After years of running Flight Simulator X (FSX) and now Prepar3D (P3D) with older SATA drives and dealing with really long load times, it seemed like the perfect time to explore how an upgrade to SSD SATA might really perform. I ran a series of tests both with P3D on its own SSD drive and also in conjunction with the OS being SSD and found some interesting results.
The drive I ended up testing with was a SAMSUNG 850 EVO 2.5″ SATA III model MZ-75E1T0B/AM. At the time back in January these sold for around $289. Basic core specs have it at 540 MBps for Max Sequential read and 520 for max sequential write. This drive uses Samsung’s 3D V-NAND flash memory architecture and has built in thermal protection. It also only uses about 4 watts of energy when running and idle 0.05 watts.
Prepar3D Test Configuration
My base machine running this test consisted of the following specs both software and hardware:
- Windows 10 x64 v1511 with all updates
- Crystal Mark 3.02
- PerformanceTest version 8.0
- Prepar3D is installed entirely to one single drive in each case
- Prepar3D version 3.1 with these installed/active: FS Global 2010 + Oceania and Europe, REX Texture Direct, REX Soft Clouds, GEX Global, UTX USA v2, UTX Europe, A2A c172, FTX Vector 1.3, FTX NCA, FTX NA Fjords, FTX CRM, FTX PNW, FTX 1WA+2WA1+1S2 + FTX KORS Oras, Aerosoft Cleveland + Chicago + Las Vegas + Los Angeles + Boston + Detroit, Maryland Megascenery V2, PA Megascenery V2, Ultra Res Pittsburgh, Ultra Res New Orleans, FSDreamTeam all airports
- Saved Flight area: KORS Orcas, in the air approaching runway
- CPU: i7-4930k Hex Core CPU (hyperthreading and all cores on) overclocked to 4.4 GHz
- RAM: 16GB (4x4GB Gskill trident 2400’s 9-11-11-31-2T)
- Motherboard: Asus P9x79 Ivy Bridge-E
- Video: EVGA Geforce 980 Ti SC+ edition 6GB 06G-P4-4995-KR (not overclocked)
- Harddrives: The Samsung SSD MZ-75E1T0B/AM, Seagate 7200 RPM SATA III
Note that I didn’t make any tweaks to the ini files and this was a bare bones fresh install of P3D with all the addons listed.
As far as my P3D settings go, these are what I had set to during the testing…
Benchmarking Test Results
Here were the results of the old SATA III drive with just P3D installed on it, using CrystalMark 3.0.2
Here was the same data, but now on the SSD Samsung drive
Reads jump from 61.73 MB/sec to 505.3 MB/sec, not too shabby.
P3D Test Results
I ran two basic tests and repeated them a few times. I would load the KORS saved flight after having just rebooted, I will call this the “cold boot”. These first launches after a reboot always take the longest and I figured would give the most accurate result between the two. For the other test, after doing the cold boot launch of the saved flight, I just relaunch the saved flight and time it until in the cockpit again.
Here are the results.
-With the regular Non-SATA drive, the basic load times “cold” were around 11 min 37 seconds (697 seconds), while the load times after that “warm” were about 3min 10 seconds (190 seconds).
-With the SSD SATA drive being just the P3D drive, the “cold boot” was about 2 min 52 seconds (172 seconds), while the warm was 2min 10 seconds(130 seconds).
-With the SSD SATA drive + SSD OS drive, the “cold boot” was basically the same, 2 min 38 seconds (158 seconds) (slightly faster), while the warm was 2min 10 seconds (130 seconds) (identical)
In the chart below, I show the results of SATA vs SSD (SSD P3D drive + OS drive being SSD)
There is a boost to using SSD, although its only slightly better when using SSD on both the OS and the P3D dedicated drive, but the benefits of a fast loading OS outweigh the only slight increase with making the OS drive SSD. Cold boots are significantly quicker, though one might argue that warm boots aren’t that much different. For me, more often than not its a cold boot load for P3D. Its a welcome thing to see the load times basically go from 11 minutes to 3 minutes roughly, basically 77% faster and a win.