Results 1 to 1 of 1
-
09-11-2019 10:31 PM #1
myfitegre PC Specs Laptop (Model) Razer Blade Stealth 13 Inch (Early 2019) Motherboard Asus Rog Zenith Extreme Processor AMD 1900x Threadripper (1st Gen) Memory (part number) F4-3200C14D-16GTZRX x4 Graphics Card #1 Powercolor RX 580 Red Dragon 8GBD5-3DHDV2/OC Monitor Sceptre 144 HZ Curved 27 Inch Gaming Monitor x3 Storage #1 Samsung 970 EVO PLUS 250 GB x3 Storage #2 Samsung 970 EVO 960 250 GB x1 CPU Cooler Enermax LiqTech TR4 V2 AIO Case Thermaltake Core P7 Power Supply EVGA SuperNOVA 850 G3, 80 Plus Gold 850W Keyboard Razer Blackwidow Elite Mouse Razer Mamba Headset ASTRO A40s TR Headset Mouse Pad Sean XXL Professional Large Mouse Pat & Computer Game Mouse Mat (35.43''W x 11.81H x 0.12TH) OS Windows 10 (Education License) Network Router Asus RT-AC66r Accessory #1 Blue Yeti Microphone
- Join Date
- Oct 2018
- Reputation
- 10
- Posts
- 3
Userbenchmark Program Causing RAID 0 to lose performance?
Hi Guys. I have Asus Rog Zenith Extreme with a 1900x TR Processor and 32gb of Gskill Trident Z RGB RAM With 3 Samsung NVME Drives.
After I boot into windows and run IOMeter at 32 bit Q Depth 1 MB file size Reads are pulling an astounding 9000 MB/Per Second. However, after benchmarking with other programs of about 20-30 mins my computer is then displaying Read speeds of about 3000 MB Per Second in iometer which is similar to using just one drive. I reboot the computer, and run Iometer again and it's back at 9K MB a second again. Anyone can explain what's doing this and how I can force my system to keep the raid array performing like it should? Thank you.
***UPDATE*** System has been running for over an hour and my C Drive is pulling 9K MB the whole time. I am wondering, is it possible that running a benchmark like userbenchmark or crystaldisk mark will lower your RAID speeds afterwards? That seems to be the only variable causing this...will test and update again. Has anyone else ever seen this?
***UPDATE 2*** Just Ran Crystal Disk Test and it did not effect my Iometer score, here were the results for documentation. (Note, a couple background apps were running during IOmeter so that's why it was a LITTLE lower no issue there)
***UPDATE 3*** YUP I just ran UserBenchmark, and Right afterwards I ran IOmeter and it dropped my RAID performance down to almost 1/3 of its' original performance. WHAT IS GOING on here that Userbenchmark kills my raid performance like that? Does anyone have a clue?
***UPDATE 4*** I have tested it repeatedly, and userbenchmark DOES permanently drop the performance of the Raid Array immediately afterwards. Also, I ran normal PC usage for a couple hours without running userbenchmark (LOL gameplay + crystal benchmark and other web browsing and file installations such as ROG CPU-Z) and then ran IOmeter at the same settings, and the Raid array had reverted to the 3000MBps low speed. That means that the performance drop is not Pinpointed to running user benchmark, and occurs as a result of other System Usage. Can someone please help? Why would performance drop during usage? Every time I reboot the computer, it's back to being fast (9k mbps). Thank you for reading this post.
***UPDATE 5*** So, I have discovered that every time I do userbenchmark something happens to the raid array and resets it to being treated as a single disk or something similar. HOWEVER, I noticed that if I open up AMD RAIDXpert2 web gui, and "Rescan" the disks, then immediately my performance is restored to the full power of the array. Help Please this is so frustrating I just want to get my RAID 0 Working.
***UPDATE 6*** AMD came out with a new version of the drivers u install during windows installation for making it a bootable RAID 0 Array. They also updated RaidXpert2. I thought this might fix it. I now have 4 NVME Samsung drives. I reinstalled, did IOMETER and am getting 10GBPS read speeds on the array. Then, I run Userbenchmark, and it's back to using the speed of only one drive at around 3GPS. Please help!! AGH!Last edited by myfitegre; 01-23-2020 at 05:03 AM. Reason: Additional Research