Our testing includes idle and load temperature measurements, in addition to acoustic benchmarks. We're using Battlefield 3 as our graphics load, with the game running under the Ultra quality preset at 1920x1080 for 10 minutes in the Operation Swordbreaker level. We log temperatures using GPU-Z, and measure noise two feet from the back of the case with a CM-130 SPL meter. Unfortunately, ambient noise is a challenge in our real-world test lab; we recorded 40.5 dB(A).
While the Arctic Accelero Xtreme fans are powered by headers on the graphics card and should vary rotational speed in response to thermal load, Deepcool's Dracula employs power supply or motherboard fan headers. We're using 12 V leads on the motherboard to drive constant RPMs.
We also plan to measure how these coolers handle a GPU overclock to 1100 MHz with a 1.2 V setting. This increase should generate significantly more heat than the stock settings.
| Test System | ||||||||
|---|---|---|---|---|---|---|---|---|
| CPU | Intel Core i7-3960X (Sandy Bridge-E), 3.3 GHz @ 4.25 GHz , Six Cores, LGA 2011, 15 MB Shared L3 Cache, Hyper-Threading enabled. | |||||||
| Motherboard | ASRock X79 Extreme9 (LGA 2011) Chipset: Intel X79 Express | |||||||
| Networking | On-Board Gigabit LAN controller | |||||||
| Memory | Corsair Vengeance LP PC3-16000, 4 x 4 GB, 1600 MT/s, CL 8-8-8-24-2T | |||||||
| Graphics | Radeon HD 7970 3 GB GDDR5 | |||||||
| Hard Drive | Samsung 470-series 256 GB (SSD) | |||||||
| Power | ePower EP-1200E10-T2 1200 W ATX12V, EPS12V | |||||||
| Software and Drivers | ||||||||
| Operating System | Microsoft Windows 7 x64, Service Pack 1 | |||||||
| DirectX | DirectX 11 | |||||||
| Graphics Drivers | Catalyst 12.7 Beta | |||||||
| Benchmarks | ||||||||
| Battlefield 3 | Operation Swordbreaker, Ultra settings, 1920x1080, 10 minutes | |||||||
This surely looks impressive (giant graphics card and oversize heat cooler), but is this "eye candy" for the technically inclined PC enthusiast really moving forward, or just another pile of copper pipes sold at a price established out of pure value perception? This article got me thinking... Are we unknowingly creating a market demand for cooling products that make little sense in the grand scheme of things, nor shows little technological advancement? Why do we get so excited when a graphics card becomes so hot during peak operation that it requires cooling beyond standard specification. In engineering terms, any system that transforms such a large amount of electrical energy into heat as a side effect would be considered inefficient. By creating a market for "aftermarket" cooling, we do not only show our tolerance for inefficiency, but also create a booming demand for lackluster "solutions".
But, would be nice to see the coolers compared to some mainstream solutions. IE the HIS IceQ X2 or Sapphire Toxic, etc. etc.
But, would be nice to see the coolers compared to some mainstream solutions. IE the HIS IceQ X2 or Sapphire Toxic, etc. etc.
btw nice article
This surely looks impressive (giant graphics card and oversize heat cooler), but is this "eye candy" for the technically inclined PC enthusiast really moving forward, or just another pile of copper pipes sold at a price established out of pure value perception? This article got me thinking... Are we unknowingly creating a market demand for cooling products that make little sense in the grand scheme of things, nor shows little technological advancement? Why do we get so excited when a graphics card becomes so hot during peak operation that it requires cooling beyond standard specification. In engineering terms, any system that transforms such a large amount of electrical energy into heat as a side effect would be considered inefficient. By creating a market for "aftermarket" cooling, we do not only show our tolerance for inefficiency, but also create a booming demand for lackluster "solutions".
this applys to all mid-high end nvidia/ati(amd) video cards
That's how you transfer heat from the shim to the unmodified Accelero III. I wonder if JB Weld would work better...although that would permanently attach the shim to the Accelero III.
@cilliers
The value is in the noise reduction at load. These processors run hot because they are doing a great deal of work pushing electrons around. Consider that incandescent bulbs work the same way - the friction causes the filament to get so hot that it glows. If you don't want a thermally hot/power hungry card for philosophical reasons, then don't buy one.
@W(h)yKnott
I imagine that "Dracula" is intended to connote sucking the heat away from the 79xx. The fact that these tests show that they are relatively inefficient at doing so makes for a humorous double entendre, like your handle.
Unless you don't like not burning the GPU, pretty much. You might get away without it, but temps would be far higher. Maybe if you really lapped the cooler and shim you could get away with it, but I'd doubt that using no thermal paste at all would be a good idea even in that situation.
Absolutely. If you don't, contact wouldn't be uniform and you'd have dead spots.
Thermal paste is important, but the trick is to use as little as necessary, not to slather on gobs.
Its very unlikely that someone with such card is going to have air cooling "upgrade", since the WC setups are afordable. If that person wants to get better cooling solution, he wouldnt spend money on air cooling, just throwing on it an radiator or connect the GPU block to an existing radiator would be sufficient, better for looks, and better performance.
If I had a card that used 1000 watts of power but was ten times faster than the Radeon 7970 in every way, it would still be the most energy efficient graphics card in the world today. I also don't think that after-market VGA cooling is a booming market for lackluster solutions. The after market VGA cooling industry probably isn't booming because even the minority of overclockers in this world tend to not use an aftermarket cooler on their graphics card(s). Even then, just because there are some lackluster solutions doesn't mean that they sell nearly as well as the good solutions.
I buy aftermarket coolers for noise reduction only, I can't stand a loud computer. Now my machine is so quiet, it's hard to tell whether it's off or on, even when it's under load.
And having more efficient hardware wouldn't help with the noise, the OEMs would just put smaller coolers and smaller/faster fans, so noise output would be similar.
Review the dracula with two 140mm fans or three 120mm fans as the design was intended, to show optimal performance.
Overclock as far as you can with each design so we can see the value added to the card.
Without that information this entire article is pointless, and leads to dumba$$ comments like 'why dracula, because it sucks?'
How about review the design as it was meant to be used?
*And what kind of a tech journalist doesn't have a few spare 120/140mm fans laying around? WTF!!!
The 7970 doesn't NEED more, temps were awesome. Noise, space, and cost also comes into play.
You think a slight temperature difference between the Dracula and Accelero will affect the overclock? That's just silly. Limits will be set by voltage unless temps are astronomical.
Overdramatic complaint is overdramatic.
OC most silicon chips, and they will lose efficiency.