thank you for taking the time to do all these benches with the different controllers NapalmV5!
Soon running 4x of these drives will approach the speeds of running things directly in ram![]()
I dont understand what you are saying... I thought you had 4 drives not 6 and all the 700+mb/s benchies are with 4 drives? I see 64gb there so 4 drives? 4x should indeed do 600, but not the 700 or 800 you are seeing. If thats 6x then sure; if its 4x then the benchmarks are false. All controllers send the same ATA read command to the drive and they can't mystically speed them up to be faster than the manufacturer's maximum specs.
your welcome..
great here i go again.. confusing the crap outta people
youre twisting that this could be 6x instead of 4x?.. unbelievable
this thread is all about 4x.. removed the 2x since it takes alot of time to do all this.. i wish i could edit the title i cant.. i didnt include any 6x in the thread.. all 4x = 64GB
4x false ? right cause im the one posting all this it must be false.. thanks alot!
manufacturer spec for the masterdrive px is 170/130 read/write
whose to say thats the limit? they probably were conservative or they upgraded their px line
p serial # performing @ 40MB/s read higher than r serial #
why is 1x @ 188MB/s read on the arc1210 which is only 256MB ??
again, how you explain that mr. One_Hertz ??
of course again you wont..
Last edited by NapalmV5; 12-10-2008 at 08:58 AM.
thanks for the update on new firmware for 5805... looks even more tempting right now![]()
---
sure youre welcome.. theres still alot room for improvement
hopefully the next bios improves read
Nice testing Napalm! Thx for sharing the results
A bit surprised at the Adaptec 5805's performance compared to the Areca1231
You are the one with the 6x comment that I quoted before (that you edited out now). That is the only place i took 6 drives from.
I really don't want to burst your bubble but I can bet any amount of money that you are not seeing 209MB/s sequential writes from each of your drives, like one of your benchies shows. 188mb/s reads is closer to reality, but still not real. If you know the first thing about SSDs you will realize how absurd it is to claim that your drives are able to write at faster speeds than to read. Unfortunately I do not think you do. The writes you are showing are a good deal faster than the Intel SLC drive. Manufacturers are not idiots and will not under rate their drives. Whatever controller you put on there will still be limited by how quickly the drive performs the operations sent to it; that part never changes.
Open your eyes. 600-650 seq reads and 500-550 seq writes is the best you can possibly see from 4 of those drives, regardless what you hook them up to.
thanks.. youre welcome
you really trying to get on my nerves arent you?
i havent edited out anything.. still there.. why are you making sheet up?
you got beef with me? for whatever reason? i certainly dont have any with you.. pm me and lets go over any beef you got with me.. or if you prefer.. this thread will do..
1x @ 1210
1x @ 1231 - ignore the cache spike whatever and thats more like 188MB/s since the marker is below the 200 line and not above it
4x @ 1231 - ignore the cache spike whatever and thats more like 750/just below 750MB/s
its best to look at the whole picture not just numbers.. etc..
Last edited by NapalmV5; 12-10-2008 at 12:04 PM.
No beef at all, just wanted to clarify for all the people watching that a controller doesnt add 100+mb/s to seq reads. And I have never mentioned that you edited anything or that you were lieing. HDTach is a worthless benchmark that simply doesnt work for raid configs. Apparently nobody told you that over the past few years that has been a very well known fact. Run iometer 100% seq reads with your array filled up to the max by the test file for about 15-20minutes to see another view...
To illustrate my point:
![]()
Last edited by One_Hertz; 12-10-2008 at 04:02 PM.
^uhhh what were you trying to say? I got hypnotized by your friggen wallpaper!!!
Like anyone's going to pay attention to your graphs when you post that LOL
what graphs![]()
LOL.... i can just barely see those graphs if i'm extremely focused![]()
---
People say that if you look really carefully, you will see a screenshot from IOmeter; I've been watching it for 10 mins and no screenshot![]()
i5 660 / Asrock P55M Pro / Ripjaws / GTS250
how can we take One_Hertz seriously with kind of display![]()
Yeah I am at a total loss here after seeing your numbers.
I just setup a EVGA X58 with 965, installed my Raid - ARC-1231ML and have 7x 32GB MTRON PRO's going.
I haven't run HDtach or HDtune yet as I really don't like those programs.
But I did run PCMark05 and PCMark Vantage on Vista Ultimate 32, fresh install.
Total stock settings with no over clock, this was my base line test, here I was I got. I have 2 SSD's on RMA which DVnation is taking it's sweet time with, so in the end it will be 9 SSD's.
PCMark05
HDD Score: 64,302
PCMark Vantage
HDD Test: 37,671
Everything is at defualt, including the ARC-1231ML for strip sizes etc. I haven't looked at firmware upgrades for some time, perhaps there is a new firmware.
Last edited by Buckeye; 12-10-2008 at 07:25 PM.
once again..
p serial px performing @ 40MB/s read higher than the r serial px
ive rmaed 2x Rxxxxxx (serial) px which were performing @ 148MB/s each
the controller didnt add any seq reads.. thats what 4x p serial px are performing at.. 750MB/s
hdtach = worthless.. fact? iometer is worse.. i cant believe even manufacturers use it..
ive used hdtach long enough and experienced performance long enough to be my preferred storage bench.. you trust iometer.. fine to each his own fetish..
the best way to bench storage is in pure mode.. filling up then benching.. thats like benching cpu @ full load
look.. we can argue till rapture whats the better benchmark or what way of doing it..
forget controllers/storage.. please post/upload your wallpaper up.. legs.. boobs.. daium!
- thread title revised.. thank you IFMU
- 1st post revised
- 5805/3520 benches revised/added
asr5805
1 aug firmware/bios/drivers
8 dec firmware/bios/drivers
![]()
Last edited by NapalmV5; 12-12-2008 at 08:49 AM.
Nice benches NapalmV5
I have a question though;
I've been using the Areca 1231ML for months now with just 2 Mtron 32GB SSDs on it, and its been fine as an OS disk since.
However I've recentry purchased a pair of velociraptors thinking I can put them onto the Areca, and use them as seconday drives.
Having now installed it and configured the Areca so that the raptors are in RAID0, I cannot find the array in windows, only the origional Mtron SSD array? Can the Areca only be configured to show one array in windows?
I'm thinking of buying the Adaptec 5805 so I can have all 4 drives sharing a hardware RAID controller, with 2 separate RAID0 arrays... is that possible on the 5805? Or it could be done on the Areca just that I suck at configuring it to show?![]()
thanks..
you gotta create the partition thru disk management.. have you tried that yet?
multi arrays is not an issue on areca controllers.. performance is and that goes for any controller
ive created and tested (real world apps/progs not benchmarks) multi array on areca controllers.. performance aint great
optimal multi array: one controller for each array
but since 1231 is an 800mhz iop maybe it handles multi arrays better but youre not gonna get full potential/performance out of any of the array
Last edited by NapalmV5; 12-14-2008 at 10:06 PM.
I have 2 SSDs and 2 raptors on my 2405 without affecting performance.
Can we get that wallpaper without the graphs? Sorry for being off-topic, but it was already requested anyways.![]()
For my part I know nothing with any certainty, but the sight of the stars makes me dream.
.
.
Bookmarks