PNY GTX 980 XLR8 Pro Review & Benchmark - Revisiting the GM204

It's been a while since our last card-specific GTX 980 review – and that last one wasn't exactly glowing. Despite the depressing reality that 6 months is “old” in the world of computer hardware, the GTX 980 and its GM204 GPU have both remained near the top of single-GPU benchmarks. The only single-GPU – meaning one GPU on the card – AIC that's managed to outpace the GTX 980 is the Titan X, and that's $1000.

This review looks at PNY's GTX 980 XLR8 Pro ($570) video card, an ironclad-like AIC with pre-overclocked specs. Alongside the XLR8 Pro graphics card, we threw-in the reference GTX 980 (from nVidia) and MSI's Gaming 4G GTX 980 (from CyberPower) when benchmarking.

Previous GTX 980 Content

Other important GTX 980 articles include the below:

• GTX 980 & Maxwell Overclocking Guide
• GTX 980 Initial Review
• 2xGTX 980 SLI vs. GTX Titan X Benchmark

Note that benchmarks between previous articles may not be comparable to today's benchmark due to system configuration changes, methodology changes, and driver updates. Also note that this is strictly a GTX 980 benchmark, and so will feature no competing video cards. For cost analysis against AMD or other nVidia solutions, visit our original GTX 980 review.

PNY GTX 980 XLR8 Pro Specs

The core specifications of all GTX 980s are going to remain equivalent. They all share the GM204 GPU, and so will host the same CUDA Cores, ROPs, TMUs, and so on. The primary differentiator between competing board partners rests within the cooling solution, followed-up by aesthetics and pre-overclocked frequencies. This is something we discuss in greater depth in our GTX 960 round-up.

PNY's GTX 980 XLR8 Pro uses a dual-fan cooler. Straying from the reference squirrel-cage fan – as most board partners have done – the PNY device employs two push fans of approximately ~80mm in diameter. White LEDs are positioned behind the fans. Both fans are push fans and propel air straight into the aluminum fins, which are split into two clusters; one cluster is mounted directly to the GPU, with the other assisting in VRM & VRAM dissipation. Five copper heatpipes are routed through the aluminum heatsinks, conjoined at the coldplate sitting over the GM204 GPU.

The design of the cooler limits air intake at the front of the card, but leaves exhaust pathways out the top, bottom, and rear (expansion slot) of the card. What would normally be a front intake grill is shielded – likely for pure aesthetic reasons – with a rounded, unibody faceplate. Improved over PNY's new GTX 960, the GTX 980 XLR8 Pro does not reveal any unsightly cabling through the cooler, making for a tidy appearance.

A backplate is mounted to the XLR8 Pro, as found on most GTX 980s, and exhibits drilled-out holes for voltage checkpoints on solders.

PNY's faceplate is actually worth mention – and that's saying something, given our generally objective-exclusive hardware coverage. The XLR8 faceplate fronts a sophisticated, unoffensive look with its metallic-speckled unibody housing. If this were combined with the design of EVGA's GTX 960 SuperSC card – which uses flanking support plates on either side of the PCB – it'd be hard to look away.

As with the XLR8 GTX 960 we reviewed, PNY's core warranty is limited to 1-year, but extends to a lifetime warranty (non-transferrable) when the purchaser registers the hardware on PNY's website.

Test Methodology

We tested using our updated 2015 GPU test bench, detailed in the table below. Our thanks to supporting hardware vendors for supplying some of the test components.

The latest 350.12 GeForce driver was used during testing. Game settings were manually controlled for the DUT. Stock overclocks were left untouched for stock tests.

VRAM utilization was measured using in-game tools and then validated with MSI's Afterburner, a custom version of the Riva Tuner software. Parity checking was performed with GPU-Z. FPS measurements were taken using FRAPS and then analyzed in a spreadsheet.

Each game was tested for 30 seconds in an identical scenario on the two cards, then repeated for parity.

Average FPS, 1% low, and 0.1% low times are measured. We do not measure maximum or minimum FPS results as we consider these numbers to be pure outliers. Instead, we take an average of the lowest 1% of results (1% low) to show real-world, noticeable dips; we then take an average of the lowest 0.1% of results for severe spikes.

We conducted a large suite of real-world tests, logging VRAM consumption in most of them for comparative analysis. The games and software tested include:

• Assassin's Creed Unity (Medium, 1080p only).
• Far Cry 4 (Ultra 1080, Very High 1080).
• GRID: Autosport (Ultra 1440, Ultra 4K).
• Metro: Last Light (Very High + Very High tessellation at 1080; High / High at 1440).
• GTA V (Very High / Ultra at 1080p).
• Shadow of Mordor (Very High, 1080p).
• 3DMark Firestrike Benchmark

We already know ACU and Far Cry 4 consume massive amounts of video memory, often in excess of the 2GB limits of some cards. GRID: Autosport and Metro: Last Light provide highly-optimized benchmarking titles to ensure stability on the bench. Shadow of Mordor, GTA V, & Battlefield Hardline (not shown here) are new enough that they heavily eat RAM. 3DMark offers a synthetic benchmark that is predictable in its results, something of great importance in benchmarking.

Games with greater asset sizes will spike during peak load times, resulting in the most noticeable dips in performance on the 2GB card as memory caches out. Our hypothesis going into testing was that although the two video cards may not show massive performance differences in average FPS, they would potentially show disparity in the 1% low and 0.1% low (effective minimum) framerates. These are the numbers that most directly reflect jarring user experiences during “lag spikes,” and are important to pay attention to when assessing overall fluidity of gameplay.

Overclocked tests were conducted using MSI Afterburner for application of settings. All devices were tested for performance, stability, and thermals prior to overclocking to ensure clean results. On the OC bench, devices were set to maximize their voltage ceiling with incremental gains applied to the core clock (GPU) frequency. MSI Kombustor, which loads the GPU 100%, was running in the background. Once stability was compromised -- either from crashing or artifacting -- we attempted to resolve the issue by fine-tuning other OC settings; if stability could not be achieved, we backed-down the core clock frequency until we were confident of stability. At this point, the device was placed on a burn-in test using Kombustor and 100% load for 30 minutes. If the settings survived this test without logged fault, we recorded the OC settings and logged them to our spreadsheet.

Final OCs were applied and tested on two games for comparison.

Thermals were reported using Delta T over ambient throughout a 30-minute burn-in period using 3DMark FireStrike - Extreme, which renders graphics at 1440p resolution. This test loads the VRAM heavily, something Kombustor skips, and keeps the GPU under high load that is comparable to gaming demands. Temperatures were logged using MSI Afterburner.

3DMark Firestrike Ultra Performance

Minimal differences here – the cards are all stressed to the limit running the 4K Firestrike benchmark.

GTX 980 Assassin's Creed Unity Benchmark – 1440p / Very High

Assassin's Creed Unity is still among the most graphically-intensive games presently on-market; part of that is Ubisoft's trademark optimization (read: limited to none), though the game does legitimately stress graphics hardware and tax VRAM.

The results for ACU across the reference, MSI, and PNY devices are relatively uninteresting and are considered within margin of error. There is no clear victor from a pure performance standpoint with ACU.

PNY GTX 980 Far Cry 4 Benchmark – 4K / Ultra

Far Cry 4 shows marginally more interesting results than ACU, but we're still on a Ubisoft title. The average FPS results span just a 2-point disparity, favoring PNY slightly over its competitors. The extra ~12MHz OC doesn't necessarily prove immediately noticeable at stock settings.

GTX 980 GRID: Autosport Benchmark – 4K / Ultra

GRID finally exits margin of error. The game is among the most optimized we've ever tested, hence its long-standing presence in our bench suite. The XLR8 Pro leads the pack in AVG, 1%, and 0.1% FPS ratings, pushing just 2FPS past the MSI Gaming 4G card and 4FPS over reference.

GTX 980 GTA V Benchmark – 4K / Ultra & 1080p / Ultra

At 1080p, GTA V is pushed at 79FPS using the XLR8 Pro GTX 980, again at a slight 2FPS gain over the Gaming 4G card. Still nothing ultra noticeable here.

PNY GTX 980 – Metro: Last Light 4K / High Benchmark

All results are within margin of error for MLL at 4K resolution, with the game sinking 0.1% FPS rates (as it always does).

GTX 980 - Shadow of Mordor 4K / Ultra High Benchmark

Shadow of Mordor exhibits a ~3FPS range, with the XLR8 Pro falling slightly behind in 0.1% low FPS. No clear victor here, yet again.

PNY GTX 980 XLR8 Pro Overclocking Benchmarks

To overclock the video cards, we maximized the power percent target, the available voltage throughput, and then tweaked the memory and base clock settings until we found stability. For a better idea of how Maxwell overclocking works, check out our Maxwell OC guide using the GTX 980.

MSI's Kombustor was used to place GPUs under 100% load, forcing each card to convoke its highest boost clock; this isn't an easily tested endeavor using games, given the intentional throttling that occurs on the GPU in favor of thermals and utilization. Once Kombustor was hitting the GPU hard, we incrementally adjusted the clockrate until a point at which the drivers crashed or other instability was detected. As soon as this point was reached, the system was restarted and adjusted to a slightly lower number. We eventually found what seemed like a stable overclock, resulting in an intensive 30-minute burn-in test to ensure stability.

Memory clocks could likely be increased further – we stopped at 500MHz and focused on OCing the core clock.

Here's a table with the final settings for each card.

Everything overclocked better than the XLR8 Pro, which has the highest starting clock. We were able to push about 1506MHz before driver crashes, but experienced artifacting (green flickering, tearing) at this frequency. Returning to 1501MHz proved completely stable, using a 135MHz offset.

The MSI card gives the greatest granularity when overclocking, but does start at a marginally lower clock speed. For someone purely interested in overclocking, it would appear that the XLR8 Pro is not the ideal solution – though it is very stable when tested. The higher base frequency may appeal marginally to users who seek no thrills in manual overclocking and would rather buy something that's pre-configured, though ultimately, the results are really not all that different in the real world:

Our PNY card had a slightly lower ASIC quality than the other two devices, which likely contributed to the disparity between reference and the XLR8 Pro. It's likely that the boost in the XLR8 over reference in 4K GRID benchmarking was a result of the higher memory clock.

PNY vs. MSI, Reference GTX 980 Thermal Benchmark Results

Thermals were averaged after a thirty-minute run of 3DMark FireStrike Extreme. Other methodology is detailed in the methodology section. Thermals were measured at stock clocks and before overclocking was attempted.

PNY's XLR8 Pro is the coolest card among those tested. The MSI card – although equipped with an impressive cooler – runs at a slightly higher idle, likely due to the aggressive stance on silence. MSI's Gaming 4G card disables the fans when load is below a wattage threshold, something EVGA and ASUS also do. This keeps the card quieter, but a bit warmer as its idle baseline. PNY's dual-fan configuration keeps the card at ~41.7C after accounting for ambient, trailed closely by the Gaming 4G at 42.97C and distantly by the reference card at 47.7C.

Note that, among these temperatures, not one of these is in any way “bad.” Even the reference card's 47.7C is exceedingly cool and well below dangerous thresholds.

Conclusion: A Gaming Card with a Professional Look, But Still a 980

We find ourselves again in the position of neutrality. As with almost every card we've looked at for the Maxwell generation, the PNY XLR8 Pro performs effectively identically to reference and a nearby competitor. Thermally, although PNY's graphics card does pull ahead, it's not like any of these devices are particularly hot to begin with. It's not until we look at overclocking and the warranty that noteworthy differences emerge.

When it comes to user overclocking, the XLR8 Pro isn't all that great. MSI's Gaming 4G (and Gigabyte's G1, which we don't have on-hand) both excel at manual overclocking and are purpose-built for the task; enthusiasts who seek nothing more than the most granularity and highest clock speeds should look elsewhere from the XLR8 Pro. Users who prefer a stable, pre-overclocked device (read: no hassle) would do well to consider PNY's solution, especially with its registration-locked lifetime warranty. Most other manufacturers provide a 2-3 year warranty, so the lifetime support is potentially worth paying for.

This is what we ran into when looking at PNY's GTX 960, too – most of our readers will likely be replacing their video card within 2-3 years anyway purely to stay ahead of the graphics curve, so the usefulness of a lifetime warranty is contingent on your use case scenarios. We can't analyze that for you.

It's the warranty that is primarily being paid for with the $20-$30 price gap, though Newegg is currently selling the PNY XLR8 Pro for ~$550 after a $30 MIR, planting it firmly at MSRP for the reference device. It's a good buy at this price-point.

The aesthetics of the card, for as little as I like discussing appearances, are actually accommodating to most color combinations and front a “professional” appearance. To this end, PNY's “Pro” nametag is befitting of the card; it's classy, professional, and isn't as flashy as cards with the “gamer” aesthetic.

I sincerely hope that PNY reconfigures its display outputs in the future. I know what they're trying to do – they want to fit as many display-outs as possible on the card, and it's just not convenient. We had to order a mini-DP to DP cable just for the purpose of testing these cards, and that's going to be a commonality for anyone using a high-end display that only accepts DP input (see: G-Sync displays). DVI is present for a more standard connection type, but that still doesn't resolve the modern trend toward DP-only displays.

Further still, a simple adapter doesn't always work – because of the way the DisplayPort interface works, the conversion must occur at the video card, not the display-end converter/cable combination. This means that adapters – we tried DVI to DP and mini-HDMI to DP – will not output video to the display, at least with the 4K display we're using. Using a mini-DP to DP cable (no adapter – just a straight cable) works fine, but requires a special order.

Annoying, but considering how many monitors still host native DVI or HDMI, it's not a huge deal. Note that the retail version of the card includes an in-line mini-DP to DP adapter.

Performance is what we'd expect for the GTX 980. PNY's warranty, professional presentation of the card, and thermals make the XLR8 Pro a worthwhile consideration for those looking into GTX 980s.

- Steve “Lelldorianx” Burke.