Test Configuration

Test Configuration – Hardware

  • Intel Core i7-6700K (reference 4.0GHz, HyperThreading and Turbo boost is on to 4.6GHz; DX11 CPU graphics).
  • ASRock Z170M OC Formula motherboard (Intel Z170 chipset, latest BIOS, PCIe 3.0/3.1 specification, CrossFire/SLI 8x+8x)
  • HyperX 16GB DDR4 (2x8GB, dual channel at 3333MHz), supplied by HyperX/Kingston
  • Gigabyte RX Vega 56 Reference Edition, 8GB, stock clocks and overclocked, on loan from ToT.
  • GTX 1070 8GBFounders Edition, reference clocks and overclocked, supplied by NVIDIA
  • Gigabyte RX Vega 64 Liquid Cooled Edition, 8GB, reference clocks.
  • GTX 1080 8GB, Founders Edition, reference clocks, supplied by NVIDIA
  • GTX 1080 Ti 11GB, Founders Edition, reference clocks, supplied by NVIDIA
  • Two 2TB Seagate 7200 rpm SSHDs for each platform
  • EVGA 1000G 1000W power supply unit (for both platforms)
  • EVGA CLC280, 280mm CPU watercooler, supplied by EVGA
  • Onboard Realtek Audio
  • Genius SP-D150 speakers, supplied by Genius
  • Thermaltake Overseer RX-I full tower case, supplied by Thermaltake
  • ASUS 12X Blu-ray writer
  • Monoprice Crystal Pro 4K
  • ACER Predator X34 GSYNC display, supplied by ACER/NVIDIA

Test Configuration – Software

  • GeForce WHQL 385.41 used for the GTX 1070 FE.  WHQL 384.94 was used for the GTX 1080. WHQL 385.28 was used for the GTX 1080 Ti.  High Quality, prefer maximum performance, single display.  See control panel images below.
  • AMD ReLive Software 17.9.1 drivers were used for the benching the Vega 56 at stock clocks and 17.8.2 was used for the overclocked results.  See control panel image below.  17.30.1051-b6 (Public Launch drivers) used for the Vega 64 Liquid edition.
  • VSync is off in the control panel.
  • AA enabled as noted in games; all in-game settings are specified with 16xAF always applied
  • All results show average frame rates including minimum frame rates shown in italics on the chart next to the averages in smaller font.  Percentage increases are calculated between the average frame rates of the RX Vega 64 at stock and at overclocked settings, and also of the GTX 1080 between stock and overclocked settings.
  • Highest quality sound (stereo) used in all games.
  • Windows 10 64-bit Home edition, all DX11 titles were run under DX11 render paths. Our four DX12 titles are run under the DX12 render path. Latest DirectX
  • All games are patched to their latest versions at time of publication.
  • WattMan
  • MSI Afterburner.
  • OCAT
  • Fraps

The 28 PC Game benchmark suite & 4 synthetic tests

Synthetic

  • Firestrike – Ultra & Extreme
  • Time Spy DX12
  • VRMark Orange Room
  • VRMark Blue Room

DX11 Games

  • Crysis 3
  • Metro: Last Light Redux (2014)
  • Grand Theft Auto V
  • The Witcher 3
  • Rainbow Six Siege
  • DiRT Rally
  • Far Cry Primal
  • Call of Duty Infinite Warfare
  • Battlefield 1
  • Resident Evil 7
  • For Honor
  • Ghost Recon Wildlands
  • Mass Effect: Andromeda

DX12 Games

  • Tom Clancy’s The Division
  • Ashes of the Singularity
  • Hitman
  • Rise of the Tomb Raider
  • Total War: Warhammer
  • Deus Ex Mankind Divided
  • Gears of War 4

Nvidia Control Panel settings

We used MSI’s Afterburner to set the power and temp limits to their maximums.

NvCP1NvCP2

AMD Radeon Global Settings:

Here are the global game settings in AMD’s ReLive control panel that we use.  We used WattMan to set the Power and temp limits to their maximums.

Using WattMan for our regular non-overclocked settings, the Power Limit is set to +50%, the temperature is allowed to reach the maximum 70C, and the fan is also allowed to spin up to its maximum, but the core and memory clocks are unchanged.  WattMan is also used to set maximum RX Vega 56 overclocks.

Even with the latest Crimson 17.9.1 drivers, WattMan does not always hold its settings between reboots, so it’s important to check and reset your overclock as needed.  And we had to revert to 17.8.2 for our overclocked benchmarking.

Calculating Percentages

There are two methods of calculating percentages.  In a recent evaluation we used “Performance Differences” to compare the GTX 1080 and the GTX 1080 Ti versus the RX Vega 64 LC.  This time we are now using “Percentage Changes” which are usually used to show improvements in average frame rates (FPS) between stock and overclocked results.

For the percentage changes, we mean the increase in frame rates between the stock and the overclocked RX Vega 64 (and stock and overclocked GTX 1080), divided by the absolute value of the original stock frame rate in FPS, multiplied by 100.  Percentage change may be expressed by the algebraic formula where “V” is Value: ( ΔV / |V1| ) * 100 = ((V2 – V1) / |V1|) * 100

We also made a BTR community forum post that explains the process of calculating percentage differences using frame rates as an example.

Let’s head to the performance charts to see how the overclocked-to-the-max GTX 1070 FE compares with the overclocked-to-the-max RX Vega 56.

Contents

6 COMMENTS

    • Yeah this site left fan on automatic at 2012mhz lol. You know that is mostly will degrade into 19XX mhz coreclock when temp goes higher.
      Also Pretty bad 1070 overclocking.
      1070 should be easily stable at least ~2050-2100mhz with proper cooling,
      Many 1070 Memory also can reach +600, +700 or so. +500 is the most basic.

  1. Hrmm.. this article might be a bit old. But the results are pretty poor for the Vega. I don’t have any experience of overclocking 1070’s personally but my reference vega56 is just fine with core @ 1650 and memory @ 1100 mhz.

    As you state here most gains will come from overclocking the memory. It scales more or less perfectly with memory frequency. But the trick is to keep the hbm modules below 82 degrees celcius. Or preferably under 80 since they are starting to loose efficiency above that. And. At 92 they will start to throttle.

    The thing about temperatures is that from my experience the hbm modules tends to run about 20 degrees hotter than the core at load. Atleast when running memoryheavy applications. Mining is a good example. When it comes to games it’s not as bad.

    But if you try to do some quick testing keeping the core between 60-65 degrees you should be able to gain way more from overclocking the Vega. My gain is typicly 23-25% above stock with makes it a clear overclockwinner. But, I want to be clear about that it comes with an acoustic toll if using the reference cooler.

    My tests are based on a msi Vega 56 running vega64 bios (air) and I have replaced the paste on the chips with liquid metal. (Only gained about 4-5 degrees).

    Over and out

Comments are closed.