Comments interspersed below -
Bruce Johnson wrote:
On Dec 9, 2011, at 8:37 AM, Bruce Godfrey wrote:
Disk Speed Bench X reported the following this morning:
ATA 100 drive mounted as /Volumes/MDD HD1 77 MB/sec
SCSI drive mounted as / 76MB/sec No difference between IDE and SCSI
speeds according to this test.
This means your disk I/O is not what's constraining your system. A
faster disk did not increase your actual trabsfer rates, so it's a bus
limit, not a disk IO limit.
Do you know something about the differences in the way Xbench and Disk
Speed Bench X actually run their tests? I can believe that the bus
might be limiting transfers to about 75MB/sec. Given that, a disk
that can actually put data down that fast will make for a better
performing system than one that puts down a lot less than that.
I see in the comparison (DSB to XB) that DSB shows both disks running at
about 76MB/sec. XB gets about this same number for the SCSI disk in the
uncached write and uncached read tests, but gets less for the IDE disk
in those tests. So there is some consistency in the measurements
there. And what look like real differences.
If you look over the comprehensive drive tests and reviews, for instance
on Storagereview.com, you find lots of cases where the specs and the
real world performance do not match up exactly. Sometimes the biggest
differences come from the tests which are designed to be most like
running desktop applications.
Xbench results when the MDD was booted from the faster ATA100 drive:
Memory Test 37.05 System 34.46 Allocate
94.16 345.79 Kalloc/sec
Xbench results when the MDD was booted from the SCSI drive running on
a UL3D card:
Memory Test 40.85 System 40.53 Allocate
111.12 408.06 Kalloc/sec
What this mainly means is that XBench is significantly non-precise,
that it is designed for systems much faster than the MDD you tested it
on, or that your system is seriously memory-bound. Disk speed should
have no bearing whatsoever on memory or system test values, as these
tests shouldn't be touching the disks at all.
I don't know what to make of the memory test results either. If you are
writing and reading data into and out of RAM for the test, where does
that data come from?
The system has the full 2GB RAM that it can hold so it shouldn't be
"memory bound".
Bruce
--
Bruce Johnson
University of Arizona
College of Pharmacy
Information Technology Group
Institutions do not have opinions, merely customs
--
You received this message because you are a member of G-Group, a group
for those using G3, G4, and G5 desktop Macs - with a particular focus
on Power Macs.
The list FAQ is at http://lowendmac.com/lists/g-list.shtml and our
netiquette guide is at http://www.lowendmac.com/lists/netiquette.shtml
To post to this group, send email to [email protected]
For more options, visit this group at
http://groups.google.com/group/g3-5-list
--
You received this message because you are a member of G-Group, a group for
those using G3, G4, and G5 desktop Macs - with a particular focus on Power Macs.
The list FAQ is at http://lowendmac.com/lists/g-list.shtml and our netiquette
guide is at http://www.lowendmac.com/lists/netiquette.shtml
To post to this group, send email to [email protected]
For more options, visit this group at http://groups.google.com/group/g3-5-list
