and btw who thinks that putting nvme drives in raid 0 is a good idea, think again or give me a good example where you not getting diminishing returns by doing that
and btw who thinks that putting nvme drives in raid 0 is a good idea, think again or give me a good example where you not getting diminishing returns by doing that
Samsung 970 Evo Boot Drive:
Corsair MP510 Game Drive:
Old Storage HDD :
Last edited by banana999; 25-03-2019 at 09:35 AM.
Flip the question, what's wrong with putting NVMe drives in Raid 0? The only time you'll hit diminishing returns is when you start saturating your available PCIe lanes and their combined bandwidth afaik?
Even with an Intel system you should have enough PCIe lanes to support a GPU at x16 as well as multiple storage options, obviously if it's going to be an issue you choose your platform more carefully.
Unlike early raid drivers from intel and amd we now have trim support. Also utilizing multiple ssd vs one = thrashing less overall read & writes per day = longer MTBF. Longevity increased with the sharing of the work load.
Before this i ran 8x 2.5" ssd's in raid 0 for close to six years on an x99 chipset. Saw little to no degradation of performance over that time. Or i didnt notice it.
Samsung 970 Pro 512 GB, NVMe, 10% over-provisioning, on an Asus Prime Z390-A:
Sequential Read (Q= 32,T= 1) : 3548.924 MB/s
Sequential Write (Q= 32,T= 1) : 2327.210 MB/s
Random Read 4KiB (Q= 8,T= 8) : 1277.728 MB/s [ 311945.3 IOPS]
Random Write 4KiB (Q= 8,T= 8) : 625.738 MB/s [ 152768.1 IOPS]
Random Read 4KiB (Q= 32,T= 1) : 307.696 MB/s [ 75121.1 IOPS]
Random Write 4KiB (Q= 32,T= 1) : 274.673 MB/s [ 67058.8 IOPS]
Random Read 4KiB (Q= 1,T= 1) : 52.743 MB/s [ 12876.7 IOPS]
Random Write 4KiB (Q= 1,T= 1) : 122.007 MB/s [ 29786.9 IOPS]
Test : 1024 MiB [C: 34.5% (147.7/428.7 GiB)] (x5) [Interval=5 sec]
Date : 2019/03/24 19:26:53
OS : Windows 10 Professional [10.0 Build 17763] (x64)
Crucial P1 500 GB, NVMe, WTF my write speeds are garbage. (Thanks Hexus for prompting the check!)
-----------------------------------------------------------------------
CrystalDiskMark 6.0.1 x64 (UWP) (C) 2007-2018 hiyohiyo
Crystal Dew World : https://crystalmark.info/
-----------------------------------------------------------------------
* MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
* KB = 1000 bytes, KiB = 1024 bytes
Sequential Read (Q= 32,T= 1) : 1588.881 MB/s
Sequential Write (Q= 32,T= 1) : 57.225 MB/s
Random Read 4KiB (Q= 8,T= 8) : 362.736 MB/s [ 88558.6 IOPS]
Random Write 4KiB (Q= 8,T= 8) : 16.928 MB/s [ 4132.8 IOPS]
Random Read 4KiB (Q= 32,T= 1) : 173.461 MB/s [ 42348.9 IOPS]
Random Write 4KiB (Q= 32,T= 1) : 54.321 MB/s [ 13262.0 IOPS]
Random Read 4KiB (Q= 1,T= 1) : 44.685 MB/s [ 10909.4 IOPS]
Random Write 4KiB (Q= 1,T= 1) : 69.404 MB/s [ 16944.3 IOPS]
Test : 1024 MiB [C: 61.8% (286.6/463.9 GiB)] (x5) [Interval=5 sec]
Date : 2019/03/24 19:52:19
OS : Windows 10 Professional [10.0 Build 17134] (x64)
My Razer 15" 2019, 512 GB
Hmmm. Seems abnormally low considering the drive isn't full so there shouldn't be an impact from the cache being overused.
https://www.anandtech.com/show/13512...-ssd-review/10
Could be any number of things causing that.The Crucial P1 relies very heavily on its SLC cache to provide high performance, and that cache shrinks as the drive fills up. When the drive is full, a very write-heavy workload can overflow the cache and severely impact write speeds and even read performance to some extent. However when the drive isn't close to full, it is nearly impossible to fill the SLC cache with a realistic workload, and the Crucial P1 performs at least as well as any other entry-level NVMe drive, and sometimes rivals high-end NVMe drives.
500GB NVMe 970 EVO
1TB NVMe Toshiba something
1TB SATA 850 EVO
Plenty fast enough
-----------------------------------------------------------------------
CrystalDiskMark 6.0.1 x64 (UWP) (C) 2007-2018 hiyohiyo
Crystal Dew World : https://crystalmark.info/
-----------------------------------------------------------------------
* MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
* KB = 1000 bytes, KiB = 1024 bytes
Sequential Read (Q= 32,T= 1) : 5364.083 MB/s
Sequential Write (Q= 32,T= 1) : 4786.400 MB/s
Random Read 4KiB (Q= 8,T= 8) : 817.743 MB/s [ 199644.3 IOPS]
Random Write 4KiB (Q= 8,T= 8) : 599.346 MB/s [ 146324.7 IOPS]
Random Read 4KiB (Q= 32,T= 1) : 242.820 MB/s [ 59282.2 IOPS]
Random Write 4KiB (Q= 32,T= 1) : 193.682 MB/s [ 47285.6 IOPS]
Random Read 4KiB (Q= 1,T= 1) : 43.868 MB/s [ 10710.0 IOPS]
Random Write 4KiB (Q= 1,T= 1) : 187.498 MB/s [ 45775.9 IOPS]
Test : 1024 MiB [C: 39.5% (550.9/1394.3 GiB)] (x5) [Interval=5 sec]
Date : 2019/03/25 15:45:32
OS : Windows 10 Professional [10.0 Build 17134] (x64)
-----------------------------------------------------------------------
CrystalDiskMark 6.0.1 x64 (UWP) (C) 2007-2018 hiyohiyo
Crystal Dew World :
-----------------------------------------------------------------------
* MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
* KB = 1000 bytes, KiB = 1024 bytes
Sequential Read (Q= 32,T= 1) : 3546.129 MB/s
Sequential Write (Q= 32,T= 1) : 2330.232 MB/s
Random Read 4KiB (Q= 32,T= 8) : 898.974 MB/s [ 219476.1 IOPS]
Random Write 4KiB (Q= 32,T= 8) : 430.110 MB/s [ 105007.3 IOPS]
Random Read 4KiB (Q= 32,T= 1) : 246.179 MB/s [ 60102.3 IOPS]
Random Write 4KiB (Q= 32,T= 1) : 152.373 MB/s [ 37200.4 IOPS]
Random Read 4KiB (Q= 32,T= 1) : 246.266 MB/s [ 60123.5 IOPS]
Random Write 4KiB (Q= 32,T= 1) : 150.528 MB/s [ 36750.0 IOPS]
Test : 1024 MiB [C: 74.3% (353.9/476.3 GiB)] (x5) [Interval=5 sec]
Date : 2019/03/26 22:37:56
OS : Windows 10 [10.0 Build 17763] (x64)
Samsung 970 PRO 512GB NVMe drive. I had it since the drives were basically available in Canada. I have never done a firmware update. Also I have used this system for almost a year now. I have been running the same windows image since I first started. I have a nearly full drive as you can see. Still the performance is impressive!
There are currently 1 users browsing this thread. (0 members and 1 guests)