Is something wrong? Raptor 150GB 10k RPM performance vs SeaGate 500GB 7200 RPM

[immersive]

Hello,
I recently built a new computer and bought a two new SeaGate 500GB, 7200 RPM, 16MB Cache, 3gbs data rate hard drives. I don't know a lot about bench numbers but it looks like this drive is out performing my WD1500ADFD Raptor 10,000 RPM 16MB 1.5gbs. Is this because of the 3gbs vs 1.5gbs data rate? Maybe I just don't understand the numbers or something?

What drive for maximum performance would you recommend as my primary drive for system and program files only? Maybe one of those 30gb 15k rpm ones? My motherboard is a ASUS P5E-VM HDMI.

My HDD setup is:
WD Raptor 150GB; For program and system files only
SeaGate 500GB; For media storage
SeaGate 500GB; Mirror backup of previous drive

Take a look at my HDTach results:

 

[jiggamanjb]

That's right.

Burst - throw out, it's a stupid measurement
average - copying large files uses this - good for media storage
random - basically your every use - good for program and system files

HD tach really needs to redo their results viewer. It throws everyone off.

 

[noegruts]

That's right.

Burst - throw out, it's a stupid measurement
average - copying large files uses this - good for media storage
random - basically your every use - good for program and system files

HD tach really needs to redo their results viewer. It throws everyone off.

There's nothing really wrong with HD Tach's results viewer, all the information is there. The problems mostly arise when people jump to conclusions without educating themselves first.

No disrespect intended to you, immersive, this is just a general observation.

 

[dominick32]

Yes. HDTach results only prove that these drives are solid performing high sustained transfer drives. Please read my full direct review on the WDRaptor 150 vs. the Seagate 7200.11 500GB drive: http://www.nextlevelhardware.com/storage/barracuda/

This should show you the real differences between the two drives. Price to performance, the Seagate is the pure winner. The Raptor is the overall winner in the random read, gaming, loading scenario making it the day to day quicker drive. But, not by much at all.

Something to note as well: If you are considering these drives in redundancy for a web or database server, the Raptor 150 is substantially faster in both reading/writing in a server environment.

Dom

Dom

 

[sajberTim]

Hehe Here's a screencap of my 7200.11 500gigs in RAID 0+1 Dunno about your WD disk tho. In the end HDTach means zero in the real world but, yea it's fast

 

[KillrBuckeye]

There's nothing really wrong with HD Tach's results viewer, all the information is there. The problems mostly arise when people jump to conclusions without educating themselves first.

No disrespect intended to you, immersive, this is just a general observation.

I tend to agree with jiggamanjb. I can't tell you how many threads I've seen like this one where the OP is saying something like "OMG, my Seagate RAID array totally owns my Raptor array", and then posts the HD Tach results showing that the Raptors have much faster random access. Random access is most important for most real-world usage, and HD Tach displays it as a tiny little line item that most people will ignore. Why not have a bar chart showing random access times instead of the useless burst speed?

 

[noegruts]

Random access is most important for most real-world usage

I don't necessarily disagree with your overall observation about HD Tach, but the statement above isn't always true - whether you need faster STR or faster random access depends entirely upon your workload (ie what you are using the drive for).

 

[noegruts]

Hehe Here's a screencap of my 7200.11 500gigs in RAID 0+1 Dunno about your WD disk tho. In the end HDTach means zero in the real world but, yea it's fast

Here is what five of them will do in RAID 0...

(the relatively low burst speed and high CPU utilization is due to the celeron 420 on my WHS machine)

 

[sajberTim]

Here is what five of them will do in RAID 0...

(the relatively low burst speed and high CPU utilization is due to the celeron 420 on my WHS machine)

Hehe nice But what's up with that humongous seek time

 

[immersive]

Thanks for the help guys! I noticed the better random access but wasn't sure if that made up for the better burst speed and average read on the SeaGate.

Does having only your system and program files on a fast drive make overall system loading better? This system isn't used for gaming but is used for web design and some recording / media playing. I was thinking about ditching the Raptor for a new 15k rpm 30gb drive instead. What are your opinions? Any suggestions for a 15k drive?

My system specs:
CPU: Q9300
MB: ASUS P5E-VM HDMI
RAM: 4GB
Video: ATI HD3450
HDDs:
WD Raptor 150GB; For program and system files only
SeaGate 500GB; For media storage
SeaGate 500GB; Mirror backup of previous drive

 

[jiggamanjb]

I don't necessarily disagree with your overall observation about HD Tach, but the statement above isn't always true - whether you need faster STR or faster random access depends entirely upon your workload (ie what you are using the drive for).

True, I figured on an overclockers forum I'd be safe with that If anyone was doing video editing, large picture editing, etc., most of the time, I'd go with the 7200.11's for sure.

 

[noegruts]

Does having only your system and program files on a fast drive make overall system loading better?

This question doesn't really make any sense. Yes, a "fast" drive is better than a "slow" drive. Were you specifically talking about access times here? If so, many people here run Raptors as their OS/Program Files drive and "slow" 7200rpm drives for their data.

This system isn't used for gaming but is used for web design and some recording / media playing. I was thinking about ditching the Raptor for a new 15k rpm 30gb drive instead.

15k rpm drive is going to be expensive and noisy. Maybe that's ok for you, I dunno. What problem are you trying to solve, specifically?

 

[noegruts]

True, I figured on an overclockers forum I'd be safe with that If anyone was doing video editing, large picture editing, etc., most of the time, I'd go with the 7200.11's for sure.

Ha, actually I don't even overclock - my E6750 is hardly ever too slow for what I need. I just come here for the beer

I do a fair amount of video and photo editing, so that might put some of my opinions in context.

btw, let me pimp the new WD6400AAKS drive here too, it is currently my favorite drive in this capacity range. It is also a fantastic value at ~20 cents/GB

 

[ratbuddy]

Hehe Here's a screencap of my 7200.11 500gigs in RAID 0+1 Dunno about your WD disk tho. In the end HDTach means zero in the real world but, yea it's fast

Any idea where those spikes come from?

 

[noegruts]

Any idea where those spikes come from?

It is just an artifact of running the "quick" test on HD Tach, which means that the line made up of a small number of samples, plus most likely Windows was running from the same drive. So whenever Windows wants to access a file, it takes a little speed away from the benchmark.

 

[noegruts]

Hehe nice But what's up with that humongous seek time

My server isn't short-stroking the drives, I just set-up the entire array as one volume, so that's the access time I'd expect with the ST3500320AS drives. Note that access time and seek time are not the same - access time = seek time + rotational latency.

 

[sajberTim]

Any idea where those spikes come from?

Yep, "Joss Stone - Mind, body & soul - 2004" in Mediaplayer Classic

 

[MadMan007]

Yes. HDTach results only prove that these drives are solid performing high sustained transfer drives. Please read my full direct review on the WDRaptor 150 vs. the Seagate 7200.11 500GB drive: http://www.nextlevelhardware.com/storage/barracuda/

This should show you the real differences between the two drives. Price to performance, the Seagate is the pure winner. The Raptor is the overall winner in the random read, gaming, loading scenario making it the day to day quicker drive. But, not by much at all.

Something to note as well: If you are considering these drives in redundancy for a web or database server, the Raptor 150 is substantially faster in both reading/writing in a server environment.

Dom

Dom

Know what I'd like to see Dom? A Matrix RAID slice test of some 7200RPM drives vs the Raptor. With small slices across multiple drives HDTach shows that seek times decrease a fair amount. You'd have to keep it realistic though, no 20GB across 4 drives as an OS/app array Maybe something like 100GB across different number of drives? It's a test I haven't seen done anywhere and would lay to rest the whole 'Matrix RAID is a fake boost' debate when access times are frequently more important.

 

[jiggamanjb]

Anyone ever done a Matrix RAID on Raptors? That should be the bar!

 

[MadMan007]

I have a quad Raptor RAID0/RAID5 array. The RAID0 slice is relatively large I made it a real size, actually a bit larger than I'd probably need, instead of a silly small benchmarking size.