GTX 660 Ti Roundup (ASUS, EVGA, Gigabyte, Galaxy, MSI)

[SKYMTL]

Performance Oddities Investigated

Performance Oddities Investigated

Throughout our in-game testing some interesting performance discrepancies were noticed. Some of these focused upon the ASUS GTX 660 Ti TOP which had the highest on-paper boost clocks but failed to consistently beat Gigabyte’s offering and at times even struggled to stay ahead of the MSI Power Edition. We also noticed that as we progressed through our four benchmark runs of each game, the EVGA GTX 660 Ti SC and NVIDIA’s reference card both showed slightly lower framerates from one repetition to the next. Granted, ASUS’ card still came out as the overall performance winner once the dust settled and EVGA’s provided more than adequate performance but we were left wondering: what was going on?

In our investigation the primary focus was upon each card’s clock speeds being the culprit and those assumptions bore fruit in short order. According to our findings, the ASUS TOP seemed to be unable to consistently hit the upper ranges of its Boost frequencies. Instead, it tended to fluctuate up and down quite a bit, with a few peaks that thrust up into extremely high clock speed ranges. Comparing and contrasting these results to those from MSI, Gigabyte and Galaxy cards puts things into stark contrast since these other products hit a mark and stay there throughout the benchmark, sometimes resulting in higher performance.

The two reference-based cards also showed an interesting side of their personas as their clock speeds gradually decreased throughout the test, resulting in the aforementioned performance drop-off up until their fans increased speed a bit. Now, the difference between maximum frequencies and where these cards end up after a few minutes is infinitesimal in the grand scheme of things and an end user will never notice anything but on paper at least, you’ll be losing a few frames per second here and there.

Within Batman, our results for the ASUS card were actually well in line with expectations as it provided class-leading framerates regardless of its constant clock speed dance. However, once again the EVGA SC and reference clocked cards exhibit a tendency to step back their clock speeds but this time it looks like the Power Limit is stepping in as well. However, be it TDP, Power Limit, temperatures or some combination thereof, we are seeing a general downgrading of Boost values over the course of our benchmark.

This brings whole exercise could bring up some worrying points about benchmarking NVIDIA’s Kepler-based cards in reviews (and charts) where every single FPS counts. Sites benchmarking with a single run or shorter sequences will likely achieve the “best” results rather than realistic performance. Luckily, we have been able to avoid this issue by using four run-throughs of every benchmark, each with somewhat long testing times. We’ll have a full article looking at GeForce Boost and AMD’s equivalent in the coming weeks but for the time being, this is certainly food for thought.

 

[SKYMTL]

Temperatures & Acoustics / Power Consumption

Temperature Analysis

For all temperature testing, the cards were placed on an open test bench with a single 120mm 1200RPM fan placed ~8” away from the heatsink. The ambient temperature was kept at a constant 22°C (+/- 0.5°C). If the ambient temperatures rose above 23°C at any time throughout the test, all benchmarking was stopped. For this test we use the 3DMark Batch Size test at its highest triangle count with 4xAA and 16xAF enabled and looped it for one hour to determine the peak load temperature as measured by GPU-Z.

For Idle tests, we let the system idle at the Windows 7 desktop for 15 minutes and recorded the peak temperature.

While there really wasn’t that much to distinguish any of these cards apart from the others in the performance testing, one would have hoped temperatures would be a deciding factor in some cases. Unfortunately, that isn’t the case since every one of the custom GTX 660 Ti’s provided excellent cooling results, beating the reference design by a significant margin. The only odd man out was the EVGA SC which actually posted higher than reference results but remember, our “reference” card is just the SC with a different BIOS so it still makes use of the upgraded internal heatsink.

Acoustical Testing

What you see below are the baseline idle dB(A) results attained for a relatively quiet open-case system (specs are in the Methodology section) sans GPU along with the attained results for each individual card in idle and load scenarios. The meter we use has been calibrated and is placed at seated ear-level exactly 12” away from the GPU’s fan. For the load scenarios, a loop of Unigine Heave 2.5 is used in order to generate a constant load on the GPU(s) over the course of 20 minutes.

If we had to split hairs, ASUS obviously won this round with some incredibly low acoustics even though it does have extremely high clock speeds. However, the competition isn’t all that far behind and regardless of what this chart shows, we can almost guarantee that the GTX 660 Ti cards from MSI, Gigabyte and Galaxy will sound just as quiet regardless of what our highly sensitive meter says. Once again, EVGA’s SC edition trails the pack but it still provides an extremely quiet gaming experience.

System Power Consumption

For this test we hooked up our power supply to a UPM power meter that will log the power consumption of the whole system twice every second. In order to stress the GPU as much as possible we once again use the Batch Render test in 3DMark06 and let it run for 30 minutes to determine the peak power consumption while letting the card sit at a stable Windows desktop for 30 minutes to determine the peak idle power consumption. We have also included several other tests as well.

Please note that after extensive testing, we have found that simply plugging in a power meter to a wall outlet or UPS will NOT give you accurate power consumption numbers due to slight changes in the input voltage. Thus we use a Tripp-Lite 1800W line conditioner between the 120V outlet and the power meter.

The GTX 660 Ti isn’t exactly the most efficient card in NVIDIA’s current lineup and with a bit of overclocking, it consumes more power than a GTX 670. Unfortunately, that causes its performance per watt to fall somewhat in comparison to other SKUs.

 

[SKYMTL]

Overclocking Results

Overclocking Results

Overclocking NVIDIA’s Kepler architecture can be a complicated process since it includes modification of Power Limits, clock speeds and Boost offsets. However, that shouldn’t dissuade anyone from trying to get the most out of their GTX 660 Ti.

For these tests, we used ASUS’ excellent GPU Tweak EVGA’s Precision, MSI’s very popular Afterburner and Galaxy’s little known but highly capable Xtreme Tuner Plus. Unfortunately, Gigabyte still doesn’t have their own overclocking utility so we ended up using Afterburner to overclock their GTX 660 Ti OC.

In all cases we added 100mV to the core and GDDR5 while MSI’s Power Edition was the only card to support a 50mV bump in PLL voltage as well. As with all other Kepler-based overclocking experiences, we set the Power Limit to the absolute maximum while the fans were left at their default speed. All results are considered 24/7 stable.

As we can see, ASUS’ TOP edition wins this section hands down. It seems like they have built higher power limits into their card which allows it to Boost to extremely high frequencies when given enough voltage. The products from MSI and Galaxy are extremely close behind while EVGA and Gigabyte bring up the rear.

 

[SKYMTL]

A Prelude to Our Conclusions / ASUS Final Thoughts

A Prelude to Our Conclusions

The pre overclocked GTX 660 Ti cards in this roundup have provided a unique cross section of cooling upgrades and performance enhancements but we have learned that certain attributes come at a steep cost. In some ways, these custom products have changed our perspective about NVIDIA’s latest graphics card, both positively and in some cases negatively. With that being said, in many situations the visible performance differences between the products in this roundup are slim to none so keep that into account while you read through these conclusions.

NVIDIA’s Boost capability has (so far) proven to be beneficial to end users since it allows Kepler-based cards to reach clock speeds they wouldn’t normally achieve. However, there is a downside due to the fact that Boost is tied at the hip to Power Limits, TDP and cooling performance so its actual reach can vary wildly from one card to another. In some cases a card that may look faster on paper but will end up tied with a less powerful (and less expensive) alternative. Temperature in particular impacts boost in a big way as evidenced by the falling clock speeds of EVGA’s SC and the reference board. Meanwhile, the custom cooled cards were able to stay at lower temperatures for longer periods of time, resulting in clock speed stability and higher performance over longer periods of time. Also take note that nearly all of these pre-overclocked cards typically consume as much or more power than a GTX 670.

Then there’s the thousand pound gorilla in the room: AMD’s HD 7950 Boost. Against it, NVIDIA’s board partners –particularly those with $315 and higher products- are coming up flat in resolutions above 1920 x 1200. While their cards are an excellent solution for lower end monitors, they just can’t compete on a price / performance level above that. This goes to prove that higher clock core speeds just can’t make up for a lack of bandwidth, which brings us back to the bewildering lack of GDDR5 overclocks on every one of these cards. While it may seem odd that board partners are shying away from the very thing that could truly improve overall performance, they are actually protecting their higher end GTX 670 SKUs. Whether or not this benefits end users is debatable but even without the aforementioned memory overclocks, every one of these GTX 660 Ti cards is appealing in its own way.

Conclusion: ASUS GTX 660 Ti DirectCU II TOP

The ASUS GTX 660 Ti DirectCU II TOP provided the most noteworthy experience of this roundup, mostly due to its flip-flopping from one test to another. When a game allowed its Power Limit some breathing room, performance was easily the best within this roundup while maintaining low temperatures and an enviable acoustical profile. However, in certain applications, its clock speeds bounced around quite a bit, causing framerates to drop ever so slightly against the competition.

With that being said, the TOP is currently the card to get if you plan on overclocking. Its BIOS-determined Power Limit is stratospheric which resulted in clock speeds that far exceeded what the competition could offer and the DirectCU II heatsink has plenty of thermal mass to spare. A price of $325 may at first seem to be on the expensive side, but you’ll be paying less than 10% more than a reference card for a feature rich, quiet and overclocker friendly GTX 660 Ti that can easily run with a GTX 670 in more than a few situations.

 

[SKYMTL]

Conclusions: EVGA & Galaxy

EVGA GTX 660 Ti SC

In a market that’s currently dominated by upgraded versions of the GTX 660 Ti, EVGA’s SC edition may not stand out as being unique but it combines good performance with a meager $10 premium over a bone stock card. Granted, gamers likely won’t notice the meager in-game framerate improvement but for $10 we really can’t complain. We also appreciate EVGA’s efforts disguise an upgraded heatsink inside of a reference shroud that exhausts hot air outside of the case. Every other design in this roundup dumps hot air into its immediate surroundings, increasing in-case temperatures.

The problem for EVGA isn’t their price being too high but rather the competition’s being so low. The $309 GTX 660 Ti SC finds itself battling similarly priced custom cards (from Galaxy and MSI) that boast heavily enhanced PWM designs, higher clock speeds and upgraded heatsinks. The fact that it goes up against these odds without the ammunition of EVGA’s lifetime warranty coverage and step-up program makes matters a bit worse. Granted, the SC has a much sought-after transferable warranty which certainly adds to its value and will be a huge selling point for anyone that upgrades frequently. Unfortunately, this could go largely unnoticed in such a competitive environment where many consumers are looking for instant gratification.

Galaxy GTX 660 Ti GC

Galaxy may not have with widespread recognition of the Gigabytes and ASUS’ of this world but they brought their a-game to the table with the GTX 660 Ti GC. This card may not provide ultra high clock speeds, nor does it have an impressive feature list like the DirectCu II TOP but it strikes a delicate balance between price and performance. Galaxy was careful not to intrude too far into the HD 7950’s market territory but they have still provided enough value to sway people away from AMD’s card.

At just $309, the GC edition is no more expensive than EVGA’s GTX 660 Ti SC but provides the lowest temperatures of any GTX 660 Ti we’ve tested, very quiet operation, impressive overclocking results and framerate improvements that are noticeable in some situations. As such, it currently represents one of the best values in a highly competitive market.

 

[SKYMTL]

Conclusions: Gigabyte & MSI

Gigabyte GTX 660 Ti OC

While we aren’t necessarily picking a winner in this roundup, Gigabyte’s showing was quite surprising since –on paper at least- it doesn’t have the highest clock speeds. What it has though is consistency. Unlike ASUS’ DirectCu II TOP, when the OC’s Boost values hit a certain level, they typically stayed there without much variance from one game scene to the next. As a result, it was in a statistical dead heat with a card that’s slightly more expensive.

Some may argue that ASUS’ feature set and impressive overclocking headroom have Gigabyte on the defensive and we would have to agree on that front. However, it is impossible to make a clear and concise recommendation of one card over another due to the similarities involved. While the lack of overclocking software is a notable exclusion from Gigabyte’s offering and the heatsink assembly feels a bit cheap, it still provides a cool, quiet and composed gaming experience while usually matching a GTX 670’s performance.

MSI GTX 660 Ti Power Edition

Much like Galaxy’s GTX 660 Ti GC, MSI’s Power Edition doesn’t stride within the upper or lower performance rungs within this roundup but it provides a consistent gaming experience. Its average framerates were within spitting distance of more expensive cards from Gigabyte and ASUS without drastic clock speed variations or flimsy heatsinks. Speaking of cooling ability, once again the Twin Frozr design has proven why it is one of the most highly regarded thermal solutions around. It went about its job quietly and provided enough capacity that the Power Edition’s fans remained at their default speed throughout our overclocking tests.

Where MSI’s card really distinguishes itself is in the value category. With in-game performance that nearly matches some more expensive cards, triple overvoltage capabilities and such a great heatsink design, a price of $309 is simply too good to pass up. In such a tight situation, a few bucks can make one hell of a difference and for the GTX 660 Ti Power Edition, that difference is award worthy.

​