MaximumPC 2007 04

28 MAXIMUMPC APRIL 2007 ➭➭

B

uying a new hard drive doesn’t have to

be an outrageously expensive proposi-

tion. In fact, we’re not going to let it be.

We’ve rounded up drives from four top manufac-

tures—with one special catch: Each drive has an

MSRP of $200 or less. Yes, it’s the battle of the

cheapies, but just because they’re inexpensive

doesn’t mean they aren’t solid hunks of storage.

The grand melee consists of Hitachi’s

500GB Deskstar E7K500 ($200, http://www.hitachi.

com ), Samsung’s 500GB SpinPoint T166 ($190,

http://www.samsung.com ), Seagate’s 400GB 7200.

Barracuda ($190, http://www.seagate.com ), and

Western Digital’s 500GB Caviar SE16 ($200, http://www.

westerndigital.com ).

Emerging from the brawl as the speed victor is

Western Digital’s Caviar drive. Not only did the drive

perform admirably on our HD Tach random access test,

but more importantly, it had the highest average read

speed of all four contenders. And that isn’t just because

it was spinning faster; the platters on each drive in our

showdown run at 7,200rpm.

As expected, the PATA

drive was the worst performer

in all categories. Come on,

Seagate—where’s the cheap,

high-capacity SATA?

F

irst, thou shall take out the external hard drive.

Then thou shall connect using three methods:

FireWire 800, USB 2.0, and eSATA. But select-

ing the speed victor among these three options isn’t

as obvious as picking out a black sheep in a white

herd—USB loses hands down, but FireWire 800 puts

up a worthy fi ght against eSATA.

We tested all three interfaces using the fastest

SATA drive in the Lab, the Western Digital Raptor X

hard drive. This helped ensure that we were testing

the interface, not the drive. We hooked the Raptor up to

a plain ol’ external enclosure that supported both USB

and eSATA interfaces and proceeded to run HD Tach on

both interfaces.

FireWire 800 is a bit more elusive on external

enclosures, so we connected a WiebeTech SATADock

to a PCI-based FireWire 800 card in our test bed and

attached the Raptor to the SATADock via SATA. Ta da!

Watching the USB benchmarks come through

brought a sad tear to our eye. If we’re talking purely

in terms of speed, USB’s usefulness as an effi cient

transfer mechanism is dead—dead, rotting, and rolling

in its grave at the thought of its faster brethren, eSATA

and FireWire 800.

As for those two, well, we fi nd ourselves at a bit

of a loss. As noted, our HD Tach benchmark showed

virtually no difference between the Raptor’s average

read speeds via a FireWire 800 or eSATA connection.

CPU usage is nearly identical, with the only difference

between the two mediums coming from burst speed,

the highest rate data will zip to the drive.

With that said, eSATA ekes out the win. Not only

is its 300MB/s cap larger than the 100MB/s cap of

FireWire 800, but it also ever-so-slightly squeaks out

faster read speeds.

THE FASTEST CONNECTION

POWERLINE

NETWORKING VS.

802.11N

NETWORKING BATTLE

HITACHI VS. SAMSUNG VS. SEAGATE VS. WD

$200 HARD-DRIVE SHOOT-OUT

BENCHMARKS HITACHI SAMSUNG SEAGATE WD

HD TACH RANDOM ACCESS (MS) 12.9 14.1 15.2 13.

HD TACH AVERAGE READ (MB/s) 64.2 70.3 61.3 72.

HD TACH BURST SPEED (MB/s) 206.4 219.9 94.0 204.

Best scores are bolded.

W

e’ve wanted to stream high-def

video from room to room for

years, but for a long time, the only

way to make that happen was to run

CAT-S Ethernet between the rooms.

Fortunately, there are alternatives to

cabling: powerline networking and the

latest Wi-Fi standard, 802.11n.

We tested using two Netgear

products, the HDX101 Powerline

Ethernet Adapter ($130 each, http://www.

netgear.com ) and the WNR854T draft-

802.11n Wi-Fi router ($160). First, we

ran a basic fi le transfer test in which

we copied a 238MB fi le from a wired

machine to a wireless machine. Next,

we tested the hardware’s ability to

stream high-defi nition video across

the interface.

Our results were interesting, to

say the least. With the HDX101, which

uses a different (and incompatible) chip-

set than previous powerline products,

we transferred our test fi le in about 87

seconds. HD streaming worked fl aw-

lessly. Wi-Fi fared poorly in our real-

world test environment, although in the

Lab it worked reasonably well. It took

about 207 seconds to transfer our test

fi le over the 802.11n connection in the

real-world environment, and we weren’t

able to stream HD video. In the Lab, with

a direct line-of-sight between the router

and the laptop, we managed to transfer

the test fi le in 36 seconds. With one wall

between the router and the laptop, that

time dropped to 65 seconds.

Granted, we’re still working with

draft-802.11n hardware, and we expect

the fi nalized spec to perform better. But,

if you want to stream HD video across

your home today, your best choice is a

powerline product using DS2’s chipset.

BENCHMARKS ESATA FIREWIRE 800 USB 2.

CPU UTILIZATION 1% 2% 0%

RANDOM ACCESS (MS) 8.3 8.1 8.

AVERAGE READ (MB/s) 77.9 76.0 36.

BURST SPEED (MB/s) 128.8 87.9 37.

Best scores are bolded.

FIREWIRE 800 VS. USB

VS. ESATA