Testing configuration and testing method
All graphics card are tested in a system
with configuration as follow:
·
Motherboard: Intel Siler DX79SI (Intel X79
Express, LGA 2011, BIOS 0559 from 03/05/2013);
·
CPU: Intel Core i7-3970X Extreme Edition,
3.5/4.0 GHz (Sandy Bridge-E, C2, 1.1 V, 2 x 256 KB L2, 15 MB L3);
·
CPU cooler: Phanteks PH-TC14PE (2 x Corsair
AF140 900 RPM fan);
·
Thermal interface: ARCTIC MX-4;
·
Graphics card: NVIDIA GeForce GTX 770 3 GB 1,046/1,085/7,012
MHz; NVIDIA GeForce GTX 760 2 GB 980/1,033/6,008 MHz and 1,095/1,148/7,128 MHz;
Gigabyte GeForce GTX 670 2 GB Ultra Durable (GV-N670OC-2GD) 915/994/6,008 MHz and
1,090/1,169/7,508 MHz; HIS 7950 IceQ X² Boost Clock 3 GB (H795QMC3G2M) 850/5,000
MHz and 1,140/5,960 MHz; ASUS GeForce GTX 660 Ti DirectCU II TOP 2 GB
(TI-DC2T-2GD5) 915/993/6,008 MHz and 1,104/1,182/6,968 MHz;
·
System memory: DDR3 4x8 GB G.SKILL TridentX
F3-2133C9Q-32GTX (2,133 MHz, 9-11-11-31, 1.6 V);
·
System drive: Crucial m4 256 GB SSD (SATA-III,CT256M4SSD2,
BIOS v0009);
·
Drive for game and programs: Western Digital
VelociRaptor (300GB, SATA-II, 10000 RPM, 16MB cache, NCQ) inside Scythe Quiet
Drive 3.5” HDD quieter and cooler;
·
Backup drive: Samsung Ecogreen F4 HD204UI (SATA-II,
2 TB, 5400 RPM, 32 MB, NCQ);
·
System case: Antec Twelve Hundred (front panel: 3
Noiseblocker NB-Multiframe S-Series MF12-S2 1020 RPM fans; rear panel: 2 Noiseblocker
NB-BlackSilentPRO PL-1 1020 RPM fans; top panel: 200 mm 400 RPM standard fan);
·
Controlling and monitoring panel: Zalman
ZM-MFC3;
·
Power: Corsair AX1200i 1200 W (with a 120 mm fan);
·
Screen: 27” Samsung S27A850D (DVI-I, 2,560x1,440,
60 Hz).
There’re 5 graphics card participating in
this time’s testing session of us, and only one of them, GeForce GTX 770 is
tested in non-overclocking mode at nominal frequency:
GeForce
GTX 770 is tested in non-overclocking mode at nominal mode
This means of approaching is pretty
logical, as GTX 770 itself is the topping reference for GTX 760 and almost some
other testing participants.
The 3 remaining graphics cards are tested
at nominal mode as well as at the max frequency of these specific graphics
cards:
Gigabyte
GeForce GTX 670 2 GB
The
3 remaining graphics cards are tested at nominal frequency
To lower the dependence of graphics card
performance on the overall speed of the platform, I overclock our 6-core 32nm CPU
with the 48x multiplier, BCLK frequency set at 100 MHz and “Load-Line
Calibration” activated up to 4.8 GHz. Processor Vcore is increased to 1.38V in
motherboard BIOS:
Overclocking
6-core 32nm CPU with 48x multiplier
Hyper-Threading tech is activated. The 16
GB DDR3 system memory works at the frequency of 2,133 GHz with the time of 9-11-11-31
and voltage of 1.6V.
The testing session began on 20/06/2013. All
tests were performed in Microsoft Windows 7 Ultimate x64 SP1 with all important
updates within that day and drivers as follows:
·
Intel Chipset Drivers 9.4.0.1017 WHQL from 03/27/2013
for motherboard chip;
·
DirectX End-User Runtimes library from 30/11/2010;
·
AMD Catalyst 13.6 beta 2 driver from 06/04/2013
+ Catalyst Application Profiles 13.5 (CAP1) AMD-based graphics card;
·
Nvidia GeForce 320.39 Beta driver for GTX 760 and
GeForce 320.18 WHQL from 05/23/2013 for other Nvidia graphics card.
We tested the
graphics card performance in 2 resolutions: 1,920x1,080 and 2,560x1,440. The
test was performed in 2 image quality modes: “Quality+AF16x” – default
structured quality in driver with 16x isotropic filter activated and “Quality+ AF16x+MSAA
4x(8x)” with 16x isotropic filter activated and full-screen 4x or 8x
antialiasing, in these cases when average fps is still high enough to
experience comfortable gaming. We activate isotropic filter and full screen
antialiasing from gaming settings. if the correlative option is absent, we
change these settings in the Control Panels of Catalyst and GeForce drivers. We
discuss Vsync. There’re no other changes in driver settings.
Our benchmarks experience some adjustments.
Apart from the traditional update of the game having its appearance in this
list, we add 2 new names. They’re GRID 2 auto car reproduction and Company of
Heroes 2 real time strategy (trial version). As a result, our standard
benchmarks consist of the semi-synthetic benchmarks and 12 games than consume
the source of many types:
·
3DMark 2013 (DirectX 9/11) – version 1.0, benchmarking
in “Cloud Gate”, “Fire Strike” and “Fire Strike Extreme” scene;
·
Unigine Valley Bench (DirectX 11) – version 1.0,
max image quality settings, AF16x and (or) MSAA 4x, 1,980x1,080 res;
·
Metro 2033: The Last Refuge (DirectX 10/11) - version
1.2, max graphics quality settings, official benchmark, “High” graphics quality
settings; tessellation tech, disabled DOF and MSAA4x; activated antialiasing
AAA, 2 consecutive “Frontline” scene;
·
Total War: Shogun 2: Fall of the Samurai
(DirectX 11) – version 1.1.0, built-in benchmark (Sekigahara battle) at max
graphics quality and activated MSAA 8x in other testing modes;
·
Battlefield 3 (DirectX 11) – version 1.4, “Ultra”
image quality setting, 2 consecutive runs of the script from the beginning of
the mission of “Going Hunting” that lasts 100s;
·
Sniper Elite V2 Benchmark (DirectX 11) – version
1.05, we use Adrenaline Sniper Elite V2 Benchmark Tool v1.0.0.2 BETA with max
graphics quality settings ( “Ultra” configuration), Advanced Shadows: HIGH,
Ambient Occlusion: ON, Stereo 3D: OFF, 2 consecutive runs;
·
Sleeping Dogs (DirectX 11) – version 1.5, we use
Adrenaline Sleeping Dogs Benchmark Tool v1.0.2.1 with max image quality
settings install the Hi-Res Textures kit, disable FPS Limiter and V-Sync, 2
consecutive runs of the integrated benchmark with antialiasing quality at Normal
and Extreme levels;
·
Hitman: Absolution (DirectX 11) – version 1.0.446.0,
integrated test with Ultra image quality settings, activated tessellation tech,
FXAA and global lighting;
·
Crysis 3 (DirectX 11) – version 1.0.1.3, all
graphics settings at their max, Motion Blur – Medium, lens blur – on, activated
FXAA and MSAA4x modes, 2 runs of the script from the beginning of mission “Swamp”
that lasts 110s.
·
Tomb Raider (2013) (DirectX 11) – version 1.1.743.0,
we used Adrenaline Benchmark Tool, all image quality settings set to “Ultra”,
V-Sync disabled, FXAA and 2x SSAA antialiasing enabled, TessFX technology
activated, two consecutive runs of the benchmark built into the game;
·
BioShock Infinite (DirectX 11) – version 1.1.21.26939,
we use Adrenaline Action Benchmark Tool v1.0.2.1, 2 continuous tests of
integrated benchmark with “Ultra” quality settings and “Ultra + DOF”;
·
Metro: Last Light (DirectX 11) – version 1.0.2, we
use the integrated benchmark for 2 runs of D6 scene. All image quality and tessellation
settings at “Very High”, activated Advanced PhysX tech, we check when there’s
and there’s no SSAA;
·
GRID 2 (DirectX 11) – version 1.0.82.5097, we
use the integrated benchmark, max graphics quality settings at all means,
perform a running test with and without MSAA 8x antialiasing with 8 cars on
the Chicago race;
·
Company of Heroes 2 beta (DirectX 11) – version 3.0.0.9614,
2 consecutive runs of built-in benchmark at max image quality and physical
effect settings.
If the game
allows recording fps readings, they will also be put on the diagram. We run
each test or benchmark 2 times and take the best result for the diagram, but
only when the differences between them don’t surpass 1%. If they go beyond 1%, we
will run at least one more time to have the repeated result.
