I seem to have some kind of hard drive fetish ;) and thus had access to three different brands of 1 TB drives. I always bench new drives and thought I could share the results...

Tested drives

• Hitachi Deskstar 7K1000: 1000 GB (931 GB), 32 MB cache, 7,200 RPM
• Samsung F1: 1000 GB (931 GB), 32 MB cache, 7,200 RPM
• Western Digital Caviar GP: 1000 GB (931 GB), 16 MB cache, 5,400 RPM

The Hitachi drive uses 5 platters with 200 GB each. The Samsung achieves the same total capacity with only 3 platters holding 333.3 GB each (hence the high transfer rates). The Western Digital drive uses 4 platters with 250 GB each. Also of note, the WD drive rotates at only 5,400 RPM according to measurements (contrary to the marketing term of "variable 5,400 to 7,200 RPM"). (Update: OWC now has a new model/higher performance 3 platter WD 1TB GP w/32MB cache - tests here used the earlier 4-platter/16MB cache model.)

I originally bought the Hitachi because of their good Single-User-Mark results at StorageReview. Unfortunately there's no benchmark with real-world access patters on the Mac AFAIK. The synthetic benchmarks I did mostly test transfer rate, and it's only natural that the Hitachi loses to the newer denser drives (or is about equal to the WD drive with higher density but lower RPM). Please have a look at this comparison of 1TB hard drives at StorageReview, especially the results of the Single-User Suite 4.0. Also of note is the exceptionally low power dissipation of the WD drive compared to the Hitachi drive, making the WD drive perfect when low power/heat/noise is more important than raw speed (for example in a RAID configuration).
The Hitachi drive is fast, but also quite hot - this is not a problem in a Mac Pro, but might be in tight spots with insufficient cooling, like some external cases.

Personally, I'd go with a Samsung F1, since it is wicked fast. Another plus is its reduced heat and power when compared to the Hitachi drive. The Western Digital drive is even quieter and uses less power than the Samsung drive, which makes it a perfect fit for external cases without active cooling.
(FYI: - check the site specials page for current 1TB Hard Drive pricing/offers including the Samsung F1, improved (32MB cache/3 platter) WD 1TB GP as well as Hitachi and Seagate models.)

Host

The drives were tested in a Mac Pro (2006) with 4 GB RAM in Mac OS X 10.4.x (all Hitachi tests except QuickBench) or Mac OS X 10.5.x (all Samsung and Western Digital tests as well as Hitachi QuickBench). All drives blank/empty.

All results are in MB/s unless specified differently.

QuickBench Results

QuickBench: Sequential Reads

QuickBench: Sequential Writes

QuickBench: Random Reads

QuickBench: Random Writes

QuickBench: Large Reads

QuickBench: Large Writes

ZoneBench Results

ZoneBench: Hitachi 7K1000

ZoneBench: Samsung F1

ZoneBench: Western Digital GP

ZoneBench Random Access: 10,000 Reads and Writes of 4 KB files

Unfortunately, I don't have a result for the Western Digital GP for this test. Due to the higher rotational latency it would probably have had a lower score than the others. Instead, I included the first 25 GiB partition of WD Raptor 150 GB with 10,000 RPM, which did much better thanks to the higher RPM (which results in lower rotational latency) and small partition.

Misc. transfer tests

In the following are results of a bunch of misc. tests. I added results of a previously tested WD Raptor 150 (10,000 RPM, 16 MB cache) drive for comparison. The tests:

• Duplicate of large file on drive: read or write as (file-size/duplicate-time), number should be duplicated by two to get combined read and write throughput.
• KONA System Test with 4 GB file, largest frame size, file system cache disabled.
• Helios LAN Test, average of 3 runs with 300 MB files.
• XBench: disk test.

XBench

(.zip'd files)
• Hitachi 7K1000 XBench result file
• Samsung F1 XBench result file
• Western Digital GP XBench result file