S
smith
Is the HD with SATAII really much faster than the IDE HD?
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
P
Paul
smith said:Is the HD with SATAII really much faster than the IDE HD?
The media (spinning platters and head assembly) control the
maximum rate of data transfer for sustained transfers. My
last drive purchased, can sustain about 70MB/sec for large
data transfers. UDMA5 on a ribbon cable is fast enough
to handle that rate, as is SATA 150MB/sec and SATA 300MB/sec.
For short data transfers, those go over the cable and are
temporarily stored in the disk drive cache memory chip.
The short data transfer can make better use of the available
bandwidth on the cable. In that case, a user might see the
data move over the 300MB/sec SATA II cable, at about 220MB/sec
or so (depends on chipset).
So, for a small amount of data, it moves at 220MB/sec, but
once the cache is full and the data is queuing up to go to
the platters, then the observed transfer rate is 70MB/sec.
If you transfer a 1GB file, the tiny 2MB or 8MB cache
means that the burst transfer characteristic is not sustained
for very long, and the majority of the transfer is at
70MB/sec.
And that is why I would claim, that it really doesn't
matter which interface type you use. When I really need
fast transfers, I am usually transferring 1GB or bigger
files, and for all of those, the interface type doesn't
matter.
Paul
B
Bob Harris
I have two SATA/150 (version I of SATA) internal hard drives. All file
transfers between these disks are faster (at least 50%) than between either
of them and an external USB2 or firewire disk or between extrnal and
external, even though all spin at 7200 rpm. So, there is more to copy speed
than just the rpm.
However, I do agree that an external USB drive that uses a 5400 rpm disk is
slower than one that uses a 7200 rpm disk. So, if everything else were
equal, rpm could matter.
Some years ago I had a Pentium II 450 Mhz with 100 Meg buss and two ATA/33
drives and controllers. The copy speed doubled when I added a PCI card with
an set of ATA/100 controllers and two new ATA/100 disks. I do not know the
speed of the old disk, but the new ones were 7200 rpm.
My current PC is a Pentium 4 2.5GHz with 400 Meg buss with SATA/150 drives
and controllers (on motherboard). The copy speed is about triple that of
the Pentium II machine with ATA/100 drives and controllers. Some of that is
SATA/150 vs ATA/100, but some is likely the buss and CPU speed differences.
From these experiences I surmise that copy speed is governed by the more
limiting of (1) rotational speed, (2) I/O speed (e.g., ATA vs SATA vs USB2),
(3) buss speed, (4) CPU speed. Additionally, the copy speed for small files
is probably impacted by the cache size in the hard drive; bigger chache is
better.
I have also heard that on current PCs SATA-II (SATA/300) is not
significantly faster than SATA-I (SATA/150). This is likely because the I/O
interface is not limiting. However, as other aspects of a PC get faster,
that may no longer be true.
transfers between these disks are faster (at least 50%) than between either
of them and an external USB2 or firewire disk or between extrnal and
external, even though all spin at 7200 rpm. So, there is more to copy speed
than just the rpm.
However, I do agree that an external USB drive that uses a 5400 rpm disk is
slower than one that uses a 7200 rpm disk. So, if everything else were
equal, rpm could matter.
Some years ago I had a Pentium II 450 Mhz with 100 Meg buss and two ATA/33
drives and controllers. The copy speed doubled when I added a PCI card with
an set of ATA/100 controllers and two new ATA/100 disks. I do not know the
speed of the old disk, but the new ones were 7200 rpm.
My current PC is a Pentium 4 2.5GHz with 400 Meg buss with SATA/150 drives
and controllers (on motherboard). The copy speed is about triple that of
the Pentium II machine with ATA/100 drives and controllers. Some of that is
SATA/150 vs ATA/100, but some is likely the buss and CPU speed differences.
From these experiences I surmise that copy speed is governed by the more
limiting of (1) rotational speed, (2) I/O speed (e.g., ATA vs SATA vs USB2),
(3) buss speed, (4) CPU speed. Additionally, the copy speed for small files
is probably impacted by the cache size in the hard drive; bigger chache is
better.
I have also heard that on current PCs SATA-II (SATA/300) is not
significantly faster than SATA-I (SATA/150). This is likely because the I/O
interface is not limiting. However, as other aspects of a PC get faster,
that may no longer be true.
V
V Green
Bob Harris said:I have two SATA/150 (version I of SATA) internal hard drives. All file
transfers between these disks are faster (at least 50%) than between either
of them and an external USB2 or firewire disk or between extrnal and
external, even though all spin at 7200 rpm. So, there is more to copy speed
than just the rpm.
However, I do agree that an external USB drive that uses a 5400 rpm disk is
slower than one that uses a 7200 rpm disk. So, if everything else were
equal, rpm could matter.
Some years ago I had a Pentium II 450 Mhz with 100 Meg buss and two ATA/33
drives and controllers. The copy speed doubled when I added a PCI card with
an set of ATA/100 controllers and two new ATA/100 disks. I do not know the
speed of the old disk, but the new ones were 7200 rpm.
My current PC is a Pentium 4 2.5GHz with 400 Meg buss with SATA/150 drives
and controllers (on motherboard). The copy speed is about triple that of
the Pentium II machine with ATA/100 drives and controllers. Some of that is
SATA/150 vs ATA/100, but some is likely the buss and CPU speed differences.
From these experiences I surmise that copy speed is governed by the more
limiting of (1) rotational speed, (2) I/O speed (e.g., ATA vs SATA vs USB2),
(3) buss speed, (4) CPU speed. Additionally, the copy speed for small files
is probably impacted by the cache size in the hard drive; bigger chache is
better.
I have also heard that on current PCs SATA-II (SATA/300) is not
significantly faster than SATA-I (SATA/150). This is likely because the I/O
interface is not limiting. However, as other aspects of a PC get faster,
that may no longer be true.
All very nice, but would you/do you notice a difference when
just doing regular stuff (net browsing, etc.) or is it just noticable
when copying large files (not something anybody does all the time)?
Let's offer our opinions based on REAL-WORLD usage, please.
Stats mean nothing when used outside the context of the device's
usage pattern.
J
JS
I know that benchmarks are not the be all end all, but that said see Tom's
Hardware for some comparison data.
http://www23.tomshardware.com/storage.html?modelx=33&model1=367&model2=137&chart=41
JS
Hardware for some comparison data.
http://www23.tomshardware.com/storage.html?modelx=33&model1=367&model2=137&chart=41
JS
D
Derek James
All very nice, but would you/do you notice a difference when
just doing regular stuff (net browsing, etc.) or is it just noticable
when copying large files (not something anybody does all the time)?
Let's offer our opinions based on REAL-WORLD usage, please.
Stats mean nothing when used outside the context of the device's
usage pattern.
No, in the real world you wont notice any appreciable difference between
SATA and IDE. But if your head is sruled by benchmarks, then get a SATA
drive.
A
Andrew E.
Actually SATA 2 isnt where one would/could ever see much of a diffrence
in a OS,the diffrences useing SATA2 opposed to SATA1 or even IDE hd is in
the controller & drive(s) configured.A good example is illustrated at
intel,chk:
http://www.intel.com/performance/desktop/platform_technologies/storage_performance.htm
As you can clearly see 1 hd really makes no diffrence,the additional ones
do.....
in a OS,the diffrences useing SATA2 opposed to SATA1 or even IDE hd is in
the controller & drive(s) configured.A good example is illustrated at
intel,chk:
http://www.intel.com/performance/desktop/platform_technologies/storage_performance.htm
As you can clearly see 1 hd really makes no diffrence,the additional ones
do.....
R
raylopez99
I have SATAII on a Core 2 Duo 6550 machine and I can say it's QUIETER
than anything I've ever run with a Pentium IV, certainly--ou cannot
hear it spin, or even know it's running, it's that quiet.
Unfortunately, because Vista (the OS I'm using) is not yet optimized
for speed like XP is, the copy process itself is slower at the user
level (behind the scenes it's faster it seems, since the copy process
jumps in spurts, fast then slow--this is not a application problem but
a problem with Vista, as has been noted elsewhere and should be fixed
in the next service pack says Microsoft).
RL
than anything I've ever run with a Pentium IV, certainly--ou cannot
hear it spin, or even know it's running, it's that quiet.
Unfortunately, because Vista (the OS I'm using) is not yet optimized
for speed like XP is, the copy process itself is slower at the user
level (behind the scenes it's faster it seems, since the copy process
jumps in spurts, fast then slow--this is not a application problem but
a problem with Vista, as has been noted elsewhere and should be fixed
in the next service pack says Microsoft).
RL
L
Lil' Dave
smith said:Is the HD with SATAII really much faster than the IDE HD?
All that matters is throughput in the end game of real computer use. 2 ide
drives (even the last ones made) are slower than 2 SATA II drives copying
data. Reason: there's no concurrent allowance with the ide bus. Only one
can operate at a time.
Singularly, there's no apparent difference.
Dave