AMD Ryzen Benchmarks: Ryzen 7 1700X Beats Core i7 6800K in 9 out of 13 Games Tested

An early Chinese review of AMD’s Ryzen 7 1700X has hit the web. The review pits AMD’s next-gen CPU against its Intel’s price equivalent counterpart, the Core i7 6800K, in 13 different games as well as includes the very first power consumption figures for Ryzen.

The Ryzen 7 1700X benchmarks showcase the $399 CPU outperforming $420 direct competitor, whilst offering two more cores, four more threads, higher clock speeds and a lower TDP. This is pretty impressive isn’t it?

Read More: AMD Ryzen 7 Wraith Max and Spire Coolers Leaked, Box Design Pictured

Before we move to the performance figures, I want to mention that these results are by no means official as the embargo on reviews hasn’t been lifted yet.

AMD Ryzen 7 1700X vs Intel Core i7 6800K Tested in 13 Games

The test results are in Chinese, but we’ve translated (thanks to VCZ) and compiled these in a separate table. First, here’s system specs used to carry out tests:

  • RX 480 8G,
  • Intel Core i7 6800K
    ASUS STRIX X99 GAMING
    16GB DDR4-2400,
  • AMD Ryzen 7 1700X
    ASUS PRIME X370-PRO
    16GB DDR4-2133,
  • Win10 64bit,
  • Crimson 17.2.1 drivers (Shared)

Ryzen 7 1700X benchmarks vs Core i7 6800K (Chinese)

Below is the list of tests in order:

– 3DMark Firestrike physics
– 3DMark11 physics
– Ludashi 5.15 CPU
– Ludashi 5.15 GPU
– Ludashi 5.15 Mem
– Ludashi 5.15 Disk
– Blender 2.78A
– 7-zip 15.05beta 1min
– LuxMark3.1 CPU OpenCL
– FritzChessbenchmark
– CinebenchR15 ST
– CinebenchR15 MT

The Ryzen 7 1700X beat the i7 6800K in 9 out of 13 games tested. The Zen based SKU couldn’t outpace its rival in Battlefield 1 and Call of Duty Infinite Warfare, while in other two games it did output higher average frame rates but lower minimum rates.

BenchmarkAMD Ryzen 7 1700XIntel Core i7 6800K1700X Lead in %
Platform Power – CPU only / full load123W126.87W3.15%
Platform Power – Standby62.77W98.74W57.30%
Platform Power – Gaming (CPU + GPU)154.66W194.2W25.57%
Platform Power – Office/Productivity81.55W113.5W39.8%
Ashes Of The Singularity (DX12) avg FPS46.845.33.31%
BF1 avg FPS81.7582.15-0.49%
COD 13 avg FPS87.388.57-1.43%
CS: GO avg FPS297.98284.124.88%
CS: GO min FPS2892755.09%
Crossfire (Lithtech Engine) 4.1.8 avg FPS1981970.51%
Crossfire (Lithtech Engine) 4.1.8 min FPS189191-1.05%
Deus Ex Human Divided (DX12) avg39.3390.77%
Deus Ex Human Divided (DX12) min32.131.80.94%
Doom (Vulkan) avg123122.50.41%
H1Z1 King of the Hill avg87.6383.54.95%
H1Z1 King of the Hill min75715.63%
Hitman (DX12) avg60.259.051.95%
Tomb Raider DX12 avg45.845.121.51%
Tomb Raider DX12 min30.130.5-1.31%
Civ 6 avg71.5262.115.17%
Civ 6 min52.994420.43%
The Division DX 12 avg63.959.67.21%
World of Tanks avg1171151.74%
World of Tanks min1071051.9%

The 1700X’s lead in gaming benchmarks is indeed great, but what impressed us more is the noticeable gap in TDPs with the CPU delivering more performance at 25-30% less power. Let’s hope we see similar benchmark results when official reviews surface on March 2nd.

AMD will start selling its new Ryzen 7 processors in early March. The first Ryzen shipment batch would entail over one million processors. The release includes three models, the Ryzen 7 1800X, 1700X and 1700, priced at US$499, US$399 and US$329, respectively.

AMD Ryzen CPU ModelCores/ThreadsBase ClockBoost ClockL3 CacheTDPSocketPrice
Ryzen 7 1800X8/163.6 GHz4.0 GHz16 MB95W-SR3+AM4$499 US
Ryzen 7 1700X8/163.4 GHz3.8 GHz16 MB95W-SR3+AM4$399 US
Ryzen 7 17008/163.0 GHz3.7 GHz16 MB65WAM4$329 US
  • Dennis Rassmann

    So, its <1.5% worse in the other games it didn't win? This would equate to 1-2 fps. Who exactly can notice that difference? 😉

    Better price, better power efficiency, better red.

    • Nicholes Pillar

      You mean orange

  • Joseph

    and the 1700x was used with 2133 mhz ram instead of 2400 mhz

    • Asgaeroth

      Could explain it’s marginal loss in a few titles. I watched a DigitalFoundry title awhile back that focused on RAM speeds for real world gaming performance and it can be a factor in certain titles. I remember them saying that in their testing, DDR4-2666 was the sweet spot to eliminate all detectable performance difference for system RAM in the range of titles they tested.