User Benchmarks
From unRAID
Contents |
Parity Check Speeds
Add your Parity Check speeds here, to help others judge if their system is "performing well"
The Drive Count column is the total number of drives you have in your system, not including cache drive. If you system has a cache drive please specify it with a "(1)" behind the total number. Since the cache drive is not figured into the parity check it was decided not to "include" it as a "real" drive but to show that it did exist by adding the "(1)."
Two new columns were added on 4/13/09 to reflect the UnRAID version and if NCQ has been disabled or not. Multiple users have noted discrepancies in parity check and transfer speeds when NCQ is enabled if AHCI mode is used in the BIOS. Others using IDE mode have not seen the discrepancies.
If you would like to estimate what your parity check speed should be, please see "Calculating parity check speeds". Use the revised equation.
If you have upgraded your system and want to update your entry here, please add a new entry immediately below your previous entry, with the updated numbers. Both the old and new entries may be helpful to different users, and it is even more helpful when comparing the two entries, in showing how the changes have affected the parity check speed.
Note: This table should be helpful, but cannot be totally relied upon for a number of reasons - rounding errors, inconsistent conversion of syslog number, grabbing an interim speed instead of final average, user vs system started parity check, etc. A parity check started by the system is usually concurrent with drive mounting and transaction replay, slowing all disk operations down.
| Drive Count | # of IDE | # of SATA | Smallest (in GB) | Largest (in GB) | Motherboard | Controller Cards | Average Parity Check Speed (in MB/sec) | UnRAID Version | Force NCQ disabled |
|---|---|---|---|---|---|---|---|---|---|
| 12 | 12 | 0 | 300 | 500 | D865GLCLK | 2xPromise TX2 | 15 | 4.0 | n/a |
| 15 (1) | 12 | 4 | 300 | 1000 | D865GLCLK | 2xPromise TX2 + Adaptec 1205SA | 11 & 58 (once past IDE drives) | 4.3 | IDE mode |
| 14 (1) | 2 | 13 | 300 | 1000 | P5B VO DO | Adaptec 1430A (4D) + SuperMicro AOC-SAT2-MV8 (1D) | 44 & 58 (once past 500GB) | ||
| 3 (1) | (1) | 3 | 1500 | 1500 | GA-EP43-UD3L | Onboard (6x ICH10 + 1x JMicron) | 88.2 | 4.5b4 | Yes/ACHI |
| 5 (1) | 1 | 5 | 500 | 750 | GA-P35-DS3 | Onboard (4x ICH9R & 2x GIGABYTE SATA2) | 72 | ||
| 8 | 0 | 8 | 250 | 500 | nForce570 | SiI3132-based | 51 | ||
| 6 | 0 | 6 | 200 | 500 | D915GAG | Promise TX4 | 54 | ||
| 3 | 0 | 3 | 500 | 1000 | MSI 945GM1 (ITX) | Promise TX4(pci) | 46 & 57 (once past 500GB) | ||
| 3 | 0 | 3 | 1000 | 1000 | ABIT AB9 PRO | ICH8 Jmicron SiI3132 | 57 | ||
| 5 | 0 | 5 | 1000 | 1000 | MSI P43 Neo3-F | Onboard (6x ICH10 + 2x Jmicron) | 73 | 4.5b4 | Yes/AHCI |
| 7 (1) | 0 | 8 | 500 | 1000 | GA-P35C-DS3R | Onboard (6x ICH9 + 2x Gigabyte SATA2 Chip) | 59 | ||
| 9 (1) | 0 | 10 | 750 | 1500 | Asus A8N-SLI Premium | Onboard (nVidia4 SATA II & SI 3114R SATA I) + 2xAdaptec 1430SA + Sabrent SATA-PE2P | 61.43 w/ 2 750G on PCI SI3114 | 4.5-b6 | IDE mode |
| 4 (1) | 0 | 5 | 1000 | 1500 | ABIT AB9 Pro | Onboard (Intel ICH8R) | 83 | 4.3.3 | AHCI |
| 4 (1) | 2 | 2 | 160 | 1000 | Foxconn 45GMX | Onboard (Intel ICH7) | 48 | ||
| 3 (1) | 0 | 3 | 1000 | 1500 | Asus P5K Deluxe | Onboard (Intel ICH9R) | 78 | ||
| 4 (1) | 0 | 5 | 250 | 500 | VIA EPIA EN15000G | Onboard + Promise TX4 | 15.5 | ||
| 4 | 0 | 4 | 1500 | 1500 | Asus P5B Deluxe | Onboard (Intel ICH8R) | 90.9 | ||
| 3 (1) | 0 | 4 | 1500 | 1500 | Gigabyte GA-MA780G-UD3H | Onboard (AMD SB700) | 82.9 | ||
| 7 (1) | 0 | 8 | 500 | 1500 | ABIT AB9 Pro | Onboard (6xICH8 + 2xJmicron SiI3132) | 53 | 4.4.2 | No |
| 3 | 0 | 3 | 1500 | 1500 | ECS A740GM-M | Onboard (AMD SB700) | 95.21 | ||
| 4 | 0 | 4 | 1000 | 1000 | Foxconn G31MX-K | Onboard (Intel ICH7) | 65.65 | ||
| 7 (1) | 0 | 7 | 1000 | 1000 | SuperMicro C2SBA+II | Onboard (Intel ICH9R) | 62.56 | ||
| 5 | 0 | 5 | 1500 | 1500 | Gigabyte MA740GM-S2 | Onboard (AMD SB700) | 76.96 | 4.5-b6 | Yes/ACHI |
| 4 | 0 | 4 | 1000 | 1000 | XFX Geforce 8300 | Onboard (nv mcp78u) | 79.25 (syslog) | 4.5-b6 | No/ACHI |
| 6 | 0 | 6 | 500 | 1000 | XFX Geforce 8300 | Onboard (nv mcp78u) | 65.35 (syslog) | 4.5-b6 | No/ACHI |
| 6 | 0 | 6 | 500 | 1000 | Abit A-S78H | Onboard (AMD SB700) | 76.28 (syslog) | Yes/Native IDE | |
| 3 | 0 | 3 | 1500 | 2000 | Gigabyte MA785G-UD3H | Onboard (AMD SB710) | 65.73 (syslog) | 4.5 | ACHI |
| 6 | 0 | 6 | 1000 | 2000 | Gigabyte MA785G-UD3H | Onboard (AMD SB710) | 62.66 (syslog) | 4.5.3 | ACHI |
| 3 | 0 | 3 | 2000 | 3000 | ASUS M5A78L-M LX AM3+ | Onboard (AMD SB710) | 59.84 | 5.0-b7 | ACHI |
Hard Drive Read Speeds
Tentative new section, unfinished, for recording read speeds for many drive brands and models. Tentative columns below.
This table may be helpful in comparing the performance of various drives, as long as users are aware that there are many other factors that can affect true performance.
- Methods to determine speed
- hdparm -tT /dev/sdX - preferably averaged over 5 times, throwing out any that significantly vary
- <example of hdparm output here> highlighting the one number we want, the very last number
- See also Check Harddrive Speed
- dd if=/dev/sdX of=/dev/null count=8192000
- <example of dd output here> highlighting the one number we want, the very last number
- hdparm -tT /dev/sdX - preferably averaged over 5 times, throwing out any that significantly vary
- http://lime-technology.com/forum/index.php?topic=4305 - discussion of drive speeds
- http://lime-technology.com/forum/index.php?topic=10352 - some drive speeds, and includes a couple of drive speed testing scripts
- Notes are optional, should include unusual factors that may have affected the speed.
| Manufacturer | Model | Size (GB) | Spin Rate (RPM) | Method | Speed (MB/s) | Notes |
|---|---|---|---|---|---|---|
| Hitachi | HD32000 | 2000 | 7200 | hdparm | 122.36 | SATA |
| Hitachi | Deskstar 7K2000 | 2000 | 7200 | hdparm | 128.73 | SATA, ICH9R, Atom D510 |
| Hitachi | Deskstar 7K2000 | 2000 | 7200 | hdparm | 123.34 | SATA, ICH9R, Atom D510 |
| Maxtor | 6L300R0 | 300 | 7200 | hdparm | 60.9 | PATA |
| Maxtor | 7V300F0 | 500 | 7200 | dd | 74.8 | |
| Maxtor | 7V300F0 | 500 | 7200 | dd | 73.0 | |
| Samsung | HD501LJ | 500 | 7200 | hdparm | 84.1 | |
| Samsung | HD501LJ | 500 | 7200 | hdparm | 84.5 | |
| Samsung | HD103UJ | 1000 | 7200 | dd | 114.0 | |
| Samsung | Spinpoint F4 HD204UI | 2000 | 5400 | hdparm | 123.80 | SATA, ICH9R, Atom D510 |
| Samsung | Spinpoint F4 HD204UI | 2000 | 5400 | hdparm | 116.98 | SATA, ICH9R, Atom D510 |
| Samsung | SSD 830 Series | 128 | 0 | hdparm | 223.87 | Using as Cache |
| Sandisk | Cruzer | 4 | 0 | dd | 24.5 | |
| Seagate | ST3250823AS | 250 | 7200 | hdparm | 65.0 | |
| Seagate | ST3320620AS | 320 | 7200 | hdparm | 77.4 | |
| Seagate | ST3400620NS | 400 | 7200 | dd | 78.5 | |
| Seagate | ST3500630AS | 500 | 7200 | hdparm | 74.4 | |
| Seagate | ST3500630AS | 500 | 7200 | hdparm | 72.5 | |
| Seagate | ST31000340AS | 1000 | 7200 | dd | 106 | |
| Seagate | ST31500341AS | 1500 | 7200 | dd | 115 | |
| Seagate | ST31500341AS | 1500 | 7200 | dd | 122 | |
| Seagate | ST31500341AS | 1500 | 7200 | dd | 82 | PCI card, SiI3124 |
| Seagate | ST31500541AS | 1500 | 5900 | dd | 94 | |
| Seagate | ST32000542AS | 2000 | 5900 | dd | 125 | |
| Seagate | ST32000542AS | 2000 | 5900 | dd | 111 | ICH8 |
| Seagate | ST32000542AS | 2000 | 5900 | dd | 84 | PCI card, SiI3124 |
| Seagate | ST32000542AS | 2000 | 5900 | dd | 120 | PCI-e card, SiI3132 |
| Seagate | ST4000DX000 | 4000 | 7200 | hdparm | 178 | |
| Seagate | ST4000DM000 | 4000 | 5900 | hdparm | 143 | |
| WDC | WD5000KS-00MNB0 | 500 | 7200 | dd | 73.5 | |
| WDC | WD5000AAKS-00TMA0 | 500 | 7200 | dd | 85.6 | |
| WDC | WD5000AAKS-00TMA0 | 500 | 7200 | dd | 83.8 | |
| WDC | WD7500AAKS-00RBA0 | 750 | 7200 | dd | 96.0 | |
| WDC | WD10EADS-00M2B0 | 1000 | 5400 | dd | 104.0 | |
| WDC | WD10EADS-00L5B1 | 1000 | 5400 | dd | 116.0 | |
| WDC | WD10EACS-00ZJB0 | 1000 | 5400 | dd | 87.3 | |
| WDC | WD1001FALS-00J7B0 | 1000 | dd | 108.5 | ||
| WDC | WD20EADS-00R6B0 | 2000 | 5400 | dd | 104.7 | |
| WDC | WD20EADS-00R6B0 | 2000 | 5400 | dd | 98.6 | |
| WDC | WD20EADS-00R6B0 | 2000 | 5400 | dd | 101.7 | |
| WDC | WD20EADS-00R6B0 | 2000 | 5400 | dd | 100 | |
| WDC | WD20EARS | 2000 | 5400 | hdparm | 105.34 | SATA, ICH9R, Atom D510 |
Preclear Times
Tentative new section, unfinished, for recording how long it takes to preclear different drive brands and models. Tentative columns below.
This table may be helpful in setting expectations as to how long it will take to preclear a drive. Obviously, various factors may affect this, such as trying to preclear multiple drives at the same time, especially if they are on a shared bus. Information about the Preclear tool can be found in the UnRAID Add Ons page, Preclear Disk section.
- where to find the number
- Preclear Disk
- preclear_disk.sh - a new utility to burn-in and pre-clear disks for quick add - contains times to transfer here
- Preclear.sh results - Questions about your results? Post them here - contains times to transfer here
- Notes are optional, should include unusual factors that may have affected the times, such as ...
| Manufacturer | Model | Size | Time | Notes |
|---|---|---|---|---|
| Hitachi | HDS721050DLE630 | 500GB | 5.05 hours | (v1.13 pre-clear and OS v5.0 RC5) (Pre:107, Zero:122, Post:53 MB/s) |
| Hitachi | HDT721010SLA360 | 1TB | 13.5 hours | Deskstar 7k1000.B (Pre:80, Zero:87, Post:40 MB/s) |
| Hitachi | HD32000 IDK/7K | 2TB | 17.8 hours | only current 7200RPM 2TB drive |
| Hitachi | HDS722020ALA330 | 2TB | 25.5 hours | Deskstar 7K2000 7200RPM, avg. time 4 drives |
| Hitachi | HDS5C30 | 2.0TB | 24.9 hours | Deskstar 5K3000, v1.13 using -A, OS 4.7, running on ST2000DL003 at same time |
| Hitachi | HDS5C30 | 2.0TB | 24 hours | -A, OS 5.0rc8a, running on WD20EFRX 2TB at same time (94MB/s, 98MB/s, 44MB/s) |
| Hitachi | HDS723030ALA640 | 3TB | 36.05 hours | drive pulled from Hitachi External Drive |
| Hitachi | HDS5C4040ALE630 | 4TB | 46 hours | Deskstar 5k4000 (Pre:90, Zero:102, Post:48 MB/s) |
| Hitachi | HDS724040ALE640 | 4TB | 36.3 hours | Deskstar 7k4000 (Pre:114, Zero:130, Post:61 MB/s) |
| Intel | SSDSA2BW080G3 | 80GB | 0.79 hours | (v1.13 pre-clear and OS v5.0 RC5) (Pre:144, Zero:93, Post:57 MB/s) |
| Samsung | HD753LJ | 750GB | 12.84 hours | (v1.13 pre-clear and OS v5.0 RC5) (Pre:67, Zero:68, Post:31 MB/s) |
| Samsung | F1 HD103UJ | 1TB | 12.88 hours | 7200 RPM model (v1.8 using -A pre-clear and OS v4.7) |
| Samsung | F1 HD103UJ | 1TB | 13 hours | 7200 RPM model (v1.8 using -A pre-clear and OS v4.7) |
| Samsung | F1 HD103UJ | 1TB | 13.15 hours | 7200 RPM model (v1.8 using -A pre-clear and OS v4.7) |
| Samsung | F1 HD103UJ | 1TB | 13.25 hours | 7200 RPM model (v1.8 using -A pre-clear and OS v4.7) |
| Samsung | F3EG HD203WI | 2TB | 32.38 hours | 5400 RPM runs cool |
| Samsung | F4 HD204UI | 2TB | 25.5 hours | 5400 RPM model |
| Samsung | F4 HD204UI | 2TB | 26 hours | w/updated firmware and v1.8 -A pre-clear |
| Samsung | F4 HD204UI | 2TB | 26 hours | w/updated firmware and v1.8 -A pre-clear |
| Seagate | ST3320620AS | 320GB | 6.1 hours | 7200.10 (Pre:55, Zero:60, Post:29 MB/s) |
| Seagate | ST3500630AS | 500GB | 9.1 hours | 7200.10 (Pre:56, Zero:64, Post:30 MB/s) |
| Seagate | ST31000524AS | 1TB | 11.52 hours | Barracuda (v1.13 using -A pre-clear and OS v4.7 avg time two cycles) |
| Seagate | ST31500541AS | 1.5TB | 22.3 hours | LP 5900RPM model |
| Seagate | ST2000DL003 | 2TB | 24.7 hours | LP 5900RPM model |
| Seagate | ST32000542AS | 2TB | 26.1 hours | LP 5900RPM model |
| Seagate | ST2000DL003 | 2TB | 25.1 hours | LP 5900RPM, v1.13 using -A, OS 4.7, running on HDS5C30 at same time |
| Seagate | ST2000DM001 | 2TB | 16.2 hours | LP 7200RPM, v1.13 using -A, OS 5.0rc8a, avg time of two drives |
| Seagate | ST3000DM001 | 3TB | 23.66 hours | LP 7200RPM, v1.13, avg time, two passes |
| Seagate | ST3000DM001 | 3TB | 23.68 hours | LP 7200RPM v1.13 avg time over two cycles |
| Seagate | ST3000DM001 | 3TB | 23.88 hours | LP 7200RPM, v1.13, OS 5.0Beta6a |
| Seagate | ST4000DX000 | 4TB | 37.26 hours | v1.13, OS 5.0-rc11, (Pre:123, Zero:134, Post:55 MB/s) |
| Seagate | ST4000DM000 | 4TB | 38.05 hours | v1.13, OS 5.0-rc11, (Pre:116, Zero:132, Post:55 MB/s) |
| Western Digital | WD1600AAJS | 160GB | 2.59 hours | (v1.13 pre-clear and OS v5.0 RC5) (Pre:73, Zero:76, Post:31 MB/s) |
| Western Digital | WD1600JD | 160GB | 3.70 hours | Caviar Blue (v1.13 pre-clear and OS v4.7) |
| Western Digital | WD2500YD | 250GB | 5.86 hours | Caviar RE (v1.13 pre-clear and OS v4.7) |
| Western Digital | WD3200JS | 320GB | 7.00 hours | Caviar Blue (v1.13 pre-clear and OS v4.7) |
| Western Digital | WD4000AAKS | 400GB | 7.50 hours | Caviar Blue (v1.13 pre-clear and OS v4.7) |
| Western Digital | WD5000AAKS | 500GB | 8.35 hours | Caviar Blue (v1.13 pre-clear and OS v4.7) |
| Western Digital | WD5000AAKS | 500GB | 8.75 hours | Caviar Blue (v1.13 pre-clear and OS v4.7) |
| Western Digital | WD5000AAKS | 500GB | 5.85 hours | Caviar Blue (v1.13 pre-clear and OS v5.0 RC5) (Pre:98, Zero:109, Post:44 MB/s) |
| Western Digital | WD1001FALS | 1TB | 4.75 hours | Caviar Black (v1.13 pre-clear using -n format and OS v4.7) |
| Western Digital | WD1001FALS | 1TB | 4.75 hours | Caviar Black (v1.13 pre-clear using -n format and OS v4.7) |
| Western Digital | WD10EADS | 1TB | 16.5 hours | Caviar Green model avg. time 3 drives |
| Western Digital | WD10EADS | 1TB | 16.5 hours | Caviar Green model avg. time 3 drives |
| Western Digital | WD10FALS | 1TB | 12.5 hours | Caviar Black (v1.8 using -A pre-clear and OS v4.7) |
| Western Digital | WD10FALS | 1TB | 13 hours | Caviar Black (v1.8 using -A pre-clear and OS v4.7) |
| Western Digital | WD1502FAEX | 1.5TB | 15.09 hours | Black model avg time over two cycles |
| Western Digital | WD20EADS | 2TB | 25.5 hours | Caviar Green (v1.11 using -A pre-clear and OS v4.7) avg time over two cycles |
| Western Digital | WD20EADS | 2TB | 27 hours | Caviar Green model avg. time 2 drives |
| Western Digital | WD20EADS | 2TB | 27.45 hours | Caviar Green model avg. time over two cycles |
| Western Digital | WD20EADS | 2TB | 29 hours | Caviar Green |
| Western Digital | WD20EADS | 2TB | 30 hours | Caviar Green (v1.6 using -A pre-clear and OS v4.7) |
| Western Digital | WD20EARS | 2TB | 25 hours | Caviar Green (v1.11 using -A pre-clear and OS v4.7) |
| Western Digital | WD20EARS | 2TB | 25 hours | Caviar Green (v1.11 using -A pre-clear and OS v4.7) |
| Western Digital | WD20EARS | 2TB | 26 hours | Caviar Green (v1.3 using -A pre-clear and OS v4.7) |
| Western Digital | WD20EARS | 2TB | 26 hours | Caviar Green (v1.6 using -A pre-clear and OS v4.7) |
| Western Digital | WD20EARS | 2TB | 27 hours | Caviar Green (v1.11 using -A pre-clear and OS v4.7) |
| Western Digital | WD20EARS | 2TB | 27 hours | (jumpered) Caviar Green model avg. time 2 drives |
| Western Digital | WD20EARS | 2TB | 29 hours | Caviar Green (v1.4 using -A pre-clear and OS v4.7) |
| Western Digital | WD20EARS | 2TB | 30.1 hours | v1.6 using -A pre-clear and OS v4.7 Atom D510 ICH9R |
| Western Digital | WD20EARS | 2TB | 31.8 hours | v1.6 using -A pre-clear and OS v4.7 Atom D510 ICH9R |
| Western Digital | WD20EARX | 2TB | 21.0 hours | v1.13 using -A pre-clear and OS v4.7 avg time two cycles |
| Western Digital | WD20EARX | 2TB | 21.69 hours | v1.13 using -A pre-clear and OS v4.7 avg time two cycles |
| Western Digital | WD20EARX | 2TB | 22.59 hours | v1.13 using -A pre-clear and OS v4.7 avg time two cycles |
| Western Digital | WD20EARX | 2TB | 23.73 hours | v1.13 using -A pre-clear and OS v4.7 avg time two cycles |
| Western Digital | WD20EARX | 2TB | 26.3 hours | v1.13 using -A pre-clear and OS v4.7 |
| Western Digital | WD20EARX | 2TB | 26.5 hours | v1.13 using -A pre-clear and OS v4.7 |
| Western Digital | WD20EARX | 2TB | 29 hours | v1.13 using -A pre-clear and OS v4.7 AthlonII Neo 1.3GHz |
| Western Digital | WD20EARX | 2TB | 29.3 hours | v1.13 using -A pre-clear and OS v4.7 |
| Western Digital | WD20EFRX | 2TB | 21.5 hours | Red, -A, OS 5.0rc8a, same time as HDS5C30 (113 MB/s, 127 MB/s,44 MB/s) |
| Western Digital | WD30EFRX | 3TB | 24.5 hours | Red, -A, OS 5.0rc8a,(pre:138 MB/s, zero:130 MB/s, post:70 MB/s) |
| Western Digital | WD30EFRX | 3TB | 28.62 hours | v1.13 using -A pre-clear and OS v4.7 avg time two cycles |
| Western Digital | WD30EFRX | 3TB | 28.75 hours | v1.13 using -A pre-clear and OS v4.7 avg time two cycles |
| Western Digital | WD30EZRS-11J00B0 | 3TB | 33.8 hours | Green (Pre:96, Zero:100, Post:49 MB/s) |
| Western Digital | WD30EZRX-00DC0B0 | 3TB | 30.75 hours | Green (Pre:105, Zero:112, Post:54 MB/s) |
Performance Tests and Discussions
- Performance benchmarking using IOzone
- SSD as a Cache Drive (starts at this post, and continues on the next page)
- And you thought your parity speed was bad... :)
