MSI GeForce GTX 1080 Ti ARMOR Review

Published: 4th May 2017, 19:25 GMT | Comments

Testing platform & methodology

Testing platform
CPU : Intel Core i7 6800K @ 4200 MHz
CPU Cooler : NZXT Kraken X61
Motherboard : ASUS X99-A II
Memory : G.Skill Trident Z DDR4 3200 MHz 16-18-18-38-2N
Display : Crossover 289K (4096×2160)
SSD : Samsung 950 Pro 512GB M.2
PSU : EVGA SuperNova 850 G2
OS : Windows 10 x64 PRO
GPUs :MSI GeForce GTX 1080 Ti Gaming X (GeForce 378.92 WHQL)
MSI GeForce GTX 1080 Ti ARMOR OC (GeForce 381.65 WHQL)
MSI GeForce GTX 1080 Gaming X (GeForce 381.65 WHQL)
 MSI GeForce GTX 1080 Gaming X (GeForce 378.92 WHQL)
COLORUFL GeForce GTX 1070 iGAME X-TOP (GeForce 378.92 WHQL)
 MSI GeForce GTX 1060 6G TOC (GeForce 378.92 WHQL)
HIS Radeon RX 480 IceQ2 Roaring Turbo  (Crimson ReLive Edition 17.3.3 BETA)
PowerColor RX 470 Red Devil (Crimson ReLive Edition 17.3.3 BETA)

New method of presenting data

Simple bar charts were replaced with more sophisticated time-based graphs, which will not only tell you the maximum, 1% low and average values but also the actual value at each stage of the test.

Under each graph, you will find four metrics: maximum, average, 1% low and delta. Those should be self-explanatory, but I will only mention 1% low is based on the 1% of the lowest values. This means that benchmark glitches, such as sudden performance glitches, should be eliminated from equations. Finally, the delta will tell what’s the difference between MAX and MIN values (the absolute values). This is a very interesting metric for clock speed or GPU temperatures.

Pay attention to the labels on the Y axis. Some charts will start from 0 (like frame-rate charts), but some will start from minimum values (like GPU temperatures or clock speeds). This is important to give you a better visual presentation of the data we collected.




by WhyCry