ASUS GTX 670 DirectCU II Mini Graphics Card
The GK104 Kepler Architecture
The GTX 670 is based on the same GK104 architecture you saw on the 680, and even 770/760 we just recently looked at as well. Although the new 7 series has a bit more horsepower in something like the 770 card, the 670 Mini holds a very special place in the fact that it can pull off fitment that most cards simply cannot do while also allowing huge performance potential.
The GTX 670 DCII Mini come to market with a street price of $324.99 via newegg at the time of this writing. This places it in a nice price area compared to other 670 models so depending upon how this card performs it could be a real steal.
A Look at the GTX 670 DCII Mini
Here you can see the full breakdown of the DirectCU Mini card and all of the components that make up the custom cooling unit. As you can see here unlike most DirectCU coolers we have seen recently this does not use heatpipes and instead employs a Vapor chamber style bladder for the GPU cooling duties. The Vapor Chamber has served very well in many applications so I am very interested to see how it performs in this space conscious build.
Made small for a reason
The DirectCU Mini card was made to be able to fit in the much smaller builds such as the tiny Mini ITX towers as seen above. Normally in order to fit in a much smaller system like this you had to go with a much smaller card which normally meant taking a hit to your Graphics card performance capability.
SAP (Super Alloy Power)
ASUS has great pride in the components they infuse into their graphics offerings, very much like the motherboards we have covered previously. ASUS has what’s called a SAP, or Super Alloy Power, which covers the components in the very VRM we are discussing. The components are hand picked for quality and performance based around not only power delivery but efficiency as well. The components in the SAP solution are super durable for a rated lifespan of up to 2.5x what reference cards would offer.
Here you get a quick rundown of the VRM explaining why key components were chosen in order to give you a more stable VRM solution and a better faster card.
One point of interest for us was the large “Direct Power” gold bar which appears to be just a direct link for the power to travel easily with a large dedicated surface to enable easy power flow without it being bottlenecked or having a restricted path which can cause heat. The restriction is called impedance, and it can be measured in ohms. Less resistance means more freely flowing power where more resistance is much like a traffic jam in that the power has a tougher time passing therefore creating hardware’s worst enemy which is heat.
Here we have some good detailing directly from ASUS, showing the improvement of the SAP choke component and how it is superior to other designs implemented presently. Other model chokes can make a whine or screeching sound due to the coil vibration, something we’re sure many enthusiasts and overclockers have heard before. The ASUS design has a sealed/filled core to inhibit vibration, which means noise free operation under extreme loading conditions. Another thing to remember is movement makes heat and friction, so less movement will mean cooler running components.
Here ASUS shows off its SAP selection for MOSFET components. These are designed to not only be more efficient but also have a higher power handling capability, which allows them to have extra head room for overclocking, should you really want to push the limit. Also note, the small package design of the SAP component as it is much smaller than the standard or generic MOSFET used on some standard or reference model cards. This allows for a larger VRM design to fit into a smaller space so that you can pack more punch without effecting the tight real estate within the chassis.
The SAP Solid state cap selection are Japanese units with a much higher MTBF than what you may see on other designs. These units are designed to have a very high expected lifetime, which allows users to rest assured that in their gaming rig with these components should have a nice happy life with a very low chance of failure for quite a long time (up to 2.5x as longer than competing components). For those of us that like to throttle our cards and really put the squeeze on them, this means we have the capability built in as the components are solid and reliable and ready to take a pounding.
Here you can see how the design for the 780 goes from the initial artistic design to the finished product and how the emphasis while being on performance also always exudes a feeling of style when picking up a DirectCU model card.
Cooltech Fan Design
Here is something pretty neat and I think ASUS deserves innovation points here is with the new Cooltech fan which from my internal discussions with them I have been told had been in development for over a year from working with it to real implementation of the unit. This fan is in essence a blower fan in the center like you have seen on reference Nvidia GPUs in the past up to present and the outer blades are flower shape blades so that you get not only the power of an axial fan but the blower as well. Thew blower fan shoots straight outward laterally while the flower blade fans tend to push air through it or down and in through the cooling grid. What this means is that now when the fan is sunk down into the cooling fin array you not only shoot air though the fins downward and onto the PCB and mosfet components but you shoot air out laterally to utilize the cooling fin area next to the fan maximizing cooling potential.
I know some will ask why not use two of these fans so I asked ASUS and their answer was what i assumed already which is that dual blower fans will actually fight against each other as the lateral air will crash into each other and cause unnecessary noise and turbulence so therefor a single unit handles that cross-flow with a standard fan in the second position. As you can see from the ASUS supplied image with the “Rice Test” they test a standard curve bladed fan design vs the Cooltech and you can see the airflow power at same RPM and the results are excellent for the Cooltech model.
These are the kind of innovations I have come to expect from ASUS and to also clearly explain why it was done is what going the extra mile is all about as users who know why it was used and not just that it is there is ultimately gonna lead to much higher consumer confidence.
ASUS offers its own software utility and its functionality has grown with generations just like the cards it supports. Of course the card can be overclocked but also voltages can be tweaked and within the GPU Tweak app, we can even launch the GPU info utility which is a full custom version of GPUz. This is a really cool feature and a benefit especially to benchmarkers, who we can definitely see using this card with some of the awesome tweaks we will discuss in the card overview.
There are many tools within the GPU Tweak menus some a little more hidden than others. First off directly form the GPU Tweak utility itself you can check for BIOS updates and even update from within the program. Anyone who’s ever updated a vBIOS before knows that having a tool that can do this in a mere few clicks is really handy.
The charts besides offering real time monitoring can also be setup to log thermals, voltages and clocks so that after benchmark or gaming runs you can see if there was any thermal issues or throttling that may have affected performance.
There is also a special ASUS-skinned version of GPU-Z that can be accessed as seen above when pressing the “GPU Info” button.
PCI-E Power LEDs
One really cool feature is that there are status LED’s on each PCI-E connector and these help you recognize that power is actively being sent to each connector of the card. The card employed today replaces the standard dual 6 pin PCIe connectors with a single 8 pin as shown in the images, therefore only a single Red or green LED will be used.
The green seen above indicate everything is working well. However if the PCI-E cable is not active or is not installed at all, it will give you the red light which tells you something is wrong. We could definitely see some value in this when diagnosing multi-card setups or even single card issues. A dead PCI-E power cable could cause you to pull your hair out without these kinds of indicators.
Existing Kepler features
Kepler features a dedicated H.264 video encoder called NVENC. Fermi’s video encoding was handled by the GPU’s array of CUDA cores. By having dedicated H.264 encoding circuitry, Kepler is able to reduce power consumption compared to Fermi. This is an important step for Nvidia as Intel’s Quick Sync has proven to be quite efficient at video encoding and the latest AMD HD 7000 Radeon cards also feature a new Video Codec Engine.
Nvidia lets the software manufacturers implement support for their new NVENC engine if they wish to. They can even choose to encode using both NVENC and CUDA in parallel. This is very similar to what AMD has done with the Video Codec Engine in Hybrid mode. By combining the dedicated engine with GPU, the NVENC should be much faster than CUDA and possibly even Quick Sync.
- Can encode full HD resolution (1080p) video up to 8x faster than real-time
- Support for H.264 Base, Main, and High Profile Level 4.1 (Blu-ray standard)
- Supports MVC (multiview Video Coding) for stereoscopic video
- Up to 4096×4096 encoding
According to NVIDIA, besides transcoding, NVENC will also be used in video editing, wireless display, and videoconferencing applications. NVIDIA has been working with the software manufacturers to provide the software support for NVENC. At launch, Cyberlink MediaExpresso will support NVENC, and Cyberlink PowerDirector and Arcsoft MediaConverter will also add support for NVENC later.
Click on the Images to View a Larger Version
Improved Software Experience
Nvidia supplied us with some really cool software and also showed us some cool stuff that we can now show off to all of you!
Here you can see the GeForce Experience program, which is actually a very cool free software for GeForce users. Once you install the program it will scan your system hardware and all installed games, and then optimize your game settings for the best experience.
This may not seem like much, but think of it this way: When you go in and just crank up the settings, are you really running the game the best that your system can? Most likely not. The GeForce Experience program is a better alternative to guessing which settings will provide a balance of performance and eye-candy. This automatically sets up your system to run well, so you don’t have to.
Do note that the GeForce Experience software has been out for BETA for some time, but as of now is moving into full user ready state.
GeForce Experience – Shadowplay
Many of us have used various different forms of recording software such as FRAPS to show our gameplay to the world. But anyone who has done this knows the limitations, as the files quickly get huge and can take up a lot of disk space very quickly.
Nvidia will be adding a feature to the GeForce Experience called ShadowPlay which is a game recording software but with a very special capability. It uses the NVENC native H.264 encoder on the Kepler based GPUs to encode the gameplay footage in real time so that the files are smaller right from the start and quality remains excellent.
The Shadowplay software has been delayed somewhat to allow fine tuning and optimizations but from what we have heard we should be seeing it very soon.
A Closer Look at the GTX 760 DirectCU II OC
Click Images to Enlarge
The ASUS box comes like the last few generation of TOP and OC cards with the red claw mark graphic, and a few key spec words. The rear covers the key feature information and a little more specific detailing.
Click Images to Enlarge
The box has a accessory box internally which sits on top of the card. The accessories with the card are simple you get a dual PCIe 6 pin into a PCIe 8 pin adapter since the card employs a single 8 pin only. Also a driver disc and quick guide but a card simply plugs in and goes so we don’t think we have ever looked at the quick guide honestly.
Click Images to Enlarge
The card itself truly is small and to know it holds the power of a full 670 card in such a small package is hard to explain as it just feels like it should be a much weaker card but knowing its a 670 just feels odd to me. Overall the cooler looks awesome and vents out radially so keep that in mind as the chassis airflow will be evacuating the heat.
The display connectivity is about standard for what we have seen on the Kepler based lineup:
- 2x DVI
- 1x HDMI
- 1x Displayport
This all-digital connector arrangement allows for maxing out the GPU with up to 4 displays and even a surround plus accessory display array should the need arise.
Click Images to Enlarge
The GTX 670 DirectCU Mini card employs a single PCIe 8 pin connection which we really like as it allows less cable clutter from a single easy connection. Also do note that the connector is rotated to allow easier releasing of the clip should the PCIe power cable need to be removed you don’t have to try and squeeze your fingers between the card and the cooler.
Here also you see the Direct Power device which allows direct power transfer on the PCB versus having to travel from point A to point B with less resistance, lets just think of it as a freeway for GPU power.
Click Images to Enlarge
Pulling the cooler you first see the vapor chamber copper layer peering through the aluminum heatsink plate. The ALU Heatsink plate works to offer supplemental cooling to the VRM along with the memory on the front side of the card, and as you can see from the paste spread on the GPU the contact surface is quite flat and offers nice coverage on the GPU die.
Click Images to Enlarge
Here we have the SAP component VRM responsible for feeding the GTX 670 DirectCU Mini card. The VRM is capable of handling up to 30% more voltage according to ASUS which means that the overclocking headroom should be very nice for this mini model.
The Hynix Memory IC’s used in the mini card are 6GHz rated but as previous 670′s have shown im betting you can get a fair bump above that rating when overclocking for even more performance.
The OS we use is Windows 8 Pro 64bit with all patches and updates applied. We also use the latest drivers available for the motherboard and any devices attached to the computer. We do not disable background tasks or tweak the OS or system in any way. We turn off drive indexing and daily defragging. We also turn off Prefetch and Superfetch. This is not an attempt to produce bigger benchmark numbers. Drive indexing and defragging can interfere with testing and produce confusing numbers. If a test were to be run while a drive was being indexed or defragged, and then the same test was later run when these processes were off, the two results would be contradictory and erroneous. As we cannot control when defragging and indexing occur precisely enough to guarantee that they won’t interfere with testing, we opt to disable the features entirely.
Prefetch tries to predict what users will load the next time they boot the machine by caching the relevant files and storing them for later use. We want to learn how the program runs without any of the files being cached, and we disable it so that each test run we do not have to clear pre-fetch to get accurate numbers. Lastly we disable Superfetch. Superfetch loads often-used programs into the memory. It is one of the reasons that Windows Vista occupies so much memory. Vista fills the memory in an attempt to predict what users will load. Having one test run with files cached, and another test run with the files un-cached would result in inaccurate numbers. Again, since we can’t control its timings so precisely, it we turn it off. Because these four features can potentially interfere with benchmarking, and and are out of our control, we disable them. We do not disable anything else.
We ran each test a total of 3 times, and reported the average score from all three scores. Benchmark screenshots are of the median result. Anomalous results were discounted and the benchmarks were rerun.
Please note that due to new driver releases with performance improvements, we rebenched every card shown in the results section. The results here will be different than previous reviews due to the performance increases in drivers.
|CPUs||Intel XEON E5-2687W|
|Motherboards||ASUS Rampage IV Extreme|
|Ram||Kingston HyperX Beast 64GB (8x8GB) 2133Mhz 11-12-11 Quad-Channel Kit|
|CPU Cooler||Custom Liquid Cooling|
|Hard Drives||8x Western Digital RE4 2TB 7200RPM 3Gb/s Hard Drives (Raid 5)|
|SSD||3x Kingston HyperX 240GB SATA III 6Gb/s SSD|
|GPU||ASUS GeForce GTX 670 DirectCU II OC Mini 2GB Video CardNvidia GeForce GTX 770 2GB Video CardNvidia GeForce GTX 780 3GB Video CardNvidia GeForce GTX 680 2GB Video CardNvidia GeForce GTX TITAN 6GB Video CardNvidia GeForce GTX 690 4GB Video Card|
|GeForce Drivers||320.14 (GTX670, 680, 690 and TITAN) 320.18 (GTX 780, 770)320.49 (GTX 760)|
|PSU||Thermaltake Toughpower XT 1475W Gold|
|Mouse||Tt eSPORTS Theron Gaming Mouse|
|Keyboard||Tt eSPORTS Meks G1 Illuminated|
Synthetic Benchmarks & Games
We will use the following applications to benchmark the performance of the ASUS GTX 670 DirectCU II OC Mini video card.
|Far Cry 3|
|Unigine Heaven 4.0|
|Batman Arkham City|
|Dirt 3 Showdown|
|Metro Last Light|
|Sniper Elite V2|
Here you can see in 3DMark 11 the ASUS GTX 670 Mini pulls in right behind the 770 and ahead of the current Gen 760 which is a pretty good result and a bit better than a reference 670 we tested quite some time ago.
In 3DMark Firestrike and Extreme, the ASUS 670 DirectCU Mini pulls similar here right above the 760 and below the 770 but still a bit higher than a reference 670. This speaks well for the gaming ability of this small card.
Unigine Heaven 4.0
Unigine Heaven is a benchmark program based on Unigine Corp’s latest engine, Unigine. The engine features DirectX 11, Hardware tessellation, DirectCompute, and Shader Model 5.0. All of these new technologies combined with the ability to run each card through the same exact test means this benchmark should be in our arsenal for a long time.
Heaven shows a little about 2FPS over a 760 here and about 6 below a 770 so its about standard for a lightly overclocked 670.
Far Cry 3
Moving over to Far Cry 3, we used a pre determined path and re-ran each card to monitor average FPS across the complete run. Each run was completed multiple times to ensure continuity and accuracy. Here the ASUS 670 pulls once again about 2FPS over the 760 and about 75 below the 770.
Even after many years, Metro 2033 is still one of the most demanding DX11 video games available to really push the limits of your system. Here the ASUS 670 shows the real abuse of Metro as it scrapes by with a single FPS over the newer 760 but it is still better than the reference 670 and very good performance honestly for teh stress metro puts on the graphics card.
Batman Arkham City
Batman Arkham City is a fun game and utilizes FXAA fully. It is nice to see the full screen anti-aliasing getting picked up by popular titles. When it comes to performance, The ASUS 670 Mini pulls in between the 760 and 770 here and a nice little jump over what we’ve seen from reference clocked 670′s in the past.
Crysis 3 is much like the original Crysis in that playing it at max settings is painful for even the highest end hardware. Once again the ASUS 670 Mini pulls in right in between the 760 and 770.
Metro Last Light
Here we have Metro Last Light which is a brand new game added to our testing platform as it just came out. If you thought Metro 2033 was stressful, you have seen nothing yet. The particles and lighting, as well as the PhysX effects all add up to an amazing looking game that can put a stun even the most powerful system in its tracks. Here the ASUS 670 Mini pulls quite good performance over the newer 760 as it pulls 5 FPS lead. This is huge in the fact that Last Light is even more stressful than Metro 2033 and to see such a large difference shows that this card is packing some major punch.
Sniper Elite V2
Sniper Elite V2 is a fun game which, when first released, was quite stressful but still playable on most higher end cards. By now, most cards have surpassed it, and most of the cards now easily break the 60FPS super smooth average. The ASUS 670 mini pulls over 90 which means no matter what it simply will be playing smoothly.
Dirt 3 Showdown
Dirt 3 Showdown is a visually beautiful game with excellent lighting and overall motion blurring creating a real visual spectacle of a game. The ASUS 670 Mini pulls just under 60FPS average so figure gameplay will be nice and smooth with no or minimal stutter.
The new Kepler Architecture with GPU boost has been a bit of a bear when it comes to pushing the clocks. This is due to the throttling mechanisms put in place and the fact that it dynamically clocks. Unlike 500 series where we could set a static clock and it just applied the card, Kepler clocks according to many variables, which means that much more care must be shown when applying overclock settings and even monitoring it during the run.
Things such as the thermal target can play a big role in stonewalling your performance during an overclock session, as everything has to be adjusted accordingly to ensure expected performance levels are met.
The ASUS GTX 670 DirectCU Mini cranked right up when requested and we were ble to pull it from a base clock of 928MHz to 1100MHz which translated to an effective boost clock of over 1250MHz on the GPU. This is an increase of over 16 percent on the base clock of the GPU and not bad whatsoever when you consider this was attained while reaching a maximum loaded temperature of only 70C peaks.
Memory ran quite well moving from the 6GHz mark up and beyond 7GHz to a final stable speed of 7052MHz on the memory which really opened up the pipeline for some great GPU performance as you can see by the 3DMark 11 score shown above.
To measure the temperature of the video card, we used ASUS GPU Tweak (Validated with EVGA Precision X) and ran Heaven Benchmark in a loop to find the Load temperatures for the video cards. The highest temperature was recorded. After looping for 10 minutes, Heaven was turned off and we let the computer sit at the desktop for another 10 minutes before we measured the idle temperatures.
|GPU Temperatures||Temperature (Idle/Load)|
|Nvidia GTX 690||32C/81C|
|Nvidia GTX TITAN||31C/67C|
|Nvidia GTX 680||31C/72C|
|Nvidia GTX 780||30C/65C|
|Nvidia GTX 770||34C/79C|
|ASUS GTX 670 DCII Mini||33C/70C|
The DirectCU Mini cooling solution while being a radial design did handle temps quite well with its integrated vapor chamber. The fact that we found the limit of the GPU while barely seeing peaks of 70C tells me that in a standard gaming rig or even a cramped mini ITX or mATX build this card will really be up to the task of knocking out some awesome gameplay.
To get our power consumption numbers, we plugged in our Kill A Watt power measurement device and took the Idle reading at the desktop during our temperature readings. We left it at the desktop for about 15 minutes and took the idle reading. Then we ran Heaven Benchmark for a few minutes minutes and recorded the highest power usage.
Here we observed that the ASUS GTX 670 DirectCU Mini pulled very low power at idle even lower than the reference 760. Under load it pulled a bit more but was below the 770 model. Overall this means when browsing and light load the power consumption of your rig would be lower for even a lower power bill.
Our Final Thoughts
The ASUS GTX 670 DirectCU Mini card is what I consider an awesome idea and executed beautifully. The reason I say this is that smaller systems are really becoming more popular and all you need to do is attend a LAN to see this, so it comes naturally to see chassis and motherboard manufacturers pumping out solutions to fill this need, but in reality we have not seen many manufacturers thinking outside the box in regards to GPU and providing a compact yet high performance solution for these mini rigs. This is where ASUS stepped in and created this card and it simply works.
Knowing the market demands smaller systems and overall solutions that still can do some serious work ASUS took the ball and ran with it. I hope this spurs the market to create more compact solutions as that means more powerful systems taking up less and less space and that’s never a bad thing.
For anybody looking to build their next small form factor rig I really think you would have a tough tiome finding a better fir for a mini system than the ASUS GTX 670 DirectCU Mini.