NVIDIA GTX 760 2GB Review

[SKYMTL]

Onscreen Frame Times w/FCAT

Onscreen Frame Times w/FCAT

When capturing output frames in real-time, there are a number of eccentricities which wouldn’t normally be picked up by FRAPS but are nonetheless important to take into account. For example, some graphics solutions can either partially display a frame or drop it altogether. While both situations may sound horrible, these so-called “runts” and dropped frames will be completely invisible to someone sitting in front of a monitor. However, since these are counted by its software as full frames, FRAPS tends to factor them into the equation nonetheless, potentially giving results that don’t reflect what’s actually being displayed.

With certain frame types being non-threatening to the overall gaming experience, we’re presented with a simple question: should the fine-grain details of these invisible runts and dropped frames be displayed outright or should we show a more realistic representation of what you’ll see on the screen? Since Hardware Canucks is striving to evaluate cards based upon and end-user experience rather than from a purely scientific standpoint, we decided on the latter of these two methods.

With this in mind, we’ve used the FCAT tools to add the timing of partially rendered frames to the latency of successive frames. Dropped frames meanwhile are ignored as their value is zero. This provides a more realistic snapshot of visible fluidity.

The first batch of FCAT results shows performance very much in-line with expectations. The GTX 750 provides a perfectly smooth gaming experience without any stuttering unless framerates go below 35 or so. AMD's HD 7950 Boost behaves the same as within previous reviews mostly because AMD hasn't rolled out a new driver for the past few months but at least it doesn't display the major stuttering seen on multi-card solutions. We do however see their issues in Far Cru 3 highlighted once again.

 

[SKYMTL]

Onscreen Frame Times w/FCAT (pg.2)

Onscreen Frame Times w/FCAT (pg.2)

When capturing output frames in real-time, there are a number of eccentricities which wouldn’t normally be picked up by FRAPS but are nonetheless important to take into account. For example, some graphics solutions can either partially display a frame or drop it altogether. While both situations may sound horrible, these so-called “runts” and dropped frames will be completely invisible to someone sitting in front of a monitor. However, since these are counted by its software as full frames, FRAPS tends to factor them into the equation nonetheless, potentially giving results that don’t reflect what’s actually being displayed.

With certain frame types being non-threatening to the overall gaming experience, we’re presented with a simple question: should the fine-grain details of these invisible runts and dropped frames be displayed outright or should we show a more realistic representation of what you’ll see on the screen? Since Hardware Canucks is striving to evaluate cards based upon and end-user experience rather than from a purely scientific standpoint, we decided on the latter of these two methods. With this in mind, we’ve used the FCAT tools to add the timing of runted to the latency of successive frames. Dropped frames meanwhile are ignored as their value is zero. This provides a more realistic snapshot of visible fluidity.

Welcome to a shocker: in our testing and by first hand experience, the HD 7950 Boost provides a smoother experience than the GTX 760 in Hitman Absolution and Metro Last Light.

Some explanation is in order here though since the results aren't as clear-cut as the charts would have you believe. The large spikes in the GTX 760's frame times happen at completely random intervals and sometimes don't even happen at all. They are however noticeable, onscreen for only a split second and are present with every NVIDIA card in these two games when using the 320.39 beta drivers. Indeed, the problem could be anything from a specific driver / system compatibility hiccup to the driver mismanaging draw calls. We've been assured a fix will be in the works after NVIDIA replicates the issue.

 

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

As with other GTX 700-series cards, the GTX 760 uses a revised fan controller which strives to keep core temperatures at an optimal 80°C without increasing the fan’s rotational speed. However, these results aren’t necessarily realistic since many GTX 760 cards will end up using custom heatsinks which will further lower temperatures and increase clock speeds.

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 Valley is used in order to generate a constant load on the GPU(s) over the course of 15 minutes.

Once again, we aren’t seeing actual results since the version NVIDIA gave us for testing uses a “reference” heatsink which may never be used by board partners. With that being said the GTX 760 attains very low acoustical levels.

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 used 15 minutes of Unigine Valley running on a loop while letting the card sit at a stable Windows desktop for 15 minutes to determine the peak idle power consumption.

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.

With a GK104 core operating at very high frequencies, the performance per watt ratio of NVIDIA’s GTX 760 doesn’t blaze a new trail. The GTX 670 consumes as much power and performs at a higher level while the HD 7950 Boost consumes just a hair more and essentially ties NVIDIA’s new card on the framerate front.

 

[SKYMTL]

Overclocking Results

Overclocking Results

Overclocking the Kepler architecture typically boils down to a somewhat frustrating balancing act, fighting NVIDIA’s Power, Voltage and Temperature limits. The GTX 760 doesn’t change this in any way.

The limits we’ve experienced in previous cards carry over en masse with a maximum power limit of 115% (an improvement over the 106% of previous cards) and there’s a mere 12mV of extra voltage overhead which won’t do a thing for clock speeds. Needless to say, overclocking is once again filled with artificially introduced caps.

The GTX 760 does however offer a good amount of clock speed overhead. We achieved an average Boost clock of 1273MHz while the memory hit 6688MHz. The performance levels at those speeds easily outpaced the GTX 670.

 

[SKYMTL]

Conclusion

Conclusion

NVIDIA’s quick and efficient GTX 700-series rollout has loaded the enthusiast market with high performance graphics cards, every one of which has carved out its own value-focused niche. Like its predecessors, the GTX 760, bucks the status quo by outperforming similarly priced solutions and replacing the well-respected GTX 660 Ti. That should make it a major bright spot for anyone who can’t justify the GTX 770’s cost.

On paper it really doesn’t look like the GTX 760 will amount to much. However, NVIDIA’s has massaged their GK104 core in such a way that a layout with fewer cores than the GTX 660 Ti can go toe to toe against a GTX 670. This is particularly impressive since at $250, it represents an affordable solution which can provide some surprisingly robust performance at 1920 x 1080. The extra 256-bit memory interface also help push framerates upwards in certain bandwidth-limited, higher resolution situations where the GTX 660 Ti tended to struggle a bit. With that being said, the GTX 760 can really stretch its legs at 1080P, this segment’s most-used resolution.

AMD has been quite aggressive with the pricing and game bundles attached to their HD 7950 Boost but we’re not convinced this will help sway gamers’ opinions in their favor. The GTX 760 is some $30 less expensive and virtually ties its overall performance but many gamers already have many of the titles AMD is offering. That may change in the not too distant future but for the time being, some of the games just aren’t that interesting anymore. NVIDIA’s newest card also has an edge in the performance per watt category, though not by a significant or meaningful amount.

The GTX 760's random choppy frametime performance in Hitman and Metro Last Light is cause for some concern but for the time being we'll withhold any assumptions until further testing can be conducted. Plus, since the so-called "spikes" only happen at widely spaced intervals, they don't overly impact the complete gameplay experience.

The most intriguing part of this launch is how NVIDIA is once again allowing their board partners to dictate their launch-day designs. In the past this has led to some impressive products becoming available for a minor premium over the suggested retail price. From all indications, Gigabyte, ASUS, EVGA and others will have custom cooled, overclocked versions available for a mere $259, providing an intriguing twist for this launch.

For all of its positive points, many will mistakenly look at the GTX 760 as an upgrade for a similarly-priced 600-series card. That would be a misplaced expectation. NVIDIA intends this to be a possible step-up for anyone using a GTX 560 Ti or older mid-priced card and in that respect it fits the bill perfectly by outperforming the GTX 580 by a significant amount.

The GTX 760 is an important product for NVIDIA’s 2013 lineup since it targets the so called “gamer’s sweet spot” price point. It will also cause gamers to reassess their expectations of what a $249 graphics card should be able to accomplish. In short, the GTX 760 brings high resolution, affordable gaming to the masses and by doing so, becomes one of this year’s best graphics cards.