The 2019 GPU Benchmark Suite & The Test

As we’re kicking off a new(ish) generation of video cards, we’re also kicking off a new generation of the AnandTech GPU benchmark suite.

For 2019 most of the suite has been refreshed to include games released in the last year. The latest iteration of the Tomb Raider franchise, Shadow of the Tomb Raider, is 2019’s anchor title and is the game used for power/temperature/noise testing as well as game performance testing. Also making its introduction to the GPU benchmark suite for the first time is an Assassin’s Creed game, thanks to Assassin’s Creed Odyssey’s extra-handy built-in benchmark.

For 2019 Ashes of the Singularity has been rotated out, so we’re empty on RTSes at the moment. But as an alternative we have Microsoft’s popular Forza Horizon 4, which marks the first time a Forza game has been included in the suite.

AnandTech GPU Bench 2019 Game List
Game Genre Release Date API
Shadow of the Tomb Raider Action/TPS Sept. 2018 DX12
F1 2019 Racing Jun. 2019 DX12
Assassin's Creed Odyssey Action/Open World Oct. 2018 DX11
Metro Exodus FPS Feb. 2019 DX12
Strange Brigade TPS Aug. 2018 Vulkan
Total War: Three Kingdoms TBS May. 2019 DX11
The Division 2 FPS Mar. 2019 DX12
Grand Theft Auto V Action/Open world Apr. 2015 DX11
Forza Horizon 4 Racing Oct. 2018 DX12

All told, I’m pleasantly surprised by the number of DirectX 12-enabled AAA games available this year. More than half of the benchmark suite is using DX12, with both AMD and NVIDIA cards showing performance gains across all of the games using this API. So this is a far cry from the early days of DX12, where using the low-level API would often send performance backwards. And speaking of low-level APIs, I’ve also thrown in Strange Brigade for this iteration, as it’s one of the only major Vulkan games to be released in the past year.

Finally, I’ve also kept Grand Theft Auto V as our legacy game for 2019. Despite being released for the PC over 4 years ago – and for game consoles 2 years before that – the game continues to be one of the top selling games on Steam. And even with its age, the scalability of the game means that it’s a heavy enough load to challenge even the latest video cards.

As for our hardware testbed, it too has been updated for the 2019 video card release cycle.

Internally we’ve made a pretty big change, going from an Intel HEDT platform (Core i7-7820X) to a standard desktop platform based around an overclocked Core i9-9900K and Z390 chipset. While we’ve used HEDT platforms for the GPU testbed for the last decade, HEDT is becoming increasingly irrelevant/compromised for gaming; while the extra PCIe lanes are nice, these platforms haven’t delivered the best CPU performance for games as of late.

By contrast, desktop processors with 8 cores now provide more than enough cores, and they also provide far better clockspeeds, delivering more of the single/lightly-threaded performance that games need. Furthermore, as SLI and Crossfire are on the rocks, the extra PCIe lanes aren’t as necessary as they once were.

On a side note, I had originally hoped to cycle in a Ryzen 3000 platform at this point, particularly for PCIe 4.0. However the timing of all of these hardware launches meant that we needed to go with an established platform, as it takes a week or so to build and validate a new GPU testbed. Plus with Ryzen 3000 not launching for another week, we wouldn’t have been able to use it for this review anyhow.

Otherwise the rest of our 2019 GPU testbed is relatively straightforward. With 32GB of RAM and a high-end Phison E12-based NVMe SSD, the system and any video cards being tested as well-fed. Enclosing all of this for our real-world style testing is our trusty NZXT Phantom 630 Windowed Edition case.

 

CPU: Intel Core i9-9900K @ 5.0GHz
Motherboard: ASRock Z390 Taichi
Power Supply: Corsair AX1200i
Hard Disk: Phison E12 PCIe NVMe SSD (960GB)
Memory: G.Skill Trident Z RGB DDR4-3600 2 x 16GB (17-18-18-38)
Case: NZXT Phantom 630 Windowed Edition
Monitor: Asus PQ321
Video Cards: NVIDIA GeForce GTX 2070 Super Founders Edition
NVIDIA GeForce GTX 2060 Super Founders Edition
NVIDIA GeForce GTX 2080 Founders Edition
NVIDIA GeForce GTX 2070 Founders Edition
NVIDIA GeForce GTX 2060 Founders Edition
AMD Radeon RX Vega 64
Video Drivers: NVIDIA Release 431.15
AMD Radeon Software Adrenalin 2019 Edition 19.6.3
OS: Windows 10 Pro (1903)
Meet the GeForce RTX 2070 Super & RTX 2060 Super Shadow of the Tomb Raider
Comments Locked

281 Comments

View All Comments

  • Korguz - Tuesday, July 2, 2019 - link

    um yea ok.. sure...
  • eva02langley - Tuesday, July 2, 2019 - link

    Well, AMD will introduce bundles, games and rebates for Ryzen and Navi launch. I will not be surprised to see Navi cut down by 50$ in August.
  • Fritzkier - Tuesday, July 2, 2019 - link

    Navi should be cheaper IMO. Navi has way more smaller die (2.5x more smaller than Vega 64 if I recall) and uses GDDR6 instead of HBM. I don't know why AMD priced them that high tho...
  • eva02langley - Tuesday, July 2, 2019 - link

    Because Nvidia was asking for 500-600$ for less performances.
  • Meteor2 - Saturday, July 6, 2019 - link

    *was*
  • edzieba - Tuesday, July 2, 2019 - link

    "the performance, partially a consequence of going with 12nm, just isn’t there"

    People should have been weaned off this by now: process shrinks stopped inherently boosting performance years ago. Power consumption drops and perf/watt increases, but 'perf/transistor' continues to decrease (due to leakage increasing as packing density grows, coupled with power density increases) as it has done for some time, and cost/transistor has been going up since 28nm. A brief period of making dies bigger and bigger (and more and more expensive) has culminated in reticle-limit dies like GV100 and TU102, but that now makes start the wall process scaling hit some time ago in reality.
    This is only going to continue as processes shrink further. Cost/transistor will rise, perf/transistor will drop, and increasing performance means dies will continue to grow. Performance gains will continue to come from architectural changes, not process changes. Unless you're hitting the reticle limit AND cannot split your die into multiple dies due to latency reasons only then does it make any sense to move to a smaller process, and you will take a hit to both cost/perf as well as perf/transistor in doing so which may eat any gains from packing more transistors in.
  • Threska - Tuesday, July 2, 2019 - link

    Chiplets.
  • Hixbot - Thursday, July 4, 2019 - link

    I agree that performance per transistor can drop with die sizes due to leakage. And that new nodes are expensive at first. But once a node is mature, the cost per transistor should be lower than previous node. If that wasn't the case than the semi conductor business would be completely sunk.
  • Hixbot - Thursday, July 4, 2019 - link

    Edit: I ageee that performance per transistor can drop with process shrinks
  • Gastec - Wednesday, July 17, 2019 - link

    Changes and improvements in software need to be done which are going very slowly because it takes a lot of work(coding) and knowledge but mostly hard work which younger generations are not willing to do (distracted by social networking and gaming).

Log in

Don't have an account? Sign up now