Jump to content


7200.3 vs. 3200BEKT vs. 7K320 - Tom's hardware benchmarks


  • You cannot start a new topic
  • Please log in to reply
7 replies to this topic

#1 Guest_Vantage72_*

Guest_Vantage72_*
  • Guests

Posted 31 August 2008 - 01:55 AM

I'd like to know which of these drives is fastest in REAL WORLD use.

On NBR there's a guy who tested the 7K320 and WD 3200BEKT on the same notebook with PC Mark 05 and PC Mark Vantage. 7K320 is the clear winner in 4 of the 5 HDD tests. In Vista and XP.

But according to Tom's hardware I/O Meter 2003 tests the WD3200BEKT is faster than 7K320.
http://www.tomshardw...r...7,2039,2037
http://www.tomshardw...r...7,2039,2037
http://www.tomshardw...r...7,2039,2037

So my question is, which benchmark is the better predictor of real world performance: I/O Meter 2003 or PC Mark/ Vantage?

Edited by Vantage72, 31 August 2008 - 01:57 AM.

#2 meteordnf1

meteordnf1

    Member

  • Member
  • 39 posts

Posted 31 August 2008 - 03:43 AM

I'd like to know which of these drives is fastest in REAL WORLD use.

On NBR there's a guy who tested the 7K320 and WD 3200BEKT on the same notebook with PC Mark 05 and PC Mark Vantage. 7K320 is the clear winner in 4 of the 5 HDD tests. In Vista and XP.

But according to Tom's hardware I/O Meter 2003 tests the WD3200BEKT is faster than 7K320.
http://www.tomshardw...r...7,2039,2037
http://www.tomshardw...r...7,2039,2037
http://www.tomshardw...r...7,2039,2037

So my question is, which benchmark is the better predictor of real world performance: I/O Meter 2003 or PC Mark/ Vantage?


A tough question. I have the same problem myself as I'm looking for a 320gb replacement of my fast but too small 7k200 (100gb). Seems all 320gb drives have a slight problem with access times probably because of the higher density. None can match the 14.x ms (including latency) of the old 7k200. They all end up between 15.5ms and 17ms. These are values of 5400 3.5" desktop drives. Especially the 7200.3 is disappointing here since it has the highest STR and Seagate usually has very good access times. So benchmark results will vary based on weight of seek times / STR / overall performance. I think the WD is the best allrounder, having the best seek times, a pretty good STR and a very good firmware making good use of hdd cache. Of course the WD is the most expensive one around here :(

But all drives are very close so you probably won't feel a difference during normal day use. I'm tempted to wait a few more months for the 7200.4 available with up to 500gb.

#3 shoarthing

shoarthing

    Member

  • Member
  • 1 posts

Posted 31 August 2008 - 06:58 AM

. . . A tough question. I have the same problem myself as I'm looking for a 320gb replacement of my fast but too small 7k200 (100gb). Seems all 320gb drives have a slight problem with access times probably because of the higher density. None can match the 14.x ms (including latency) of the old 7k200. They all end up between 15.5ms and 17ms. These are values of 5400 3.5" desktop drives. Especially the 7200.3 is disappointing here since it has the highest STR and Seagate usually has very good access times . . . But all drives are very close so you probably won't feel a difference during normal day use. I'm tempted to wait a few more months for the 7200.4 available with up to 500gb.


If your 7K200 is the 100Gb-per-platter jobbie, these measures [from a respectable-enough OSX benching utility that'll be unfamiliar to you, with its own 'Nix & OSX quirks] may be of interest.

Worth noting:

1) each HDD attached via the same ICH7R channel to the same box
2) each HDD newly partitioned/formatted to single/GUID/HFS+ [a journalled scheme]
3) 7K200 has Apple firmware [as can pretty obviously be seen]

7K200: Posted Image

7200.3: Posted Image

. . FWIW [Macbook Pros have pretty good monitoring h/w & s/w] over the average working day the 7200.3 is if anything cooler than the best-in-class 7K200 by around 1C or a little better, & is about equally noisy/quiet

Edited by shoarthing, 31 August 2008 - 07:01 AM.

#4 Guest_Vantage72_*

Guest_Vantage72_*
  • Guests

Posted 02 September 2008 - 05:51 AM

So my question is, which benchmark is the better predictor of real world performance: I/O Meter 2003 or PC Mark/ Vantage?


No one knows this?

#5 continuum

continuum

    Mod

  • Mod
  • 3,467 posts

Posted 03 September 2008 - 11:45 AM

What kind of "real world" performance? I/O Meter traditionally is considered better at capturing server/multiuser workloads, but both should give useful numbers. The Tom's Hardware review you are reading naturally uses different hardware than what the guy on NBR or what SR is using, but furthermore the Tom's Hardware I/O Meter results are intensely workstation and server role biased rather than the NBR's desktop-performance-biased testing with PC Mark Vantage.

To help clarify things Tom's Hardware even indicates that their I/O Meter results are for workstation, Database, and File Server patterns rather than for desktop use.

.
So long as the benchmark used is adequate, the far more useful thing is to ensure the rest of the testbed remains the same so that you can compare results from different drives on the same testbed properly... so what's faster? Honestly it depends how you're using it. For 99.9% of people here it will be desktop use that's the question-- even under "heavy" desktop multitasking there simply isn't enough load to get anywhere outside the realm of desktop loads, period.

#6 Guest_Vantage72_*

Guest_Vantage72_*
  • Guests

Posted 05 September 2008 - 10:27 AM

Yes I am talking about normal desktop use to medium-heavy multitasking desktop use. Thanks for your answer continuum.

So PC Mark 05 and Vantage are an ok indication if I understand you correctly.

Well then it's interesting that two other users have posted PC Mark scores for the Seagate 7200.3 and it comes out as the fastest. They measured a Win XP startup speed of 9,7 to 10,0 MB/sec. Much faster than what Tom's Hardware measured: 7,8 Mb/sec.

Posted Image

Posted Image
Free Image Hosting by ImageBam.com

#7 continuum

continuum

    Mod

  • Mod
  • 3,467 posts

Posted 05 September 2008 - 07:11 PM

Yes, but unless all test results you're using were on the exact same testbed, you can't directly compare them.

#8 rdzeng

rdzeng

    Member

  • Member
  • 8 posts

Posted 13 December 2008 - 07:14 PM

I have both hard drives (7k200 and 7200.3). I have noticed a response difference between the two of them in regular use. The 7k200 is noticeably more responsive in comparison to the 7200.3, and I believe it's reflective of the differences in the random access times.

My guess is that the electronics in the 7200.3 are more highly optimized in comparison to the 7k200, which means that in combination with the higher areal density, allows for increases in performance which covers up the difference in random access times in most benchmarks where the amount of data requested is more important than how quickly that data is initially accessed.

I have both hard drives (7k200 and 7200.3). I have noticed a response difference between the two of them in regular use. The 7k200 is noticeably more responsive in comparison to the 7200.3, and I believe it's reflective of the differences in the random access times.

My guess is that the electronics in the 7200.3 are more highly optimized in comparison to the 7k200, which means that in combination with the higher areal density, allows for increases in performance which covers up the difference in random access times in most benchmarks where the amount of data requested is more important than how quickly that data is initially accessed.

And I did use both hard drives on the same exact hardware.

I have both hard drives (7k200 and 7200.3). I have noticed a response difference between the two of them in regular use. The 7k200 is noticeably more responsive in comparison to the 7200.3, and I believe it's reflective of the differences in the random access times.

My guess is that the electronics in the 7200.3 are more highly optimized in comparison to the 7k200, which means that in combination with the higher areal density, allows for increases in performance which covers up the difference in random access times in most benchmarks where the amount of data requested is more important than how quickly that data is initially accessed.

And I did use both hard drives on the same exact hardware.



0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users