why are the reviews on that server BS Guru3D correct?

Quote:

---

Originally Posted by xoqolatl

Runs cooler but draws more power? where's the energy going?

---

given a constant power draw, the more air is moving through the cooler, the lower the temperature

Quote:

---

Originally Posted by OBR

Sorry but this is BS like every GFX review on that server ...

---

ah hi OBR. good to see you commenting on our reviews not only on tpu.
how do you measure power? wall? card power? whats the resolution, repeatability, calibration of your device? how often do you sample power draw? how do you calculate your average? how do you load the card? which app? which resolution? whats the rest of your test system like?

Quote:

---

he modded the test platform to meassure current draw through the pciE slot, so lets see if he has time to meassure the actual current draw through the slot and power connectors

---

unfortunately i dont have a dual pcb card anymore to test that. i can get you the card only measurements for the single pcb card, but that wont be useful for comparisons

Yeah, I was never too enthusiastic about the dual PCB 295 cooler. This one just seems to be much more efficient.

The cooler is more effective so I expected as much. Its nice to see they are a tad quieter at idle but I had been hoping that the load levels would have decreased. I had grown accustomed to the 295 dual pcb when I had used one for a short time but I was hoping for something more in the way of progress with the revised design other than a cost cutting measure for board partners.

If their claimed power figure are true it would put the usage on the single pcb 295s in line with the 4870x2s.

Quote:

---

Originally Posted by xoqolatl

Runs cooler but draws more power? where's the energy going?

---

better heatsink maybe? :P

interesting .. i got a single pcb gtx 295 from zotac yesterday ... ran it through the power consumption test and got lower power consumption than the card from inno3d .. idle= -5W, load= -8W, peak= -10W

both card are identical except for the sticker. i noticed something similar when i reviewed the gtx 280 (or 260?) .. different cards have wildly different power consumption, no idea why

Quote:

---

Originally Posted by OBR

Guru3D says:

---

Guru3D also says:

is exactly that... an update. Nothing new is to be found on the product when it comes to clocks, thermals, power consumption and so on.

you tell him Wizz, this self proclaimed "reviewer" needs to learn a thing or two, he thinks hes hot because he gets hardware pre-release and doesnt do good testing :up: