|
|||||||
| Hardware Discussion & Support Discuss your computer - its components or ANY hardware, past/current/future you want, or ask our forum experts if you have a general problem with your hardware. |
![]() |
|
|
Thread Tools |
|
|
#1 |
|
confutatis maledictis
|
OK, in Sandra, it gives 3 different sets of read/write speeds after it tests your HD.
Buffered, Sequential, and Random. How much weight should each of these be given? I figure Buffered is probably least important, but what about the other two? (Do people write to their HD more often Sequentially or Randomly?) And how much of a difference is in their importance? For example: 50% / 35% / 15% ? ...hope I wasn't too confusing... thanks
__________________
Digitalis 3.3 Athlon 64 3000 // ASUS K8V SE Deluxe // 1024MB PC3200 (2-2-2-10 1T)
ATI All-In-Wonder 9700 Pro // 20" Dell 2005FPW (DVI) M-Audio Revo 7.1 + Philips Acoustic Edge // Klipsch ProMedia 2.1 320/16 Western Digital WD3200KS + 120/8 Seagate 7200.7 NEC ND-3550A 16x DVD±RW + Lite-On 52x24x CD-RW Antec Sonata case // 480W Antec TruePower personal bests || Aq'3: 46796 | 3D'01: 20461 | 3D'03: 6336 | 3D'05: 2677 | PC'04: 4605 | PC'02: 7691,9092,1250 |
|
|
|
|
|
#2 |
|
Twice the fun!
Join Date: Jul 2002
Posts: 1,404
Rep Power: 0 ![]()
|
I don't bother using SiSoft Sandra to test my hd speeds
I use nBench from http://www.raid.com (under .EDU - benchmark utilities - windows) it tells you exactly how many MB/sec read and write your drives can handle I test my WesternDigital 30G 7200 on my P4 at 29 write and 30 read on a 100mb test file. My 120G WD's in my server, however, test much higher. ![]() Sandra is too confusing to convert its data into a general comparison. |
|
|
|
|
|
#3 |
|
E Pluribus Unum
Join Date: May 2002
Posts: 3,203
Rep Power: 0 ![]() |
True that. I could never get SANDRA to actually work. HDTach is excellent, although you can't test write speeds without wiping out your data. ^_^ It's good for testing new drives because it'll work even with the drive unpartitioned and unformatted.
|
|
|
|
|
|
#4 | |
|
HardwareHeaven Senior Member
Join Date: May 2002
Posts: 916
Rep Power: 0 ![]() |
Quote:
Sequential represents reading and writing involving large files that are generally defragmented. Large game map files are a good example. These files will have their sectors arranged one after another, so this measurement represents linear track-to-track speed. If there is any fragmentation, however, the drive must switch from sequential to non-sequential or random methods. Random is generally the most common type of hard-drive access. This represents the many small files that are scattered all over your hard drive, particularly files such as internet browser caches. It also represents the state of most users' hard drives, with many files being fragmented. As such, this measurement is crucial in determining an overall performance rating, as it's probably the most used. In terms of real-world performance, this is one to watch carefully. Some drives perform exceptionally well going sector-to-sector, track-to-track in linear fashion, but fall apart when asked to randomly seek out a particular sector. There are ways to improve your performance. The cheapest is probably a good disk defragmenter, such as Diskeeper 7. You can download a trial at www.diskeeper.com and see for yourself. Drive striping can enhance performance tremendously, but is limited by factors such as hardware versus software RAID, individual drive performace, and your hard drive controller's performance. If you're considering drive striping, get a decent RAID controller if your board doesn't already have one. And try if at all possible to use matching hard drives and drive sizes for this - it really does make a difference. Another performance enhancer is partitioning. Yes, the simple act of changing that 80 gig drive into 4 20 gig drives can make all the difference in the world in terms of performance. With the larger drive comes larger sweeps of the drive arm for those random reads and writes, which dramatically slows you down. Make the area it's working with smaller, however, and you can add zip beyond the dreams of avarice. So don't sit there drooling because someone's got a single 100 gig drive and you've got 3 40's - chances are if you set yours up correctly, you'll be in playing that game before they finish with the splash screen
|
|
|
|
|
|
|
#5 | |
|
Always newbie & cocky kid
Join Date: Jul 2002
Posts: 439
Rep Power: 0 ![]() |
Quote:
Thank you for writing all the above, very understandable, i have qted the best part that i like to learn more, also like to try it.. so i would apprec if you can explain exactly of how to? to the "Drive striping" my MB doesn't has RAID.. it's Asus CUSL2, but i got 2 HDD.. IBM 120gxp both in same size. Regards, |
|
|
|
|
|
|
|
|
|
confutatis maledictis
|
Quote:
Buffered Read ....... 84 MB/s Sequential Read .... 46 MB/s Random Read ......... 8 MB/s Buffered Write ...... 85 MB/s Sequential Write ... 44 MB/s Random Write ....... 13 MB/s Which part is confusing? And JavaFox, why can't you get it to work? Just double-click File System Benchmark, right? Do you mean it would hang or something? ![]() Shaith... LOL, I knew what the different speeds meant I just wanted to know how important they each are. Are you saying most important is Random, then Sequential, then Buffered?Also, what range of Random speeds is considered to be "good?" Thanks for the reply, btw
__________________
Digitalis 3.3 Athlon 64 3000 // ASUS K8V SE Deluxe // 1024MB PC3200 (2-2-2-10 1T)
ATI All-In-Wonder 9700 Pro // 20" Dell 2005FPW (DVI) M-Audio Revo 7.1 + Philips Acoustic Edge // Klipsch ProMedia 2.1 320/16 Western Digital WD3200KS + 120/8 Seagate 7200.7 NEC ND-3550A 16x DVD±RW + Lite-On 52x24x CD-RW Antec Sonata case // 480W Antec TruePower personal bests || Aq'3: 46796 | 3D'01: 20461 | 3D'03: 6336 | 3D'05: 2677 | PC'04: 4605 | PC'02: 7691,9092,1250 |
|
|
|
|
![]() |
| Bookmarks |
| Thread Tools | |
|
|