HDD benchmark

[Alec§taar]

i will do it, but i'm a bit busy today and tomorrow i'm going on vacation again, keep ppl updated while i'm not here if you can

Heh, ok... I THINK I can handle "juggling" 3 of these tests charts (ScienceMark 2.0, AquaMark 3, & HD Tach 3.0), if need be!



(OK, if/when I do this? We are "even-steven/squared-up" for your providing the idea to add CPU types/RAM etc. in the ScienceMark 2.0 test too, deal?)

LOL!

* Gotta set the contract now & all that... lol!

APK

P.S.=> You build the "base template" for the categories, & I can maintain them when you split & go enjoy life (and all that good stuff)... apk

 

[_33]

scores updated, and new rules (can enter more than one hdd)
can anyone tell me how 2x80gb hitachi beated 2xraptors?
_33 wich score should i post?

Didn't you want to use the fast test scores? I personally prefer the longer test scores as they are probably little more accurate (if that is possible). I would also (if I were you) indicate "average read rate" which you can get from everyone's screen capture.

Good luck!

EDIT: Forgot to mention that the title is inapropriate IMHO because we are dealing here with HDTACH and not ANY HDD performance utility. ATTO Disk Benchmark is another one I trust for brute force HDD performance.

 

[pt]

Heh, ok... I THINK I can handle "juggling" 3 of these tests charts (ScienceMark 2.0, AquaMark 3, & HD Tach 3.0), if need be!

ok, i'm going to be out for more than a week, if you can "juggle" with them, and want to, do it

i think the sm2 thread is dying but when i get the psu and cooler it will probabily be back to life

 

[Alec§taar]

ok, i'm going to be out for more than a week, if you can "juggle" with them, and want to, do it

Sure/certainly/will do... why not?

Again: As I find harddisks the MOST interesting portion of a system... it piques the curiousity to see how mine stacks up, vs. others here.

Consider it done.

i think the sm2 thread is dying but when i get the psu and cooler it will probabily be back to life

Oh, I agree, and in a way? I am actually GLAD it's finally "dying down" some... it kept me extremely busy (relative term) @ times... sometimes, too much so.

* Anyhow - go enjoy your free time man! We'll see you on the bounce...



APK

 

[pt]

we have to get a memory only benchmark too, so all covers are based

 

[Alec§taar]

we have to get a memory only benchmark too, so all covers are based

Yes, I think that's in order... but, I think that Super-Pi 1.5 mod test that giorgos.th ran is such a test.

APK

 

[_33]

we have to get a memory only benchmark too, so all covers are based

There is a memory quote/bench done in the Sciencemark 2.0. Also, there are memory benchmarks in Everest and in Sisoft Sandra.

Cheers!

 

[cdawall]

i just found this here my old shool maxtor
ata 133
not bad for ancient tech its almost 6-7yrs old

 

[Steevo]

Between his system and your, I would say yours is badly configurated, and his is perfectly set up. How big is your stripe size?

CRAPPY RAID PERFORMANCE THREAD

Burst size is extremely high on his due to the ICHR7 being integrated into the southbridge controller and having full bandwidth available, and the fact that it is reading out of the onboard cache, Hitachi drives are very quick, and seem to make better use of the cache than other drives.

http://www.osnn.net/forum/showthread.php?t=81436

However, look at Average speed, and sustained output for the drives. Sure based on burst speed it kicks my systems ass, but how often is the next level preloaded into the drives cache?

 

[Deleted member 24505]

i guess i'm still at the top then.lol

come on there must be someone else who has hitachi drives to compete.

how come no one else is having this "anomoly"

 

[Alec§taar]

Good Morning Everyone (8 a.m. here, & I just got up, having my coffee & re-reading this thread now):

Cdawall, & _33: I'll be charting in those results in lieu of pt (man's on vacation enjoying his life) today, so rest assured, it will be done shortly.

Steevo: VERY interesting point & read!

Tigger69: Again, "DAMN YOU" lol... (see my post about you above!)



* Anyhow, gimme some time... I will:

1.) Add new scores

2.) Bust the chart up more (don't think we have QUITE enough "takers" yet for that though so I may hold off on it (making SATA I separated, from SATA II disks))

3.) I am also going to add in MORE than just "READ SPEEDS" but also CPU usage and SEEK/ACCESS time as well...

4.) Lastly, I will add a section with "avg. read" too!

APK

P.S.=> Let's make this one more comprehensive & more fun, so we draw more from it by way of comparison to aid us in future decision making on parts to buy, as well as have a "competition amongst one another" of course... apk

 

[_33]

Burst size is extremely high on his due to the ICHR7 being integrated into the southbridge controller and having full bandwidth available, and the fact that it is reading out of the onboard cache, Hitachi drives are very quick, and seem to make better use of the cache than other drives.

http://www.osnn.net/forum/showthread.php?t=81436

However, look at Average speed, and sustained output for the drives. Sure based on burst speed it kicks my systems ass, but how often is the next level preloaded into the drives cache?

Actually I wouldn't argue a score like what Ti66er is showing in that thread or this one here. It's just a high score that is achievable. But I tought yours is pretty low in comparison since you have a raid0 setup (but I understood *PCI* bus limited = 133MB/sec). So, what's the stripe size and cluster size?

 

[regan1985]

 

[Alec§taar]

I'll update (sorry for delay guys, busy here @ home & yes, in other posts on this forums)... if pt does not & he MAY not.

(Waiting to see & doing the breakout I mention above now, finally)



* Again - apologies for delay!

APK

 

[Alec§taar]

HD TACH 3 TEST ALL TAKERS BY CATEGORIES CHART (not broken out by disktypes)

REPOSTING TO LAST PAGE... apk

 

[Alec§taar]

Hd Tach 3 Test By Disktypes Used

REPOSTING TO LAST PAGE... apk

 

[_33]

Are you sure you read my benches properly dude? Maybe a little confused?

POST #42 dude

 

[Alec§taar]

Are you sure you read my benches properly dude? Maybe you want me to be in the last spots? What's your excuse?

LOL, do you know how much data there is, AND formatting to make it look nice?

TONS! Believe you me: It is NOT that simple...

Maybe a little confused?

Honestly, @ times on this one (& the others I did in AquaMark 3 & ScienceMark 2.0?):

YES! lol...

(Especially that latter one, broken out by disks!)

The 2nd set is FAR from done & ordered properly, first "ALL" group is near done (just doing avg.'s now).

(Feel free to take over man, or be patient!)

APK

 

[_33]

LOL, do you know how much data there is?

It is NOT that simple...

Especially that latter one, broken out by disks!

The 2nd set is FAR from done & ordered properly, first "ALL" group is near done (just doing avg.'s now).

(Feel free to take over man, or be patient!)

APK

Well, you decided to take this. I never even considered doing groupings of benches and making lists. Maybe I got upset at first. But being mistaken is a very bad thing. Let's say you're at the airport, prepared to go to Hawaï, and at your arrival, your luggages ended up in Los Angeles...

 

[Alec§taar]

Well, you decided to take this.

I was asked to by pt, did it as a favor. This is all...

(I figured PT did me, & ALL OF US, right by making a suggestion in the ScienceMark 2.0 test, to add CPU types & such. He did the data gather there)

Thus, I owed him one!

(As it made it a MUCH better test, & most viewed thread on these forums to date, plenty of participants, & GREAT overall test!)

You started the thread man, you ought to know - I did the work there though... wasn't ANY simpler than this one... believe me!

I never even considered doing groupings of benches and making lists. Maybe I got upset at first. But being mistaken is a very bad thing. Let's say you're at the airport, prepared to go to Hawaï, and at your arrival, your luggages ended up in Los Angeles...

TELL You what - just be patient.

I hate doing the second chart to be honest... but, it needs doing for the heck of it, so it is more "fair" to all concerned.

Please, just wait, nearly done!

APK

P.S.=> Do you guys NEED or WANT to keep the 2nd chart?

We can!

The hardest part of these? RIGHT NOW!

The INITIAL designing of the charts (thinking it out), formatting them nicely, & the "data-gather" really... trust me, it is a BEETCH!

BUT, after that it's pretty much cake (even the averages really, the groups here are TINY thusfar, 12 numbers to punch in is NOT bad, now 40 or so is (like in ScienceMark 2, whew))... apk

 

[_33]

I was asked to by pt, did it as a favor. This is all...

(I figured PT did me, & ALL OF US, right by making a suggestion in the ScienceMark 2.0 test, to add CPU types & such. He did the data gather there)

Thus, I owed him one!

(As it made it a MUCH better test, & most viewed thread on these forums to date, plenty of participants, & GREAT overall test!)

You started the thread man, you ought to know - I did the work there though... wasn't ANY simpler than this one... believe me!



TELL You what - just be patient.

I hate doing the second chart to be honest... but, it needs doing for the heck of it, so it is more "fair" to all concerned.

Please, just wait, nearly done!

APK

P.S.=> Do you guys NEED or WANT to keep the 2nd chart?

We can!

The hardest part of these? RIGHT NOW!

The INITIAL desing of the charts, formatting, & the "data-gather" really, it is a BEETCH!

BUT, after that it's pretty much cake (even the averages really, the groups here are TINY thusfar, 12 numbers to punch in is NOT bad, now 40 or so is (like in ScienceMark 2, whew))... apk

You don't have to do 2 charts, you can do 1 chart like it was before. I don't see the need for exagerating the significance of HDTACH since it's not even accurate. 1 accurate chart with the exact values is really enough IMHO, and you do have some fixing to do on my part already. If you want my hints, 1 overall chart woul dbe a start, and then charts segregated by system type, say IDE? SATA1, SATA2, SCSI. RAID or NO RAID, I think it makes no difference. A 15K HDD can squash a RAID0 setup. And a RAID1 isn't necessarely aimed at performance.

 

[Alec§taar]

You don't have to do 2 charts, you can do 1 chart like it was before.

IT IS HOW PT HAD IT BEFORE MAN - he even broke it out! See first page... I kept it as he had it with formatting & info mods.

I don't see the need for exagerating the significance of HDTACH since it's not even accurate.

Accurate enough!

We ALL run those same "inaccuracies"...

(Thus, test conditions are the same for us all I would think @ least).

1 accurate chart with the exact values is really enough IMHO, and you do have some fixing to do on my part already.

Long ago done!

I am now in the averages stages on second longer chart - only 2 groups away now from finish.

If you want my hints, 1 overall chart woul dbe a start, and then charts segregated by system type, say IDE? SATA1, SATA2, SCSI. RAID or NO RAID, I think it makes no difference.

I agree, 1 chart would be simplest, but I will wait out others's thoughts.

If we had more takers? I'd bust out SATA 1 vs. SATA 2 & suggested it earlier... this is when pt handed it over to me (man's on his vacation, & did me a fav. in the ScienceMark 2.0 test, I owed him one (I am "big" on that)).

A 15K HDD can squash a RAID0 setup. And a RAID1 isn't necessarely aimed at performance.

We don't have any 15k UltraScSi Cheetahs in here... but, it's not outta the question in SOME areas (seeks especially)...

APK

P.S.=> The hard part's done now, designing charts, tag formatting them, doing the data gather... rest is cake, even AVERAGES since the groups are puny! apk

 

[Alec§taar]

HDTach 3.0 Results/Scores/Standings ALL users by Seek, CPU, READS

REPOSTING TO ADD WOLVERINE SCORE... apk

 

[Alec§taar]

HDTach 3.0 Scores/Results/Standings by SEEK, CPU, & READS (by disktype used)

REPOSTING TO ADD WOLVERINE SCORE & CORRECT _33's PLACEMENT IN GROUP HIS SYSTEM IS IN... apk

 

[_33]

=========================================
RAID SATA & SATA2 CLUB (ACCESS/SEEK DATA)
=========================================

• 3. _33 = 13.5ms - (Seagate Barracuda 160gb sata II,set sata I)

**************************************************
AVG SATA/SATA 2 RAID 0 ACCESS-SEEK DATA (5 members): 12.68ms
**************************************************
=========================================
RAID SATA & SATA2 CLUB (CPU USAGE DATA)
=========================================

• 1. Alec§taar =0% - (2xWD "Raptor 'X'" RAID 0 (Promise Caching 128mb ECC RAM + onboard CPU Controller))

• 1. YiK =0% - (WD 80GB IDE)

• 2. YiK =2% - (Maxtor 80GB IDE)

• 2. regan1985 =2% - (Maxtor Duel SATA2)

• 3. tigger69 =3% - (2x hitachi 80gb sataII raid 0(ich7r))

• 4. steevo =4% - (2xMaxtor 300gb sataII raid 0 (raid0 silicon 3114 controller))

**************************************************
AVG SATA/SATA 2 RAID CPU USE DATA (6 members): 1.83%
**************************************************

=========================================
SATA & Sata2 CLUB (CPU USAGE DATA)
=========================================

• 1. _33 = 2% - (Seagate Barracuda 160gb sata II,set sata 1)

=========================================
SATA & Sata2 CLUB (READ DATA)
=========================================

• 1. _33 = 304.1mb/s - (Seagate Barracuda 160gb sata II,set as sata I)

=========================================
SATA & Sata2 CLUB (AVG READ DATA)
=========================================

• 1. _33 = 99.5mb/s - (Seagate Barracuda 160gb sata II,set sata I)

APK

P.S.=> DONE! apk

Those are not my config. I should appear in the RAID section.