[jclarkv]

I found a good link over on the Apple G5 Support forum I wanted to pass along. Storage Review is the web site and they have an article entitled: Setting a New Standard - A Revised Look at Serial ATA Performance at http://www.storagereview.com/article...ANewStd_1.html

Wester Digital seemed to be the consistent winner of the SATA drives tested so far and my newly purchased and installed Seagate Barracuda the worst (although it is the quietest).

[TC]

I was wondering if anyone else could comment on whether the G5 hard disk is any faster than what was used in the G4s?
I know the interface should be faster but isn't the drive the bottleneck at the moment?

I've seen a couple of people pass comments that the G5 will also benefit in performance test due to its faster drive but from xbench scores it doesn't seem to be true, or does xbench need to be rewritten to get the best out of SATA drives?

Anyway, here is the xbench test I was looking at which compared a dual G4 867 to a G5 1.8:
http://ladd.dyndns.org/xbench/merge....amp;doc2=30728
( Comes from this thread: http://forums.macnn.com/showthread.p...hreadid=174750 )

[Tom N]

The Western Digital that was tested is a 10,000 rpm 36 GB drive, while the others were 7200 rpm drives of much larger storage size. Hardly an apples for orange test of the drives.

SATA as currently implemented can only provide a marginal increase in performance over a single ATA-133 interfaced drive. But performance looks much better when comparing two SATA drives against two ATA-133 interfaced drives. The SATA drives have none of the contention issues that the ATA-133 drives have do to the interface being used.

SATA-2 should provide a potential speed boost, assuming hard drive manufacturers will produce higher performance drives. The question I think becomes to the drive manufacturers is where SATA and SATA-2 drive interfaces fit in against SAS (Serial attached SCSI).

I think SATA and SATA-2 will be large drives operating at lower performance levels geared towards desktop users, while SAS drives which can use the same SATA physical interface, will be marketed to those needing higher performance and/or reliability, such as in the server, and large storage markets.

Tom N.

[Arkham_c]

Originally posted by TC:
I was wondering if anyone else could comment on whether the G5 hard disk is any faster than what was used in the G4s?
I know the interface should be faster but isn't the drive the bottleneck at the moment?

Access time will essentially be the same for all drives. However, for sustained reads and writes (large files, for example) the SATA interface should allow more throughput. I don't think any of the benchmark tools do a "Copy a 500MB file from A to B", which is yet another reason that benchmarks don't mimic real-world performance.