For much of the last year now, the story of the high-end video card market has been the story of NVIDIA. In September of 2014 the company launched the GeForce GTX 980, the first and at the time most powerful member of their Maxwell 2 architecture, setting a new mark for both power efficiency and performance, securing their lead of high-end of the video card market. NVIDIA then followed that up in March with the launch of the GeForce GTX Titan X, NVIDIA’s true flagship Maxwell part, and a part that only served to further cement their lead.
Based on the very powerful (and very large) GM200 GPU, GTX Titan X is currently untouched in performance. However priced at $1000, it is also currently untouched in price. In NVIDIA’s current lineup there is a rather sizable gap between the $550 GTX 980 and $1000 GTX Titan X, and perhaps more significantly GTX Titan X was the only GM200 part on the market. With NVIDIA launching their fully enabled flagship card first, it was only a matter of time until they released a cheaper card based on a cut-down version of the GM200 GPU in order to fill that pricing hole and to put salvaged GM200s to good use.
Now just a bit over two months since the launch of the GTX Titan X, NVIDIA launching their second GM200 card, GeForce GTX 980 Ti. Based on the aforementioned cut-down version of GM200, GTX 980 Ti is the expected junior version of GTX Titan X, delivering GM200 at a cheaper price point. But calling GTX 980 Ti a cheaper GM200 may be selling it short; “cheaper” implies that GTX 980 Ti is a much lesser card. At $649, GTX 980 Ti is definitely cheaper, but the card that is launching today is not to be underestimated. GTX 980 Ti may be intended to be GTX Titan X’s junior, but with the excellent performance it delivers, GTX 980 Ti may as well be GTX Titan X itself.
NVIDIA GeForce Specification Comparison
GTX Titan X
GTX 980 Ti
GTX 980
GTX 780 Ti
CUDA Cores
3072
2816
2048
2880
Texture Units
192
176
128
240
ROPs
96
96
64
48
Core Clock
1000MHz
1000MHz
1126MHz
875MHz
Boost Clock
1075MHz
1075MHz
1216MHz
928Mhz
Memory Clock
7GHz GDDR5
7GHz GDDR5
7GHz GDDR5
7GHz GDDR5
Memory Bus Width
384-bit
384-bit
256-bit
384-bit
VRAM
12GB
6GB
4GB
3GB
FP64
1/32 FP32
1/32 FP32
1/32 FP32
1/24 FP32
TDP
250W
250W
165W
250W
GPU
GM200
GM200
GM204
GK110B
Architecture
Maxwell 2
Maxwell 2
Maxwell 2
Kepler
Transistor Count
8B
8B
5.2B
7.1B
Manufacturing Process
TSMC 28nm
TSMC 28nm
TSMC 28nm
TSMC 28nm
Launch Date
03/17/2015
06/01/2015
09/18/2014
11/07/2013
Launch Price
$999
$649
$549
$699
Taking a look at GTX 980 Ti from a specifications perspective, NVIDIA’s latest card is in a somewhat unusual place. Its direct predecessor, GTX 780 Ti, was a fully enabled GK110 card, differing from that generation’s Titan only in double precision compute capabilities and a trivial clockspeed difference. However with GM200 being a pure graphics chip – and hence GTX Titan X not pulling double-duty as a prosumer level compute card – NVIDIA has needed to turn to cutting down the chip itself in order to differentiate the products. In this sense the GTX 980 Ti is probably closer to being the GTX 780 of its generation, a very timely situation given the fact that GTX 780 launched almost exactly 2 years ago.
In any case, compared to GTX Titan X NVIDIA has made just two changes to create GTX 980 Ti. The first is that the company has disabled 2 of GM200’s 24 SMMs, bringing it down to 22 SMMs for a total of 2816 active CUDA cores. The second change was to dial back the amount of VRAM, from GTX Titan X’s 12GB to the 6GB we see on GTX 980 Ti. And that’s it.
In every other aspect the GTX 980 Ti is identical to the GTX Titan X. Both are clocked at 1000MHz, with a boost clock of 1075MHz. Both feature their VRAM on a 384-bit memory bus with their respective VRAM modules clocked at 7GHz. Both are set for a 250W TDP, and are equipped with NVIDIA’s high end metal-shrouded cooler. And finally, GTX 980 Ti retains all 96 ROPs and 3MB of L2 cache, which means that in pixel-heavy situations like 4K it is as no disadvantage compared to GTX Titan X.
Consequently, looking at the specifications tells us that we should expect GTX 980 Ti to be 8% slower than GTX Titan X, a result of losing 2 SMMs. Coupled with the difference in VRAM between the two cards, this would put a decent gap between the two cards. However as we’ll see in our benchmarks, reality begs to differ. Thanks to the combination of a couple of factors GTX 980 Ti ends up coming much closer to GTX Titan X than what the specifications tell us to expect. In the end what we find is that it delivers 97% of GTX Titan X’s performance. This, in a nutshell, is what makes GTX 980 Ti a deceptive card, and is why it’s going to have such a large impact on the high-end market.
Shifting gears, let’s talk about pricing, availability, and the competition. The GTX 980 Ti will be a hard launch, with cards going on sale on June 1st. Due to the fact that Computex is taking place this week in Taiwan and GTX 980 Ti is one of the products NVIDIA is launching at the show, NVIDIA has lifted the embargo on GTX 980 Ti at an atypical 6pm Eastern, which for Taiwan and Computex is June 1st, 6am local time. NVIDIA is launching the card globally on the 1st, so in a reversal of typical launches APAC buyers will get first dibs on the card, followed by European and North/South American buyers several hours later. Along with the GTX 980 Ti reference cards launching today, expect to see semi-custom cards launching very soon thereafter.
Meanwhile for pricing, the GTX 980 Ti will be launching at $649. This is an increasingly persistent price point for NVIDIA that has fluctuated a bit over the last couple of years, with the GTX 780 launching at $649 as well, only for the GTX 780 Ti to launch at $699. The launch of GTX 980 Ti at $649 will be putting pressure on the rest of NVIDIA’s product stack from both above and below. In response to this launch NVIDIA is officially cutting the price of the GTX 980 from $549 to $499 in order to open up a bit more room between the cards and to keep GTX 980 Ti from making GTX 980 redundant. At the same time however GTX 980 Ti puts enormous pressure on GTX Titan X; GTX 980 Ti’s performance is close enough to GTX Titan X that the latter’s only practical advantage is its 12GB of VRAM, and that’s not a lot to justify the Titan’s $350 (54%) price premium.
Not stopping there, in an unusual move for NVIDIA the GTX 980 Ti is getting a game bundle right off the bat. The card isn’t getting NVIDIA’s full Two Times The Adventure bundle that comes with the GTX 980, but it is getting a copy of the forthcoming Batman: Arkham Knight, another one of this year’s major GameWorks titles. The end result is that NVIDIA is being more aggressive than usual this time around, offering what amounts to a GTX Titan X and a game for $649.
Of course the competition may have something to do with it. AMD is pretty much shouting from the rooftops that they are launching a new high-end video card this quarter, which at this point means the card is due by the end of June. While NVIDIA does have other financial incentives for releasing GTX 980 Ti now that GTX Titan X has been on the market for a couple of months, by all appearances this looks to be NVIDIA making the first move. We’ll have to see just what AMD delivers next month, but what is clear is that whatever they do, NVIDIA will not be making it easy by delivering flagship performance at $649.
In the meantime AMD and their partners are also still selling the Radeon R9 295X2 for around $700, though it looks like this is part of an effort to sell off the remaining inventory of cards. AMD has no other cards in this price range, so the GTX 980 Ti is otherwise uncontested until AMD’s new card launches.
Like the rest of NVIDIA’s high-end cards in this generation, the reference GeForce GTX 980 Ti is launching with NVIDIA’s standard metal cooler. This design has served NVIDIA well since the launch of the GTX Titan in 2013 and continues to be the blower design to beat the high end, easily handling the 250W TDP of NVIDIA’s high-end cards without generating a ton of noise in the process.
As with so many other aspects of the GTX 980 Ti, the GTX 980 Ti’s cooler and build is a near-copy of the GTX Titan X. The only difference in the cooler is the paint job; GTX Titan X got a unique black paint job, while GTX 980 Ti gets the more standard bare aluminum finish with black lettering and a black-tinted polycarbonate window.
Otherwise there’s very little to be said about the GTX 980 Ti’s design that hasn’t been said before, so we’ll just recap what we said about the cooler design from our review of the GTX Titan X.
For GTX 980 Ti, NVIDIA has opted to leave well enough alone, having made virtually no changes to the shroud or cooling apparatus. And truth be told it’s hard to fault NVIDIA right now, as this design remains the gold (well, aluminum) standard for a blower. Looks aside, after years of blowers that rattled, or were too loud, or didn’t cool discrete components very well, NVIDIA is sitting on a very solid design that I’m not really sure how anyone would top (but I’d love to see them try).
In any case, our favorite metal shroud is back once again. Composed of a cast aluminum housing and held together using a combination of rivets and screws, it’s as physically solid a shroud as we’ve ever seen. The card measures 10.5” long overall, which at this point is NVIDIA’s standard size for high-end GTX cards.
Drilling down we have the card’s primary cooling apparatus, composed of a nickel-tipped wedge-shaped heatsink and ringed radial fan. The heatsink itself is attached to the GPU via a copper vapor chamber, something that has been exclusive to NVIDIA’s 250W cards and provides the best possible heat transfer between the GPU and heatsink. Meanwhile the rest of the card is covered with a black aluminum baseplate, providing basic heatsink functionality for the VRMs and other components while also protecting them.
Finally at the bottom of the stack we have the card itself, complete with the GM200 GPU, VRAM chips, and various discrete components. The GM200 PCB places the GPU and VRAM chips towards the front of the card, while the VRMs and other discrete components occupy the back. As with the GTX Titan X, GTX 980 Ti features NVIDIA’s reworked component placement to improve airflow to the discrete components and reduce temperatures, along with employing molded inductors.
NVIDIA once again employs a 6+2 phase VRM design, with 6 phases for the GPU and another 2 for the VRAM. This means that GTX 980 Ti has a bit of power delivery headroom – NVIDIA allows the power limit to be increased by 10% to 275W – but hardcore overclockers will find that there isn’t an extreme amount of additional headroom to play with. Based on our sample the actual shipping voltage at the max boost clock is a bit higher than GTX Titan X, coming in at 1.187v, so in non-TDP constrained scenarios there is some additional headroom through overvolting, up to 1.23v in the case of our sample.
In terms of overall design, unlike GTX Titan X and its 24 VRAM chips, for the GTX 980 Ti NVIDIA only needs to use 12 VRAM chips to get the card’s 6GB of VRAM, so all of the VRAM is located at the front of the card. Halving the RAM capacity simplifies the card a bit – there are now no critical components on the back – and it brings down the total VRAM power consumption slightly. However despite this, NVIDIA has not brought back the backplate from the GTX 980, having removed it on the GTX Titan X due to the VRAM chips it placed on the rear.
Moving on, in accordance with GTX 980 Ti’s 250W TDP and the reuse of the metal cooler, power delivery for the GTX 980 Ti is identical to its predecessors. This means a 6-pin and an 8-pin power connector at the top of the card, to provide up to 225W, with the final 75W coming from the PCIe slot.
Meanwhile display I/O follows the same configuration we’ve seen on the rest of the high-end GTX 900 series. This is 1x DL-DVI-I, 3x DisplayPort 1.2, and 1x HDMI 2.0, with a total limit of 4 displays. In the case of GTX 980 Ti the DVI port is somewhat antiquated at this point – the card is generally overpowered for the relatively low maximum resolutions of DL-DVI – but on the other hand the HDMI 2.0 port is actually going to be of some value here since it means GTX 980 Ti can drive a 4K TV. Meanwhile if you have money to spare and need to drive more than a single 4K display, GTX 980 Ti also features a pair of SLI connectors for even more power.
Finally, taking a look at the long term, I wanted to quickly hit upon the subject of the VRAM capacity difference between the GTX 980 Ti and the GTX Titan X. Essentially NVIDIA’s only remaining selling point for the GTX Titan X, the Titan will remain their only 12GB card for some time to come. For NVIDIA this means that they can pitch the GTX Titan X as a more future-proof card than the GTX 980 Ti, as it would be hard-pressed to run out of VRAM.
The question for the moment then is whether 12GB is worth a higher premium, let alone the GTX Titan X’s $350 premium. The original GTX Titan by comparison was fortunate enough to come out with 6GB right before the current-generation consoles launched, and with them their 8GB memory configurations. This lead to a rather sudden jump in VRAM requirements in games that the GTX Titan was well positioned to handle, whereas GTX 780 Ti and its 3GB of VRAM can struggle in the very latest games at 4K resolutions. Much like 6GB in 2013, 12GB is overkill in 2015, all the while 6GB is a more practical amount for a 384-bit card at this time.
But to answer the question at hand, unlike the original GTX Titan, I suspect 12GB will remain overkill for a much longer period of time, especially without a significant technology bump like the consoles to drive up VRAM requirements. And consequently I don’t expect GTX 980 Ti to have any real issues with VRAM capacity in games over the next couple of years, making it better off than the GTX 780 Ti, relatively speaking.
Alongside the launch of the GTX 980 Ti, NVIDIA is also taking advantage of Computex to make a couple of other major technology announcements. Given the scope of these announcements we’re covering these in separate articles, but we’ll quickly go over the high points here as they pertain to the GTX 980 Ti.
G-Sync Variable Overdrive & Windowed Mode G-Sync
NVIDIA is announcing a slew of G-Sync products/technologies today, the most important of which is Mobile G-Sync for laptops. However as part of that launch, NVIDIA is also finally confirming that all G-Sync products, including existing desktop G-Sync products, feature support for G-Sync variable overdrive. As the name implies, this is the ability to vary the amount of overdrive applied to a pixel based on a best-effort guess of when the next frame will arrive. This allows NVIDIA to continue to use pixel overdrive on G-Sync monitors to improve pixel response times and reduce ghosting, at a slight cost to color accuracy while in motion from errors in the frame time predictions.
Variable overdrive has been in G-Sync since the start, however until now NVIDIA has never confirmed its existence, with NVIDIA presumably keeping quiet about it for trade secret purposes. However now that displays supporting AMD’s Freesync implementation of DisplayPort Adaptive-Sync are out, NVIDIA is further clarifying how G-Sync works.
Meanwhile being freshly rolled out in NVIDIA’s latest drivers is support for Windowed Mode G-Sync. Before now, running a game in Windowed mode could cause stutters and tearing because once you are in Windowed mode, the image being output is composited by the Desktop Window Manager (DWM) in Windows. Even though a game might be outputting 200 frames per second, DWM will only refresh the image with its own timings. The off-screen buffer for applications can be updated many times before DWM updates the actual image on the display.
NVIDIA will now change this using their display driver, and when Windowed G-Sync is enabled, whichever window is the current active window will be the one that determines the refresh rate. That means if you have a game open, G-Sync can be leveraged to reduce screen tearing and stuttering, but if you then click on your email application, the refresh rate will switch back to whatever rate that application is using. Since this is not always going to be a perfect solution - without a fixed refresh rate, it's impossible to make every application perfectly line up with every other application - Windowed G-Sync can be enabled or disabled on a per-application basis, or just globally turned on or off.
GameWorks VR & Multi-Res Shading
Also being announced at Computex is a combination of new functionality and an overall rebranding for NVIDIA’s suite of VR technologies. First introduced alongside the GeForce GTX 980 in September as VR Direct, NVIDIA will be bringing their VR technologies in under the GameWorks umbrella of developer tools. The collection of technologies will now be called GameWorks VR, adding to the already significant collection of GameWorks tools and libraries.
On the feature front, the newly minted GameWorks VR will be getting a new feature dubbed Multi-Resolution Shading, or Multi-Res Shading for short. With multi-res shading, NVIDIA is looking to leverage the Maxwell 2 architecture’s Multi-Projection Acceleration in order to increase rendering efficiency and ultimately the overall performance of their GPUs in VR situations.
By reducing the resolution of video frames at the edges where the human eye is less sensitive, NVIDIA says that using multi-res shading can result in a 1.3x to 2x increase in pixel shader performance without noticeably compromising the image quality. Like many of the other technologies in the GameWorks VR toolkit this is an implementation of a suggested VR practice, however in NVIDIA’s case the company believes they have a significant technological advantage in implementing it thanks to multi-projection acceleration. With MPA to bring down the rendering cost of this feature, NVIDIA’s hardware can better take advantage of the performance advantages of this rendering approach, essentially making it an even more efficient method of VR rendering.
Getting Behind DirectX Feature Level 12_1
Finally, though not an outright announcement per-se, from a marketing perspective we should expect to see NVIDIA further promote their current technological lead in rendering features. The Maxwell 2 architecture is currently the only architecture to support DirectX feature level 12_1, and with DirectX 12 games due a bit later this year, NVIDIA sees that as an advantage to press.
For promotional purposes NVIDIA has put together a chart listing the different tiers of feature levels for DirectX 12, and to their credit this is a simple but elegant layout of the current feature level situation. The bulk of the advanced DirectX 12 features we saw Microsoft present at the GTX 980 launch are part of feature level 12_1, while the rest, and other functionality not fully exploited under DirectX 11 are part of the 12_0 feature level. The one exception to this is volume tiled resources, which is not part of either feature level and instead is part of a separate feature list for tiled resources that can be implemented at either feature level.
The press drivers for the launch of the GTX 980 Ti are release 352.90, which other than formally adding support for the new card is otherwise identical to the standing 352.86 drivers.
CPU:
Intel Core i7-4960X @ 4.2GHz
Motherboard:
ASRock Fatal1ty X79 Professional
Power Supply:
Corsair AX1200i
Hard Disk:
Samsung SSD 840 EVO (750GB)
Memory:
G.Skill RipjawZ DDR3-1866 4 x 8GB (9-10-9-26)
Case:
NZXT Phantom 630 Windowed Edition
Monitor:
Asus PQ321
Video Cards:
AMD Radeon R9 295X2
AMD Radeon R9 290X
AMD Radeon HD 7970
NVIDIA GeForce GTX Titan X
NVIDIA GeForce GTX 980 Ti
NVIDIA GeForce GTX 980
NVIDIA GeForce GTX 780 Ti
NVIDIA GeForce GTX 780
NVIDIA GeForce GTX 680
NVIDIA GeForce GTX 580
Video Drivers:
NVIDIA Release 352.90 Beta
AMD Catalyst Cat 15.5 Beta
OS:
Windows 8.1 Pro
Kicking off our benchmark suite is Battlefield 4, DICE’s 2013 multiplayer military shooter. After a rocky start, Battlefield 4 has since become a challenging game in its own right and a showcase title for low-level graphics APIs. As these benchmarks are from single player mode, based on our experiences our rule of thumb here is that multiplayer framerates will dip to half our single player framerates, which means a card needs to be able to average at least 60fps if it’s to be able to hold up in multiplayer.
Battlefield 4 is going to set the pace for the rest of this review. In our introduction we talked about how the GTX 980 Ti may as well be the GTX Titan X, and this is one such example why. With a framerate deficit of no more than 3% in this benchmark, the difference between the two cards is just outside the range of standard run-to-run experimental variation that we see in our benchmarking process. So yes, it really is that fast.
In any case, after stripping away the Frostbite engine’s expensive (and not wholly effective) MSAA, what we’re left with for BF4 at 4K with Ultra quality puts the 980 Ti in a pretty good light. At 56.5fps it’s not quite up to the 60fps mark, but it comes very close, close enough that the GTX 980 Ti should be able to stay above 30fps virtually the entire time, and never drop too far below 30fps in even the worst case scenario. Alternatively, dropping to Medium quality should give the card plenty of headroom, with an average framerate of 91.8fps meaning even the lowest framerate never drops below 45fps.
Meanwhile our other significant comparison here is the GTX 980, which just saw its price cut by $50 to $499 to make room for the GTX 980 Ti. At $649 the GTX 980 Ti ideally should be 30% faster to justify its 30% higher price tag; here it’s almost exactly on that mark, fluctuating between a 28% and 32% lead depending on the resolution and settings.
Finally, shifting gears for a moment, gamers looking for the ultimate 1440p card will not be disappointed. GTX 980 Ti will not get to 120fps here (it won’t even come close), but at 77.7fps it’s well suited for driving 1440p144 displays. In fact and GTX Titan X are the single-GPU cards to do better than 60fps at this resolution.
Still one of our most punishing benchmarks, Crysis 3 needs no introduction. With Crysis 3, Crytek has gone back to trying to kill computers and still holds “most punishing shooter” title in our benchmark suite. Only in a handful of setups can we even run Crysis 3 at its highest (Very High) settings, and that’s still without AA. Crysis 1 was an excellent template for the kind of performance required to drive games for the next few years, and Crysis 3 looks to be much the same for 2015.
Once more we find the GTX 980 Ti and GTX Titan X virtually tied. Across all settings and resolutions the GTX 980 Ti stays within 97-98% of the Titan’s performance. Consequently GTX Titan X is ever so marginally better, but not enough to make any real difference.
This also means that GTX 980 Ti continues with its very strong lead over the GTX 980. Once more we’re looking at a 26-31% performance advantage for the latest member of the GTX 900 series, in-line with its price premium.
Meanwhile on an absolute basis, as one of our most punishing games this is also a good reminder of why even GM200 cards can’t quite pull off high quality 4K gaming with a single GPU today. Even without MSAA and one step below Crysis 3’s Very High quality settings, the GTX 980 Ti can only muster 40.9fps. If you want to get to 60fps you will need to drop to Low quality, or drop the resolution to 1440p. The latter will get you 83.2fps at the same quality settings, which again highlights GTX 980 Ti’s second strength as a good card for driving high refresh rate 1440p displays.
Our next benchmark is Monolith’s popular open-world action game, Middle Earth: Shadow of Mordor. One of our current-gen console multiplatform titles, Shadow of Mordor is plenty punishing on its own, and at Ultra settings it absolutely devours VRAM, showcasing the knock-on effect of current-gen consoles have on VRAM requirements.
Another game, another set of benchmarks where the GTX 980 Ti and GTX Titan X are more or less tied. In this case the latest GM200 card closes the tiny gap even more, bringing the difference between the two down to 1-2% in favor of the GTX Titan X. Meanwhile the GTX 980 Ti’s advantage over the GTX 980 is as strong as ever, beating the most powerful of the GM204 cards by 30% or more.
On an absolute basis, as with Crysis 3 GTX 980 Ti won’t be enough for 60fps at 4K, but at 47.9fps it’s closer to 60fps than 30fps, representing a significant improvement in 4K performance in only a generation. Turning down the game’s quality settings to Very High does improve performance a bit, but at 53.7fps it’s still not quite enough for 60fps. The biggest advantage of Very High quality is alleviating some of the high VRAM requirements, not that the GTX 980 Ti seems to mind even at 6GB. Otherwise dropping to 1440p will give us a significant bump in performance, pushing framerates over 80fps once again.
Looking at minimum framerates, we find the one and only place under which the GTX 980 Ti may be struggling to keep up with its Titan sibling. While it held very close to the GTX Titan X in average framerates, the minimum framerate finds a larger, distinct gap between the two, with the GTX 980 Ti trailing by 8%. That said, minimum framerates are inherently more unreliable than averages, and other than a momentary dip the GTX 980 Ti is doing quite well here, so while it’s a less-than-perfect showing, I don’t believe we’re seeing any kind of real impact from VRAM differences. Note that the 4GB cards don’t seem to be worse off despite being short a further 2GB of VRAM.
Shifting gears from action to strategy, we have Civilization: Beyond Earth, the latest in the Civilization series of strategy games. Civilization is not quite as GPU-demanding as some of our action games, but at Ultra quality it can still pose a challenge for even high-end video cards. Meanwhile as the first Mantle-enabled strategy title Civilization gives us an interesting look into low-level API performance on larger scale games, along with a look at developer Firaxis’s interesting use of split frame rendering with Mantle to reduce latency rather than improving framerates.
Stop me if you’ve heard this one before: the GTX 980 Ti and GTX Titan perform within a few percent of each other. Under Civilization the gap between the two is a hair larger than in other games, at 3-4%, but this is also as large of a gap as you’ll see for average framerates. Even here the two are for all meaningful purposes tied.
Meanwhile on an absolute basis, the GM200 twin remain the only single-GPU cards to crack 60fps, with GTX 980 Ti delivering 70.5fps at the game’s most extreme setting. This is once again well ahead of the GTX 980 – beating it by 34% at 4K, though by less at lower resolutions where we start to get CPU-bottlenecked.
I also want to quickly touch upon how the GTX 980 Ti compares to the last-generation high-end GK110 Kepler cards, the GTX 780 and GTX 780 Ti. Against GTX 780 in particular, in this test we see the GTX 980 Ti deliver 70-80% better performance. With this being the 2 year anniversary of the GTX 780’s release, this is especially notable since it’s such a good example of how performance has improved specifically at this $649 price point in the last 2 years. GM200 in general is not this fast versus GK110 – there’s only so much to be done at 28nm – but against GTX 780 in particular NVIDIA’s latest card looks quite good. Even GTX 780 Ti is not entirely immune, with GTX 980 Ti beating it by around 45% at 4K.
When it comes to minimum framerates the same story continues; the two GM200 cards are very close to each other, staying within 5%. At worst, you can say that the 7-17% performance advantage over the GTX 980 isn’t very impressive, though this is admittedly a game that’s not too far off from being CPU-bottlenecked.
Our RPG of choice for 2015 is Dragon Age: Inquisition, the latest game in the Dragon Age series of ARPGs. Offering an expansive world that can easily challenge even the best of our video cards, Dragon Age also offers us an alternative take on EA/DICE’s Frostbite 3 engine, which powers this game along with Battlefield 4.
Meanwhile on an absolute basis what we find at 4K with Ultra quality is that the GTX 980 Ti is once more hitting framerates in the 40fps range. At 40.7fps the GTX 980 Ti is the only other single-GPU card to average better than 30fps at these settings, with the next-closest card being the GTX 980 at exactly 30fps. Otherwise users looking for higher framerates can either turn down the quality setting one notch to high, which gets us 52.6fps, or drop down to 1440p, which is good for 77.9fps.
Croteam’s first person puzzle and exploration game The Talos Principle may not involve much action, but the game’s lush environments still put even fast video cards to good use. Coupled with the use of 4x MSAA at Ultra quality, and even a tranquil puzzle game like Talos can make a good case for more powerful video cards.
At 4K Ultra quality the GTX 980 Ti won’t quite break 60fps, but at 52.5fps and like the GTX Titan X, it’s not too far off. Compared to the GTX 980 this is another 31% performance advantage as well.
The next game in our 2015 GPU benchmark suite is Far Cry 4, Ubisoft’s Himalayan action game. A lot like Crysis 3, Far Cry 4 can be quite tough on GPUs, especially with Ultra settings thanks to the game’s expansive environments.
For those of you just joining us, the GTX 980 Ti and GTX Titan are still neck-and neck. The GTX 980 Ti trails the Titan by no more than 4%, at 1440p and 2560, with the two cards taking the top two spots in our charts for single-GPU cards as one would expect.
On an absolute basis, at 4K Ultra this happens to be another case where the GTX 980 Ti delivers framerates around 40fps, in this case coming in at 40.6fps. Otherwise the GTX 980 Ti is going to come up a hair short of 60fps at medium quality – hitting 59.5fps – and finally going over 70fps at 1440p Ultra.
This also ends up being another case where the GTX 980 Ti looks very good as compared to the GTX 780. Here it beats NVIDIA’s last $649 card by as much as 86% at 1440p, highlighting NVIDIA’s performance gains at this price point over the last 2 years.
The second strategy game in our benchmark suite, Total War: Attila is the latest game in the Total War franchise. Total War games have traditionally been a mix of CPU and GPU bottlenecks, so it takes a good system on both ends of the equation to do well here. In this case the game comes with a built-in benchmark that plays out over a large area with a fortress in the middle, making it a good GPU stress test.
Switching out to another strategy game, even given Attila’s significant GPU requirements at higher settings, GTX 980 Ti still doesn’t falter. It trails GTX Titan X by just 2% at all settings.
That said, on an absolute basis even GTX Titan X couldn't get past 30fps here at 4K with max quality settings, so the GTX 980 Ti is going to fare no better.. To get single card performance above 30fps we have to drop a notch to the “Quality” setting, which gets the GTX 980 Ti up to 44.2fps. In any case, at these settings the GTX 980 Ti makes easy work of the single-GPU competition, beating the GTX 980 once again by 28%+.
For the racing game in our benchmark suite we have Codemasters’ GRID Autosport. Codemasters continues to set the bar for graphical fidelity in racing games, delivering realistic looking environments with layed with additional graphical effects. Based on their in-house EGO engine, GRID Autosport includes a DirectCompute based advanced lighting system in its highest quality settings, which incurs a significant performance penalty on lower-end cards but does a good job of emulating more realistic lighting within the game world.
Switching out to another strategy game, even given Attila’s significant GPU requirements at higher settings, GTX 980 Ti still doesn’t falter. It trails GTX Titan X by just 2% at all settings.
As was the case with all of our other games so far, our racing benchmark of choice does no better in separating the two GM200 cards, with GTX 980 Ti yet again trailing GTX Titan X by no more than 3%. Even with everything cranked up to max, the GTX 980 Ti makes easy work of GRID at 4K, hitting 70.6fps at 4K Ultra and making it the cheapest card to crack 60fps. This also continues to be a solid lead for the GTX 980 Ti over the GTX 980 and GTX 780, beating the two cards by 28% and 76% respectively at 4K.
The final game in our review of the GTX 980 Ti is our most recent addition, Grand Theft Auto V. The latest edition of Rockstar’s venerable series of open world action games, Grand Theft Auto V was originally released to the last-gen consoles back in 2013. However thanks to a rather significant facelift for the current-gen consoles and PCs, along with the ability to greatly turn up rendering distances and add other features like MSAA and more realistic shadows, the end result is a game that is still among the most stressful of our benchmarks when all of its features are turned up. Furthermore, in a move rather uncharacteristic of most open world action games, Grand Theft Auto also includes a very comprehensive benchmark mode, giving us a great chance to look into the performance of an open world action game.
On a quick note about settings, as Grand Theft Auto V doesn't have pre-defined settings tiers, I want to quickly note what settings we're using. For "Very High" quality we have all of the primary graphics settings turned up to their highest setting, with the exception of grass, which is at its own very high setting. Meanwhile 4x MSAA is enabled for direct views and reflections. This setting also involves turning on some of the advanced redering features - the game's long shadows, high resolution shadows, and high definition flight streaming - but it not increasing the view distance any further.
Otherwise for "High" quality we take the same basic settings but turn off all MSAA, which significantly reduces the GPU rendering and VRAM requirements.
After initially expecting Grand Theft Auto to be a walk in the park performance wise, the PC version of the game has instead turned out to be a very demanding games for our GPUs. Even at 1440p we can’t have very high quality with MSAA and still crack 60fps, though we can get very close.
Ultimately GTA doesn’t do any better than any other game in setting apart our GM200 cards. GTX 980 Ti trails GTX Titan by 4% or less, essentially the average outcome at this point. Also average is the GTX 980 Ti’s lead over the GTX 980, with the newest card beating the older GTX 980 by 29-31% across our three settings. Finally, against the GTX 780 the GTX 980 Ti has another strong showing, with a 69-79% lead.
On an absolute basis we can see that at 4K we can’t have 4x MSAA and even crack 30fps on a single-GPU card, with GTX 980 Ti topping out at 27.8 fps. Taking out MSAA brings us up to 46.2fps, which is still well off 60fps, but also well over the 30fps cap that this game was originally designed against on the last-generation consoles.
Along with an all-around solid benchmark scene, the other interesting benchmarking feature of GTA is that it also generates frame percentiles on its own, allowing us to see the percentiles without going back and recording the game with FRAPS. Taking a look at the 99th percentile in this case, what we find is that at each setting GTA crushes some group of cards due to a lack of VRAM.
At 4K very high quality, 4GB cards have just enough VRAM to stay alive, with the multi-GPU R9 295X2 getting crushed due to the additional VRAM requirements of AFR pushing it over the edge. Not plotted here are the 3GB cards, which saw their framerates plummet to the low single-digits, essentially struggling to complete this benchmark. Meanwhile 1440p at high quality crushes our 2GB cards, with less VRAM than a Radeon HD 7970 falling off of the cliff.
As for what this means for the GTX 980 Ti, the situation finds the GTX 980 Ti trailing the GTX Titan X in 99th percentile framerates by anywhere between 3% and 10%. This test is not designed to push more than 6GB of VRAM, so I’m not entirely convinced this isn’t a wider than normal variance (especially at the low framerates for 4K), though the significant and rapid asset streaming this benchmark requires may be taking its toll on the GTX 980 Ti, which has less VRAM for additional caching.
As always we’ll also take a quick look at synthetic performance. Being a virtual copy of the GTX Titan X, GTX 980 Ti should perform very similarly here, just as we've seen in our gaming tests.
Compared to GTX Titan X, GTX 980 Ti does technically lose 2 Polymorph Engines as a result of losing 2 SMMs. However as with our games, this doesn’t really hinder GTX 980 Ti, leading it being within a few percent of GTX Titan X on tessellation performance.
As for texel and pixel fillrates, the results are both as-expected and a bit surprising. On the expected side, we see the GTX 980 Ti trail GTX Titan X by a bit, again taking a hit from the SMM loss. On the other hand we’re seeing a larger than expected drop in the pixel fill rates. GTX 980 Ti loses some rasterization throughput from the SMM loss, but a 15% drop in this test is much larger than 2 SMMs. Just to be sure we checked to make sure the ROP/MC configuration of GTX 980 Ti was unchanged at 96 ROPs, so we’re at a bit of a loss to explain the difference at this time. Though in either case, despite what 3DMark is telling us, we aren’t seeing any signs of GTX 980 Ti struggling at 4K versus GTX Titan X. So if there is a meaningful difference in pixel fillrates, it’s not impacting game performance.
Shifting gears, we have our look at compute performance. Since GTX Titan X has no compute feature advantage - no fast double precision support like what's found in the Kepler generation Titans - the performance difference between the GTX Titan X and GTX 980 Ti should be very straightforward.
Starting us off for our look at compute is LuxMark3.0, the latest version of the official benchmark of LuxRender 2.0. LuxRender’s GPU-accelerated rendering mode is an OpenCL based ray tracer that forms a part of the larger LuxRender suite. Ray tracing has become a stronghold for GPUs in recent years as ray tracing maps well to GPU pipelines, allowing artists to render scenes much more quickly than with CPUs alone.
With the pace set for GM200 by GTX Titan X, there’s little to say here that hasn’t already been said. Maxwell does not fare well in LuxMark, and while GTX 980 Ti continues to stick very close to GTX Titan X, it none the less ends up right behind the Radeon HD 7970 in this benchmark.
For our second set of compute benchmarks we have CompuBench 1.5, the successor to CLBenchmark. CompuBench offers a wide array of different practical compute workloads, and we’ve decided to focus on face detection, optical flow modeling, and particle simulations.
Although GTX T980 Ti struggled at LuxMark, the same cannot be said for CompuBench. Though taking the second spot in all 3 sub-tests - right behind GTX Titan X - there's a bit wider of a gap than normal between the two GM200 cards, causing GTX 980 Ti to trail a little more significantly than in other tests. Given the short nature of these tests, GTX 980 Ti doesn't get to enjoy its usual clockspeed advantage, making this one of the only benchmarks where the theoretical 9% performance difference between the cards becomes a reality.
Our 3rd compute benchmark is Sony Vegas Pro 13, an OpenGL and OpenCL video editing and authoring package. Vegas can use GPUs in a few different ways, the primary uses being to accelerate the video effects and compositing process itself, and in the video encoding step. With video encoding being increasingly offloaded to dedicated DSPs these days we’re focusing on the editing and compositing process, rendering to a low CPU overhead format (XDCAM EX). This specific test comes from Sony, and measures how long it takes to render a video.
Traditionally a benchmark that favors AMD, GTX 980 Ti fares as well as GTX Titan X, closing the gap some. But it's still not enough to surpass Radeon HD 7970, let alone Radeon R9 290X.
Moving on, our 4th compute benchmark is FAHBench, the official Folding @ Home benchmark. Folding @ Home is the popular Stanford-backed research and distributed computing initiative that has work distributed to millions of volunteer computers over the internet, each of which is responsible for a tiny slice of a protein folding simulation. FAHBench can test both single precision and double precision floating point performance, with single precision being the most useful metric for most consumer cards due to their low double precision performance. Each precision has two modes, explicit and implicit, the difference being whether water atoms are included in the simulation, which adds quite a bit of work and overhead. This is another OpenCL test, utilizing the OpenCL path for FAHCore 17.
Folding @ Home’s single precision tests reiterate GM200's FP32 compute credentials. Second only to GTX Titan X, GTX 980 Ti fares very well here.
Meanwhile Folding @ Home’s double precision test reiterates GM200's poor FP64 compute performance. At 6.3ns/day, it, like the GTX Titan X, occupies the lower portion of our benchmark charts, below AMD's cards and NVIDIA's high-performnace FP64 cards.
Wrapping things up, our final compute benchmark is an in-house project developed by our very own Dr. Ian Cutress. SystemCompute is our first C++ AMP benchmark, utilizing Microsoft’s simple C++ extensions to allow the easy use of GPU computing in C++ programs. SystemCompute in turn is a collection of benchmarks for several different fundamental compute algorithms, with the final score represented in points. DirectCompute is the compute backend for C++ AMP on Windows, so this forms our other DirectCompute test.
We end up ending our benchmarks where we started: with the GTX 980 Ti slightly trailing the GTX Titan X, and with the two GM200 cards taking the top two spots overall. So as with GTX Titan X, GTX 980 Ti is a force to be reckoned with for FP32 compute, which for a pure consumer card should be a good match for consumer compute workloads.
As always, last but not least is our look at power, temperature, and noise. Next to price and performance of course, these are some of the most important aspects of a GPU, due in large part to the impact of noise. All things considered, a loud card is undesirable unless there’s a sufficiently good reason – or sufficiently good performance – to ignore the noise.
As the GM200 flagship card, GTX Titan X gets the pick of the litter as far as GM200 GPUs go. GTX Titan X needed fully-functional GM200 GPUs, and even then needed GPUs that were good enough to meet NVIDIA’s power requirements. GTX 980 Ti on the other hand, as a cut-down/salvage card, gets second pick. So we expect to see these chips be just a bit worse; to have either functional units that came out of the fab damaged, or have functional units that have been turned off due to power reasons.
GeForce GTX Titan X/980 Voltages
GTX Titan X Boost Voltage
GTX 980 Ti Boost Voltage
GTX 980 Boost Voltage
1.162v
1.187v
1.225v
Looking at voltages, we can see just that in our samples. GTX 980 Ti has a slightly higher boost voltage – 1.187v – than our GTX Titan X. NVIDIA sometimes bins their second-tier cards for lower voltage, but this isn’t something we’re seeing here. Nor is there necessarily a need to bin in such a manner since the 250W TDP is unchanged from GTX Titan X.
GeForce GTX 980 Ti Average Clockspeeds
Game
GTX 980 Ti
GTX Titan X
Max Boost Clock
1202MHz
1215MHz
Battlefield 4
1139MHz
1088MHz
Crysis 3
1177MHz
1113MHz
Mordor
1151MHz
1126MHz
Civilization: BE
1101MHz
1088MHz
Dragon Age
1189MHz
1189MHz
Talos Principle
1177MHz
1126MHz
Far Cry 4
1139MHz
1101MHz
Total War: Attila
1139MHz
1088MHz
GRID Autosport
1164MHz
1151MHz
Grand Theft Auto V
1189MHz
1189MHz
The far more interesting story here is GTX 980 Ti’s clockspeeds. As we have pointed out time and time again, GTX 980 Ti’s gaming performance trails GTX Titan X by just a few percent, this despite the fact that GTX 980 Ti is down by 2 SMMs and is clocked identically. On paper there is a 9% performance difference that in the real world we’re not seeing. So what’s going on?
The answer to that is that what GTX 980 Ti lacks in SMMs it’s making up in clockspeeds. The card’s average clockspeeds are frequently two or more bins ahead of GTX Titan X, topping out at a 64MHz advantage under Crysis 3. All of this comes despite the fact that GTX 980 Ti has a lower maximum boost clock than GTX Titan X, topping out one bin lower at 1202MHz to GTX Titan X’s 1215MHz.
Ultimately the higher clockspeeds are a result of the increased power and thermal headroom the GTX 980 Ti picks up from having two SMMs disabled. With those units no longer consuming power or generating heat, and yet the TDP staying at 250W, GTX 980 Ti can spend its power savings to boost just a bit higher. This in turn compresses the performance gap between the two cards (despite what the specs say), which coupled with the fact that performance doesn't scale lineraly with SMM count or clockspeed (you rarely lose the full theoretical performance amount when shedding frequency or functional units) leads to the GTX 980 Ti trailing the GTX Titan X by an average of just 3%.
Starting off with idle power consumption, there's nothing new to report here. GTX 980 Ti performs just like the GTX Titan X, which at 74W is second only to the GTX 980 by a single watt.
Meanwhile load power consumption is also practically identical to the GTX Titan X. With the same GPU on the same board operating at the same TDP, GTX 980 Ti ends up right where we expect it, next to GTX Titan X. GTX Titan X did very well as far as energy efficiency is concerned – setting a new bar for 250W cards – and GTX 980 Ti in turn does just as well.
As was the case with power consumption, video card temperatures are similarly unchanged. NVIDIA’s metal cooler does a great job here, keeping temperatures low at idle while NVIDIA’s GPU Boost mechanism keeps temperatures from exceeding 83C under full load.
Finally for noise, the situation is much the same. Unexpected but not all that surprising, the GTX 980 Ti ends up doing a hair worse than the GTX Titan X here. NVIDIA has not changed the fan curves or TDP, so this ultimately comes down to manufacturing variability in NVIDIA’s metal cooler, with our GTX 980 Ti faring ever so slightly worse than the Titan. Which is to say that it's still right at the sweet spot for noise versus power consumption, dissipating 250W at no more than 53dB, and once again proving the mettle of NVIDIA's metal cooler.
Finally, no review of a high-end video card would be complete without a look at overclocking performance.
From a design standpoint, GTX 980 Ti already ships close to its power limits. NVIDIA’s 250W TDP can only be raised another 10% – to 275W – meaning that in TDP limited scenarios there’s not much headroom to play with. On the other hand with the stock voltage being relatively low, in clockspeed limited scenarios there’s still some room for pushing the performance envelope through overvolting. And neither of these options addresses the most potent aspect of overclocking, which is pushing the entire clockspeed curve higher at the same voltages by increasing the clockspeed offsets.
GTX Titan X by comparison ended up being a good overclocker, and while we'd expect GTX 980 Tis to use slightly lower quality chips as part of the binning process, it should still overclock rather well.
GeForce GTX 980 Ti Overclocking
Stock
Overclocked
Core Clock
1000MHz
1250MHz
Boost Clock
1075Mhz
1326MHz
Max Boost Clock
1202MHz
1477MHz
Memory Clock
7GHz
8GHz
Max Voltage
1.187v
1.23v
Overall we're able to get another 250MHz (25%) out of the GTX 980 Ti's GPU, and another 1GHz (14%) out of its VRAM. This pushes the GTX 980 Ti's clockspeeds up to 1326MHz for the standard boost clock, and 1477MHz for the maximum boost clock. The card is heavily TDP limited at this point, so it's unlikely to sustain clockspeeds over 1400MHz, but working clockspeeds in the 1300MHz range are certainly sustainable. Meanwhile interestingly enough, this is actually a slightly better overclock than what we saw with the GTX Titan X; the Titan was only able to get another 200MHz out of its GPU and 800MHz out of its memory. So GTX 980 Ti ends up being the better overclocker by 50MHz.
The gains from this overclock are a very consistent across all 5 of our sample games at 4K, with the average performance increase being 20%. Though not quite enough to push the GTX 980 Ti above 60fps in Shadow of Mordor or Crysis 3, it is enough to crack 60fps on Battlefield 4 and The Talos Principle.
The cost of that 20% overclock in terms of power and noise is similarly straightforward. You're looking at an increased power cost of 30W or so at the wall – in-line with the 25W increase in the card’s TDP – while on the noise front the GTX 980 Ti is pushed out of its sweet spot. Card noise levels will increase by around 4.5dB(A).
Bringing this video card review to a close, through the last 14 pages of benchmarks we have seen the same result time and time again. While on paper the GeForce GTX 980 Ti should trail the GeForce GTX Titan X by several percent, what we get in the real world is much, much closer. With an average performance deficit of just 3%, GeForce GTX 980 Ti is for all intents and purposes GTX Titan X with a different name.
Typically NVIDIA engineers a wider gap between their cards, and while there is plenty of room for speculation here as to why they’d let GTX 980 Ti get so close to GTX Titan X – and make no mistake, it is intentional – at the end of the day none of that changes the final result. With a launch price of $649, the GTX 980 Ti may as well be an unofficial price cut to GTX Titan X, delivering flagship GeForce performance for 35% less.
As it stands GTX Titan X does have one remaining advantage that precludes it from being rendered redundant: its 12GB of VRAM, versus GTX 980 Ti’s 6GB. However without any current games requiring more than 6GB of VRAM – and any realistic workload running out of GPU throughput before running out of VRAM – the GTX Titan X’s place in this world now hinges on an uncertain degree of future-proofness. For this reason GTX Titan X isn’t going anywhere, it will still be around for buyers who need the very best, or even compute users after a cheap 12GB card, but for everyone else the GTX 980 Ti is now going to be the card all other high-end video cards are measured against.
Meanwhile for prospective high-end buyers who haven’t already picked up a GTX Titan X, GTX 980 Ti comes at an interesting time for new buyers and upgrades alike. NVIDIA’s previous $649 card, the GTX 780, has just turned two years old, which is about the bare minimum for upgrading a video card these days. Gamers looking to replace the GTX 780 will find that the GTX 980 Ti offers around a 70% performance improvement, which compared to the gains we saw with GTX Titan X and NVIDIA’s other Titan cards is actually ahead of the curve. It’s still not enough to double GTX 780’s performance, nor are we going to get there until 16nm, but it’s a bright spot for those who may want to upgrade a bit sooner than 2016. On the other hand GTX 780 Ti owners will almost certainly want to hold off for the next generation, despite the name.
That said however, today’s launch is just the first part of a larger battle between NVIDIA and AMD. With AMD scheduled to launch their next-generation high-end card in June, the launch of the GTX 980 Ti is in many ways NVIDIA striking first and striking hard. By pushing GTX Titan X-like performance down to $650, NVIDIA has set the bar for AMD: AMD needs to either beat GTX 980 Ti/Titan X if they want to take back the performance crown, or they need to deliver their card for less than $650. It goes without saying that NVIDIA has given AMD a very high bar to beat, but AMD has proven to be quite resourceful in the past, so it shall be interesting to see just what AMD’s response is to the GTX 980 Ti.
As for this moment, the high-end video card market is essentially in a holding pattern. The GeForce GTX 980 Ti is a fine card for the price – a GTX Titan X for $649 – however with AMD’s new flagship card on the horizon buyers are likely better off waiting to see what AMD delivers before making such a purchase, if only to see if it further pushes down video card prices.