@OldFart

Here are my results:

Sam460ex
Amiga OS 4.1.4
OCZ Vertex 3 on a Silicon image 3114 4 port Sata II PCI card


2.Amiga_OS4:> scsispeed DRIVE sii3114ide.device:0 FAST BUF1=8192 BUF2=16384 BUF3=32768 BUF4=65536
MKSoft ScsiSpeed 4.2 Copyright © 1989-92 MKSoft Development
AmigaOS4 port by Stéphane Guillard
------------------------------------------------------------
CPU: 68020 AmigaOS Version: 53.25
Device: sii3114ide.device:0


Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 61476864 bytes/sec

Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 83858227 bytes/sec

Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 102688358 bytes/sec

Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 114363596 bytes/sec

No too bad :)

My I TB seagate actually scored better:

Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 70921420 bytes/sec

Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 92840755 bytes/sec

Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 108606259 bytes/sec

Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 118335078 bytes/sec

But the SSD experience is faster overall due to the fact there is almost non existent access and seek times.

Drewlio77

Last edited by Drewlio77 on 27-Apr-2012 at 08:01 AM.

@Drewlio77

Very impressive figures, Sir! Very impressive. Your claim has sufficently been substantialised.

Thanks

OldFart

_________________
More then three levels of indigestion and you're scroomed!

@OldFart

Not that impressive. just 1mb/s higher than on my sam440ep

And the X1000 result was a bit disapointing.

_________________
I'm an antique. Don't light my fuse

@Antique

Quote:

Not that impressive.

I know, I know! Just trying to keep his spirit up. After 3 months of toying around and finally getting his act together he's now over the moon with the results.

Quote:

just 1mb/s higher than on my sam440ep

What is this, young boys peeing from the bridge to see who comes farthest? A d*ck measuring contest? Tsk, tsk. Hm, come to think of it: why do I have a µ?

Quote:

And the X1000 result was a bit disapointing.

Showing it is not as much a H/W issue as a S/W issue.

OldFart

_________________
More then three levels of indigestion and you're scroomed!

@Antique

"And the X1000 result was a bit disapointing."

To me it "only" indicates that x1000 runs the sata at sata1 speed for now.
When 150MB is the theorethical maximum for SATA1, 100MB values are already very good.

What is interesting is system behaviour during real life situations.
My Dualcore 2Ghz x86 can transfer 17MB/s from partition to another in linux and the system almost freezes during the transfer.
My SAM can do similar transfer 14MB/s while maintaining full system responsiveness.
It has been also said that x1000 has performed partition to partition copy at 70MB/s. That's 4x faster than my "good" x86 HW from y2008 or so. Being physically faster + give better user experience is nice combo.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@OldFart

Offcourse. Just like my father is bigger than yours,a dn my ferriari is faster than yours.

Just that i've seen much higher values on the x1000 myself.

_________________
I'm an antique. Don't light my fuse

Sam440ep-flex 800 mhz . Western Digital Caviar Blue


10.WorkBench:> Programs:Tools/Scsispeed4/scsispeed DRIVE sii3114ide.device:0 FAST BUF1=8192 BUF2=16384 BUF3=32768 BUF4=65536
MKSoft ScsiSpeed 4.2 Copyright © 1989-92 MKSoft Development
AmigaOS4 port by Stéphane Guillard
------------------------------------------------------------
CPU: 68020 AmigaOS Version: 53.25
Device: sii3114ide.device:0


Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 48327475 bytes/sec

Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 69065113 bytes/sec

Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 86830284 bytes/sec

Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 69301043 bytes/sec
10.WorkBench:>

Last edited by Spectre660 on 29-Apr-2012 at 08:41 AM.

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

@OldFart

X1000 update 5.
Samsung ssd disc. Don't remember exact model etc.


scsispeed DRIVE sb600sata.device:1 FAST BUF1=8192 BUF2=16384 BUF3=32768 BUF4=65536
MKSoft ScsiSpeed 4.2 Copyright © 1989-92 MKSoft Development
AmigaOS4 port by Stéphane Guillard
------------------------------------------------------------
CPU: 68020 AmigaOS Version: 53.26
Device: sb600sata.device:1


Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 41402368 bytes/sec

Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 93029990 bytes/sec

Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 151199744 bytes/sec

Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 177350246 bytes/sec

_________________
I'm an antique. Don't light my fuse

@Antique

Ok, now we are above SATA1 speed.
Are we nearing the fastest PPC desktop speed ever?? Or do we need RAID enabled?

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@Antique

SSD vs a sata hard drive, nice speedup.

_________________
Hate tends to make you look stupid...

@Antique

Now THAT's what I would really call impressive! Wow! Congratulations with these figures.

Post #1 dutifully updated.

(I'm sorry for the non-formatted look of the table: <pre> and </pre> don't seem to work anymore)

OldFart

_________________
More then three levels of indigestion and you're scroomed!

@Spectre660

Quote:

Sam440ep-flex 800ghz

Specially made for you? You bribed ACube? Did they sell it to you that way? Or is it an ordinary tipo? Should probably read Sam440ep-flex 800MHz, isn't it?

Anyway thanks for your contribution, which is taken to post #1.

OldFart

_________________
More then three levels of indigestion and you're scroomed!

@OldFart


indeed mhz . Corrected to avoid chaos .

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

Amiga model: X1000
OS-version: OS4.1.5
Disk brand, model and size: ADATA S510, 120Gb, 2.5 SATA 6Gb/s SSD (Solid State Drive)

I read somewhere that normal disk tests arent reliable for SSD's. But i have not much knowledge in SSD's. Anyway heres the stuff you wanted.

MKSoft ScsiSpeed 4.2 Copyright © 1989-92 MKSoft Development
AmigaOS4 port by Stéphane Guillard
------------------------------------------------------------
CPU: 68020 AmigaOS Version: 53.26
Device: sb600sata.device:0


Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 61863116 bytes/sec

Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 97891942 bytes/sec

Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 134358630 bytes/sec

Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 164367564 bytes/sec

Last edited by Kicko on 15-May-2012 at 08:45 PM.
Last edited by Kicko on 15-May-2012 at 08:44 PM.

Here is a test with a normal Harddisk (not SSD)

Amiga model: X1000
OS-version: OS4.1.5
Disk brand, model and size: WESTERN DIGITAL WD10EARS-00MVWB0 HARD DRIVE 1TB

MKSoft ScsiSpeed 4.2 Copyright © 1989-92 MKSoft Development
AmigaOS4 port by Stéphane Guillard
------------------------------------------------------------
CPU: 68020 AmigaOS Version: 53.26
Device: sb600sata.device:0


Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 64316211 bytes/sec

Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 99812147 bytes/sec

Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 119781785 bytes/sec

Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 120307712 bytes/sec

Sam440ep-flex 800mhz OS4.1 Update 5 .Seagate ST350041 500MB Harddrive .

scsispeed DRIVE sii3114ide.device:0 FAST BUF1=8192 BUF2=16384 BUF3=32768 BUF4=65536
MKSoft ScsiSpeed 4.2 Copyright © 1989-92 MKSoft Development
AmigaOS4 port by Stéphane Guillard
------------------------------------------------------------
CPU: 68020 AmigaOS Version: 53.34
Device: sii3114ide.device:0


Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 19458867 bytes/sec

Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 35511500 bytes/sec

Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 71678361 bytes/sec

Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 109687603 bytes/sec
------------------------------------------------------------------------------------------------------------------------------------------
Sam440ep-flex 800mhz OS4.1 Update 4 .Seagate ST350041 500MB Harddrive .

scsispeed DRIVE sii3114ide.device:0 FAST BUF1=8192 BUF2=16384 BUF3=32768 BUF4=65536
MKSoft ScsiSpeed 4.2 Copyright © 1989-92 MKSoft Development
AmigaOS4 port by Stéphane Guillard
------------------------------------------------------------
CPU: 68020 AmigaOS Version: 53.25
Device: sii3114ide.device:0


Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 73239756 bytes/sec

Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 94144921 bytes/sec

Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 107713331 bytes/sec

Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer.
Read from SCSI: 117070233 bytes/sec

Quote:

Spectre660 wrote: Sam440ep-flex 800 mhz . Western Digital Caviar Blue 10.WorkBench:> Programs:Tools/Scsispeed4/scsispeed DRIVE sii3114ide.device:0 FAST BUF1=8192 BUF2=16384 BUF3=32768 BUF4=65536 MKSoft ScsiSpeed 4.2 Copyright © 1989-92 MKSoft Development AmigaOS4 port by Stéphane Guillard ------------------------------------------------------------ CPU: 68020 AmigaOS Version: 53.25 Device: sii3114ide.device:0 Testing with a 8192 byte, MEMF_FAST, LONG-aligned buffer. Read from SCSI: 48327475 bytes/sec Testing with a 16384 byte, MEMF_FAST, LONG-aligned buffer. Read from SCSI: 69065113 bytes/sec Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer. Read from SCSI: 86830284 bytes/sec Testing with a 65536 byte, MEMF_FAST, LONG-aligned buffer. Read from SCSI: 69301043 bytes/sec

Last edited by Spectre660 on 25-Aug-2012 at 03:31 PM.

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

@Spectre660

Rather strange that your system is performing far worse with the latest update applied! On all fronts its lower. While the first figures seem about to double the speed, from 19 to 36 and from 36 to 72, the rather high speed of 117 in highest mode with update 4 you do not reach by far!

OldFart

_________________
More then three levels of indigestion and you're scroomed!

@KimmoK

Quote:

My Dualcore 2Ghz x86 can transfer 17MB/s from partition to another in linux and the system almost freezes during the transfer. My SAM can do similar transfer 14MB/s while maintaining full system responsiveness. It has been also said that x1000 has performed partition to partition copy at 70MB/s. That's 4x faster than my "good" x86 HW from y2008 or so.

It only means that either is your "good" x86 HW broken or badly configured or both.

For comparison, my ancient desktop at work achieves about 50-60MB/s when copying files within the same partition or between different partitions (the limit is most likely the harddrive itself), about 30MB/s when copying files between harddrive and remote server through gigabit ethernet. Another machine does harddrive transfers at up to 110-115MB/s. Another 4 years old linux server does perform at about 100-110MB/s either.

Ah yes, the system is fully responsive.

@michalsc

Yes, it could be that default configurations in linux sucks.
Will try again after I update to latest Mint xfce.

(will be buying extra 500G hard drive just because the partition to partition copy is so slow currently, I imagine two Seagate SATA Barracude 7200rpm should move files 100MB/sec between each other)

Last edited by KimmoK on 27-Aug-2012 at 11:27 AM.
Last edited by KimmoK on 27-Aug-2012 at 11:25 AM.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@michalsc

On Linux hdparm gives about 120MB/s on your typical cheapo dell using ordinary cheapo sata disks. Even on my 5 year old low power via proc board I get about 30MB/s with a ten(!) year old IBM disk. (man I need to retire that disk)

While hdparms skips the file system layer it is a pretty good indicator of disk speed and in my experience block file i/o mirrors it closely.

All that said, the Amiga numbers for the sam460, which i am hoping to buy, look respectable and are probably pretty close to the hardware limits.

Lawrence

_________________
Blog