1 |
On Sat, 28 Jun 2008 23:19:26 +0200 |
2 |
Volker Armin Hemmann <volker.armin.hemmann@××××××××××××.de> wrote: |
3 |
|
4 |
> On Samstag, 28. Juni 2008, Florian Philipp wrote: |
5 |
> > Hi list! |
6 |
> > |
7 |
> > I've got a little question. Here I've got a PC with two 40GB hard |
8 |
> > disks, a CD-burner and a DVD-burner, all PATA. Now I'm running out |
9 |
> > of disk space and I'm thinking about ways to increase it: |
10 |
> > |
11 |
> > 1. I could replace the older 40GB disk with a new model but I'd |
12 |
> > hate to trash it because it's working alright. |
13 |
> > |
14 |
> > 2. I could replace the really old and rather useless CD-burner with |
15 |
> > a disk. However, that would mean I'd have to attach the DVD-burner |
16 |
> > and one of the disks to the same controller. Now I'm asking myself |
17 |
> > (and you ;) ) how that would affect performance. |
18 |
> |
19 |
> wikipedia can answer that. |
20 |
|
21 |
"'One operation at a time' |
22 |
|
23 |
This is caused by the omission of both overlapped and queued feature |
24 |
sets from most parallel ATA products. Only one device on a cable can |
25 |
perform a read or write operation at one time, therefore a fast device |
26 |
on the same cable as a slow device under heavy use will find it has to |
27 |
wait for the slow device to complete its task first. |
28 |
|
29 |
However, most modern devices will report write operations as complete |
30 |
once the data is stored in its onboard cache memory, before the data is |
31 |
written to the (slow) magnetic storage. This allows commands to be sent |
32 |
to the other device on the cable, reducing the impact of the "one |
33 |
operation at a time" limit. |
34 |
|
35 |
The impact of this on a system's performance depends on the |
36 |
application. For example, when copying data from an optical drive to a |
37 |
hard drive (such as during software installation), this effect probably |
38 |
doesn't matter: Such jobs are necessarily limited by the speed of the |
39 |
optical drive no matter where it is. But if the hard drive in question |
40 |
is also expected to provide good throughput for other tasks at the same |
41 |
time, it probably should not be on the same cable as the optical drive." |
42 |
|
43 |
So I can hope that caching will negate any negative effects as long as |
44 |
one of the drives is much faster than the other (e.g. optical and hard |
45 |
disk drive) even when I'm streaming a (theoretically) constant byte |
46 |
stream from one to the other (ripping, burning)? |
47 |
|
48 |
> |
49 |
> But why not a cheap sata controler and a nice big sata disk? They are |
50 |
> cheaper now than PATA disks. |
51 |
|
52 |
cheap SATA-controller: EUR 35 |
53 |
cheap 160GB SATA-disk: EUR 35 |
54 |
cheap 160GB PATA-disk: EUR 40 |
55 |
cheap 250GB PATA-disk: EUR 45 |
56 |
|
57 |
Not cheap enough it seems ;) |
58 |
It would be more price efficient to trash one of the current disks for |
59 |
the new one than buying a new controller. |
60 |
|
61 |
By the way: Actually the mobo has two SATA-controllers but one of them |
62 |
crashes on POST while the other one doubles the time the POST takes. |
63 |
That's why they are both deactivated. |