R9 270x secrets

[adamsjus1]

I know people like that in the gtx 900 cards you can turn off the fans

 

Well I did it with my R9 270x windforce card from gigabyte.  

 

AMD fanboy all the way. Did it in MSI afterburner. Only works when its open though when you make a custom fan curve.

[Bogica]

And the temps? My Sapphire R9 270 is silent even while gaming so no need for that. My CPU cooler seems to be a lot louder (Hyper 212+).

[Albatross]

Okay, so? You can cook bacon on that card too.

[Zeke]

What if I told you you can turn fans off on any card with active cooling? You don't even need the Afterburner, just a pair of wirecutters.

[adamsjus1]

 

 

 

Temps not running a game are at 40 C   I have a fan set to turn on at 45

[adamsjus1]

But it makes no difference in noise because I have 6 case fans and two corsair sp 120mm running at full speed on my hyper 212 evo

[Bogica]

Temps not running a game are at 40 C   I have a fan set to turn on at 45

Nice! Not bad at all. I wish my CPU cooler was quieter, but that's what I get for using stock. I'll have to rebuild my PC during the summer.

[Majestic]

Except, Maxwell can do this because of the low TDP*. AMD VRM's run so hot, they require constant cooling.

Don't be a wiseguy and do something the card wasn't designed for.

 

*edit

[Sakkura]

Okay, so? You can cook bacon on that card too.

 

The R9 270X draws significantly less power than the GTX 970 (and only slightly more than the GTX 960). So if you can make passive 970s that don't overheat, you can do the same with the 270X.

[Sakkura]

Except, Maxwell can do this because of the low power consumption. AMD VRM's run so hot, they require constant cooling.

Don't be a wiseguy and do something the card wasn't designed for.

 

Yes the GTX 970s 175W power consumption is amazing, so much lower than the 120W power consumption of the R9 270X.

 

[Zeke]

Yes the GTX 970s 175W power consumption is amazing, so much lower than the 120W power consumption of the R9 270X.

 

*snip*

Except 970 has about 1.7-1.8x performance of 270X.

[Sakkura]

Except 970 has about 1.7-1.8x performance of 270X.

 

Sure, the efficiency is great. But that has nothing to do with whether you can passively cool the card.

[Zeke]

Sure, the efficiency is great. But that has nothing to do with whether you can passively cool the card.

Actually it does, you get the same performance while maintaining lower temps. Or better performance with same temps.

 

Everything can be passive cooled if you're brave enough (that's a 680 btw).

[Sakkura]

Actually it does, you get the same performance while maintaining lower temps. Or better performance with same temps.

 

But that has nothing to do with whether you can semi-passively cool the card. All that matters there is what the power consumption is, and how good the heatsink is. An R9 270X is easier to semi-passively cool than a GTX 970. But of course, Nvidia will offer the best performing cards that can realistically be semi-passively cooled. You wouldn't want to try this with an R9 290X.

[Tim Drake]

270x is okay but literally anything higher, I doubt this would work.

 

And don't even mention trying this with a reference cooler, you will need a half decent heatsink on anything higher than a 270x to passively cool it.

[Majestic]

Yes the GTX 970s 175W power consumption is amazing, so much lower than the 120W power consumption of the R9 270X.

 

 

TDP =! actual power usage. And I did say VRM's not power consumption, maybe a particular model works, maybe it doesn't. Disabling a fan however will make sure that if it doesn't work, you brick it because something will burn out. If it's not designed for it, don't do it. Not without modding or at the very least atleast and making sure everything (not just GPU temp) is functioning properly and adding passive heatsinks where needed. You don't just grab a random card and start shutting down the fans, that's retarded.

 

Also, the construction of the H-bridges and PWM motors come with a hardcap. They won't try to turn on below a certain point/duty cycle. Because they might potentially damage or wear out quick (coils/bearings). Truly, even for maxwell cards I don't recommend doing this to a card which was not designed for it. Not without consulting the corresponding factory/producer. I really hate Linus's video about this, he takes absolutely nothing technical into consideration.

 

Your cheap "sarcasm" is really out-of-place.

[Sakkura]

TDP =! actual power usage.

 

Exactly. The Nvidia TDP is much more "optimistic" than the AMD TDP. According to the TDPs, the R9 270X should draw more power than the GTX 970, but as I've demonstrated above, in reality the GTX 970 draws significantly more power than the R9 270X.

[blackadder]

More gpus should have secrets, easter eggs too.

[R_Golden]

Exactly. The Nvidia TDP is much more "optimistic" than the AMD TDP. According to the TDPs, the R9 270X should draw more power than the GTX 970, but as I've demonstrated above, in reality the GTX 970 draws significantly more power than the R9 270X.

Power draw isn't the only thing to be concerned about when turning the fans of a GPU off...

[Majestic]

Exactly. The Nvidia TDP is much more "optimistic" than the AMD TDP. According to the TDPs, the R9 270X should draw more power than the GTX 970, but as I've demonstrated above, in reality the GTX 970 draws significantly more power than the R9 270X.

 

You're missing the point. And AMD generally underdefines their power usage. R9-285 is rated at 190W TDP but can use 200W actual wattage. Whilst the 960 is defined 120W TDP but only uses about 80 when you properly test it (intercepted power PCI-E and PEG)

 

Source; http://www.pcper.com/reviews/Graphics-Cards/NVIDIA-GeForce-GTX-960-2GB-Review-GM206-199/New-Level-Power-Testing

 

GTX 760 is rated at 170W, but uses way less. And we're not even talking about the 295X, which uses >500W and exceeds all kinds of ATX regulations and specifications. That 960 was overclocked to 1500something btw.

 

And again (the important bit), the card was designed for having a fan blowing on the PCB + components. You can't just fucking disable the fans and think nothing will happen. Stop telling people to do this. Not without having knowledge of what you're doing and making sure the components are up for it.

 

This is the same reason fullcover blocks exist, and the AIO GPU-coolers require a shroud to blow on the components. 

[Sakkura]

You're missing the point. And AMD generally underdefines their power usage

 

As I've just proven, Nvidia is much more aggressive about "underdefining" their power usage. They put a lower TDP on the GTX 970 than AMD put on the R9 270X, but in reality power consumption is the other way around.

[Majestic]

As I've just proven, Nvidia is much more aggressive about "underdefining" their power usage. They put a lower TDP on the GTX 970 than AMD put on the R9 270X, but in reality power consumption is the other way around.

 

Stop this damn cherrypicking and get the hint already, you're being obnoxious. You miss the point entirely and go on about this NVIDIA vs. AMD TDP red herring.

You're not even making sense, the GTX 970 also uses less than advertised. You're just comparing two totally different cards and mixing TDP with power consumption.

 

Thermal design power does not reflect the power consumption. It's an indicator for the cooling system that should be used. Not all power consumption is tranfered to heat, that depends on the chip and it's efficiency.

[Sakkura]

Stop this damn cherrypicking and get the hint already, you're being obnoxious. You miss the point entirely and go on about this NVIDIA vs. AMD TDP red herring.

You're not even making sense, the GTX 970 also uses less than advertised. You're just comparing two totally different cards and mixing TDP with power consumption.

 

 You're the one missing the point and making all kinds of strawman arguments.

 

You said:

 

 

Except, Maxwell can do this because of the low power consumption.

 

Which it can... but the R9 270X has a lower power consumption than the GTX 970 that does this. And no, this is not an Nvidia vs. AMD thing from my side. I could just as well make the same argument in favor of a GTX 550 Ti. That thing could be passively cooled too. I wouldn't want to use one... but it would work. On Fermi, not the Maxwell architecture you claim is somehow the only GPU architecture capable of passive cooling.

 

I have shown a graph with the power consumption of the R9 270X and the GTX 970. The R9 270X drew significantly less power than the GTX 970, even though the GTX 970 has the lower TDP of those two cards. Yet you continue to claim that "the GTX 970 also uses less than advertised", as if the measured power consumption of >170W is less than the 145W TDP.

 

But okay, if that graph wasn't good enough, we can have another:

 

[Majestic]

@Sakkura That power consumption is calculated, not measured. They make rough stabs at what the platform powerconsumption is and substract that from the system power consumption they measured to get the card's power consumption. This is because it's extremely difficult to measure actual power consumption without making custom intercept boards. Those PCper graphs I displayed are intercepted power, and are thus much more reliable. The 970, 980 and TitanX all have slightly lower power consumption than TDP suggests on average. But that's not really the point i was trying to make when I said maxwell has lower TDP. The reason this is important is that they were THUS designed for 0dB operation because manufacturers felt it interesting to do so where they haven't done this before.

 

That quote you linked was meant to be compared to similarly performing cards. The 970 uses significantly less than a 290 but performs the same. The gist of this is, they're operating at different pricepoints. 270X cards have such little margin on them, that the chance of them fitting passive VRM cooling on them is relatively smaller, especially the cheap ones. For example the HIS R9 270 iPower IceQ X2 Boost has no VRM heatsinks, neither does the DirectCU II, however the VaporX TOXIC does. It depends on the card and should therefor be checked prior to just firing up MSI afterburner and disabling the fans like a total lemon.

 

The grand point; don't shut down fans on a graphics card without checking the board layout and adding heatsinks where needed. I later also explicitly said the same for maxwell so i'm not discriminating. You're just hit on the fanboy feels and somehow have some vinidication to settle.

 

Also, i never misrepresented your argument, I don't think you know what a strawman means :') I may have represented my argument better at the start, that's the only complaint that's valid.

[Sakkura]

@Sakkura That power consumption is calculated, not measured. They make rough stabs at what the platform powerconsumption is and substract that from the system power consumption they measured to get the card's power consumption.

 

No.

 

Here's a quote from the Techpowerup article:

 

 

For this test, we measure the power consumption of only the graphics card via the PCI-Express power connector(s) and PCI-Express bus slot. A Keithley Integra 2700 digital multimeter with 6.5-digit resolution is used for all measurements. Again, the values here only reflect the card's power consumption as measured at its DC inputs, not that of the whole system.

 

And here's a quote from the Tom's Hardware article that yielded very similar numbers:

 

 

We use a riser card on the PCIe slot (PEG) to measure power consumption directly on the motherboard for the 3.3 and 12 V rails. The riser card was built specifically for this purpose.

 

In addition, we separately measure the voltage and current at each of the two individual PCIe power connectors.

 

 

Both of those sites use more sophisticated hardware than PC Perspective does, and they've done so for a longer period of time. PC Perspective is just getting started with the direct measurement of GPU power consumption. At the launch of the R9 290X, PC Perspective only measured the total system power consumption. Techpowerup and Tom's Hardware were already measuring the isolated GPU power consumption at that point.

 

As for misrepresenting my argument, you did. You said:

 

 

You can't just fucking disable the fans and think nothing will happen. Stop telling people to do this

 

I never told anyone to do this. I simply pointed out that one of your claims was wrong, and I have provided clear proof from two independent, reliable sources.