IDE RAID – Is There A Benefit?

Pages: 1 2 3 4 5 6

Non RAID baseline

Just as a base line, I ran the benchmarks with first a single drive configured as a master on the first HPT port, then with 2 drives. Only the benchmark was run from the second drive, while the first had Window’s installed on it and was used for any swapping of data to the drives and from memory. I ran the test with the second drive configured as a slave to the first and with both drives as masters on their own ports.

Single drive

2 drives, master and master

2 drives, master and slave

HD Tach 2.60

Random Access (ms)

10.90

10.70

11.90

Read Burst Sped (mbps)

84.60

84.10

84.40

Read Speed (kbps)

Max

33103.00

33268.00

32274.00

Min

24319.00

22748.00

22784.00

Avg

25982.00

25892.00

25985.00

CPU Utilization (%)

6.00

7.30

6.40

Winbench 99

Bus Disk Winmark (kbps)

8260.00

7760.00

7820.00

HE Disk Winmark (kbps)

19300.00

19100.00

19300.00

Disk Transfer rate (kbps)

Beginning

33300.00

33300.00

33300.00

End

33400.00

33200.00

33200.00

Disk Access time (ms)

11.00

11.00

11.10

Disk CPU Utiliz. (%)

3.06

3.13

3.23

Bus Winstone 2001

24.70

25.10

25.30

CC Winstone 2001

24.40

24.30

24.20

No real surprises here, pretty much what I’d expect – in essence the performance is the same, one or two drives, master or slave. One note: Disk test under Windows are very difficult to run with consistency, sometimes I had to run the test a number of times to get a score that I felt was an accurate result. Note the slightly higher Winbench Business disk score that is not reflected anywhere else.

Looking at these results, I think it’s safe to say with any disk use profiled by my setup and the benchmarks used (and I think that covers most) there is no advantage to using more than one disk or using a disk as a master or a slave.


Pages: « Prev   1 2 3 4 5 6   Next »

Be the first to discuss this article!