Home Page
  • March 28, 2024, 11:18:41 am *
  • Welcome, Guest
Please login or register.

Login with username, password and session length
Advanced search  

News:

Official site launch very soon, hurrah!


Author Topic: Hardware performance speed tests  (Read 8978 times)

Dakusan

  • Programmer Person
  • Administrator
  • Hero Member
  • *****
  • Posts: 535
    • View Profile
    • Dakusan's Domain
Hardware performance speed tests
« on: January 02, 2015, 02:15:30 am »


So I got a new computer back in April and have finally gotten around to doing some speed tests to see how different applications and settings affect performance/harddrive read speed.


The following is the (relevant) computer hardware configuration:
  • Motherboard: MSI Z87-GD65
  • CPU: Intel Core i7-4770K Haswell 3.5GHz
  • GPU: GIGABYTE GV-N770OC-4GD GeForce GTX 770 4GB
  • RAM: Crucial Ballistix Tactical 2*8GB
  • 2*Solid state hard drives (SDD): Crucial M500 480GB SATA 2.5" 7mm
  • 7200RPM hard drive (HDD): Seagate Barracuda 3TB ST3000DM001
  • Power Supply: RAIDMAX HYBRID 2 RX-730SS 730W
  • CPU Water Cooler: CORSAIR H100i
  • Case Fans: 2*Cooler Master MegaFlow 200, 200mm case fan

Test setup:

I started with a completely clean install of Windows 7 Ultimate N x64 to gather these numbers.

The first column is the boot time, from the time the start of the "Starting Windows" animation shows to when the user login screen shows up, so the BIOS is not included. I used a stopwatch to get these boot numbers (in seconds), so they are not particularly accurate.

The second and third columns are the time (in seconds) to run a "time md5sum" on cygwin64 on a 1.39GB file (1,503,196,839 bytes), on the solid state (SDD) and 7200RPM (HDD) drives respectively. They are taken immediately after boot so caching and other applications using resources are not variables. I generally did not worry about running the tests multiple times and taking lowest case numbers. The shown milliseconds fluctuations are within margin of error for software measurements due to context switches.


Results:

Boot times are affected between multiple steps, as seen below, but not too bad. The only thing that affected the MD5sum was adding the hardware mirror raid on the SSDs, which dropped the time of the md5 by half. So overall, antivirus and system encryption did not have any noticeable affect on the computer's performance (at least regarding IO on a single file and number crunching).


Numbers:
What was addedBootSSDHDDNotes
Initial installation4--
NIC Drivers and Cygwin74.6648.393I'm not sure why the boot time jump so much at this point. The initial number might have been a fluke.
All Windows updates + drivers + 6 monitors144.6188.393The boot time jumped up a lot due to having to load all the monitors
Raid 1 mirror[Windows] on SSDs + no page file174.6188.393This was removed once I realized Truecrypt could not be used on a dynamic disk (Windows software) RAID
Raid 1 mirror[hardware] on SSDs + no page file172.2468.408
Truecrypt System Volume Encryption (SSD Raid Only)17-182.2788.424
Antivirus182.3248.408Kaspersky 2014
Logged