Notebook hard disc weakest component

I always thought Seagate delivered when it came to hard drives.. even the reviews were praising. But now this happens.




my question: Is it possible to get better hard disc performance?


Model : SEAGATE ST9500420AS
S/N : 5VJ127MN
Firmware : 0003HPM1
Serial ATA : Yes
Serial ATA version : 2.0 - (SATA-300)
Support : ATA8-ACS
Size: 500 GB
Solid State Disk (SSD) : No
Format : 2,5"
Speed : 7200 rpm
Cache : 16 384 KB

 

1) 5.9/7.9 is a very good score
2) WEI is not a good benchmark for anything
3) Get an SSD if you really want to get good performance

 

Do you mean without actually changing HDD? If so, there isn't much to do except regular defrags really and eliminating the junk.

If you can change the disc, get an SSD; they offer great performance and the degradation over time isn't like those of HDDs.

 

Besides getting an SSD? Edit your WEI scores

On topic:
WEI shouldn't be taken seriously though. It changes from time to time as well. At least whenever I do my WEI test... >_>

 

haha, I guess I took WEI a little bit too seriously?
Didnt know it doesnt represent anything. Why isnt it a good benchmark?
Oh well, SSD technology is far too expensive if you see how much GB u can get for a euro.
I think those prices wont come down anytime soon.. Ive got a Cowon S9 pmp with 16 gb ssd in it though.
Ill enjoy that for the time being, Thanks for the replies and letting me adjust my opinion on the Windows facade.

 

Don't take WEI seriously. I have 2 Cheetahs in RAID 0, and they only score 6.1.

 

To bench your HDD get HDTune or CrystalMark and see how fast your HDD is at read/write. If it's over 20 MB/s usually then it should be fine.

You can just edit the WEI scores by changing an XML file and make it whatever # u want.

 

as of late, it appears that the disk is the bottle neck. an ssd will increase speed significantly.

increase performance on a laptop -
- get an ssd
- get 8gb ram
- get the fastest proc.

 

Get an SSD if u want super performance. WEI is not a good benchmark... my crap stock drive gets 5.5 although it is way slower.. Use HD tune to benchmark and post your results. Also remember to regularly defrag the hard drive. I recommend Smart Defrag as a defragmentation programme.

 

So? SSDs are faster Win7 was designed with SSDs in mind.

WEI is a very good benchmark to get an idea if something is seriously wrong.

Say you had a powerful NVidia card and scored 1.0 on the graphics - then you know something is wrong.

On the other hand, I think my X3100 get's a 3.0 and 3,5 or so in Vista - does it matter? Well not at all, I could possibly live with a weaker card - in the end benchmarks don't mean a thing - you should be happy with the performance you get when you use your computer.

E.g. if you buy a netbook, don't expect gaming.
Buy a light laptop (rather than a netbook) for "General use" buy something more powerful for heavy use etc.

 

The problem with WEI is that it's even more erratic than other synthetic benchmarks.

It is highly affected by drivers and the actual scale itself is variable. When I was first introduced to WEI, I think it went up to 5 or somewhere along those lines, now it goes higher IIRC.

I recall updating m ATI Catalyst driver and all my benchmarks & performance went up slightly, but my WEI score actually went down noticeably(well 1.0 point if you call that noticeable).

Nonetheless, benchmarks will be benchmarks(WEI or not), real life performance is what matters.

 

Vista is 5.9, 7 is 7.9. WEI doesn't seem to count drives in RAID, either, I get the same rating from my 1tb samsung's raid or not.

 

Please stop paying attention to WEI harddrive score

As said before, go for CrystalMark or the IOPS benchmark in HDTune Pro.

 

Funny thing is, a few months ago, 5.9 was considered good. Then they raise the ceiling in Win7 and all the sudden people are like,

"WHERE IS MY PERFORMANCE!!!????"

Everything is so relative....

 

Yeah I get 7.6 on everything and a 5.9 in HDD on my desktop, slightly depressing

 

HAHA, so if isnt a good indicator.. and effected by the smallest things.
MS kinda fails, it looked as if it had a function and promising before I posted
my question, oh well. HD Tune Pro it is, lets see:

Maximum - 34.6 MBs
Minimum - 97.3 MBs
Average - 71 MBs

Well, thats good I believe.

Acces time - 16.5 ms
Burst Rate - 56.9 MBs
CPU Usage - -1% ( ?.. does that mean its a power source? lol )

Thnx for the support people! im off to watch the new Better off Ted

 

I hope you all know the Mechanical Hard Drives are limited to 5.9 in the WEI, even if it performs better than that.

It's just another reason to not take the scores too seriously since they're clearly using the scores as a method of increasing hardware upgrades.

 

Never said if before, saw it a lot.. but its true.
Micro$oft i$ $cheming.

 

You may want to compare it to some of my benchmarks if you want?

See:
http://forum.notebookreview.com/showthread.php?t=441674

First link in post above is the benchmark scores for some of my most recent HD's.

You might also want to download CrystalDiskMark 3.0 beta 1 from here:
http://www.ntcompatible.com/CrystalDiskMark_3.0_Beta_1_s136986.html


this will allow you to compare in more detail the scores I got using that tool.

(In CDM 3.0 b1, make sure you run the appropriate 32bit or 64bit version and click Edit, Copy and then paste that into a new txt document to see more detailed information than the main screen of CDM shows).

 

The IOPS benchmark in HD Tune Pro is located on the page "random acces" and says a lot more than these scores.

Here's the IOPS benchmark for a Scorpio Black and Seagate 7200.4





@ tilleroftheearth, could you post your 7K500 IOPS screenshot here for comparison?

 

Phil, sorry I don't know how!

But here is the text version - btw, the 160GB Scorpio Black must feel very fast.

512 bytes:.54 IOPS, 18 ms, 0.026 MB/s
4 KB:........53 IOPS, 18 ms, 0.208 MB/s
64 KB:.......51 IOPS, 19 ms, 3.241 MB/s
1 MB:........32 IOPS, 31 ms, 32.105 MB/s
RAND:.......40 IOPS, 24 ms, 20.152 MB/s

As originally reported here:
http://forum.notebookreview.com/showthread.php?p=5608617#post5608617

 

I have a hard time believing the 7K500 is slower than the Scorpio Black and Seagate 7200.4. I'm expecting there's something wrong with the benchmark setup. Especially the acces times seem too high, thus impacting the MB/s negatively.

Are you sure there were no background processes running?

 

Phil, please understand that I don't 'benchmark' as most people do here.

This is on a live system with a full working set of programs and utilities running as they 'should'. At the time the benchmark was started, the only thing extraordinary I did was execute a processidletasks command to finish any pending operations that might be running or start in the middle of the run.

This may help you to understand my setup better too:
http://forum.notebookreview.com/showthread.php?p=5617217#post5617217

Maybe its simpler if you tell me how your benchmarks were run? I might try to duplicate the same environment for you on the Hitachi.

 

I ran the Scorpio Black benchmarks on an Atom netbook with Windows XP with about 25 processes running. I terminated most processes manually.

If you could terminate as many processes as possible I'd be interested to see if that improves your IOPS results.

 

25 processes! Wow, that's 1999 computing for me.

I'm running at 127 right now, but I may try this for you later tonight (working currently).

Even if it increases the IOPS, how would that compare to any real world use?

I'm even betting that they won't match the Scorpio Black's scores because I have noticed with increasing capacity, much less 'snap' than what I was used to with my 7K200. The drive manufacturers simply don't care in tuning the drives for 'feel' anymore - benches are what they're chasing, IMO.

 

The IOPS benchmark is the best synthetic indicator of real life performance I know of.

As long as no one has done any real life comparisons between the fastest hard drives, this as good as it gets.

 

The harddrive is always the slowest part of the system. I have over 400mb/sec data transfer on my laptop, and its still the bottleneck. It doesnt stand a chance againg 8 cores and 16gb of memory.

The Hitachi 7K500 is a good drive. Ive got one for my E1505. Its a bit better than my 7K320. Not really a noticeable upgrade though. Going to sell it and get an ssd for it.

K-TRON

 

how about taking WEI seriously, and realising why it might stop giving better points for hdds? because they ARE much slower than other options? maybe a <1ms access time deserves 2 points more in WEI?

all blaming WEI are just stupid. it's a benchmark. if you fail at the benchmark, it's your (systems configs) fault. not the benchmarks. always blame someone else.. and at best, always blame microsoft.

 

lets see, my maximum is 250MB or so, my minimum is 70MB. and my access time is 0.085ms.

so why, again, do you think your numbers are good?

 

No, not 0.085ms, ever. Not on a G1, anyway.

 

uhm, why not? G2 is 0.065ms. mtron is 0.1ms.

directly out of the intel document:

intel 50nm ssd, read latency: 85microseconds, or 0.085milliseconds.

here

 

Sorry, I stand corrected (with head spinning after 22 hrs straight on computers)!

 

it's quite shocking, not? current ssds have 0.065 ms access time. typical hdds 16ms. that means it can access 250 files while the hdd still tries to access the first one.

 

Yes, numerically it is stunning.

Using one though, I could swear I could see the electrons moving from one part of the memory to the other...


Seriously,

Those speed advantages really do not show up like that in real use. But benchmark programs sure make the most of them though, huh?

 

well, they do show up. in that the disk is always ready to access anything i want to have. in that it's able to read/write tons of data in parallel so it never blocks what i want to use.

i don't need to quote the famous "starting tons of apps simultaneously" video. that's a best case of course, but shows very well exactly THAT difference.

 

Well... if you have to access plenty of small files then you will notice it.

Now the question is, how many apps consist of a few hundred or even thousand small files...

 

det: firefox, for one.. starting on a 4200rpm disk with 30ms random access time, with tons of plugins, it took 1 minute. on an mtron ssd it took 2 seconds. on the intel, it takes <=1 second.

firefox is extreme, but most apps are not "one file" when starting. most use system components (like for gui, for file accesses, etc). those will be accesses. config files, other accesses. the apps own dlls and plugins and resources, own accesses.

and each file access means >1 disk access.

it quickly escalates quite a bit. not that extreme like firefox does, but all apps access >10 to >50 files when you start an app, easily.

 

Wow, Phil,

Sorry I've taken so long to get back to this thread - awake too long and couldn't find it for the life of me... anyway, I'm here.

I still can't figure out how to attach an image here but I did the test you requested on the 7K500, I don't think the scores changed much actually but here they are:

512 bytes:.54 IOPS, 18 ms, 0.027 MB/s
4 KB:........53 IOPS, 18 ms, 0.208 MB/s
64 KB:.......51 IOPS, 19 ms, 3.241 MB/s
1 MB:........32 IOPS, 31 ms, 32.041 MB/s
RAND:.......40 IOPS, 24 ms, 20.115 MB/s


the above was run on Win 7 x64 with 8GB RAM and 38 processes running (from 81 originally - the computer feels like crap like this, btw) with processidletasks run at the cmd line about 10 minutes before I ran the above test.

I tried many times to get it lower than 30 processes, but all I succeeded at was crashing the computer - hope this helps you.

ps. As I mentioned before, fast (7200 RPM) and large HD's do not mix well in my experience in terms of access times or IOPS (they are related, correct)?

Anyway, off to bed for me... 'G night!

 

In my opinion WEI is probably the worst hard drive benchmark out there. I don't see any reason to take it seriously.

Free alternatives like CrystalMark or even HDTune deliver much more, and more accurate information.

Thanks for that.

I think they are yes. I still expect the 7K500 to do better.

 

opinions don't matter. WEI determines how the os gets configured. based on WEI, stuff like defragmentation on ssds get disabled, superfetch gets configured etc.

it's not about having accurate information, for this, one can use other tools. but WEI shows what its supposed to: if a device works as it should, and if the device has the minimum requirements for certain features.

it allows to compare. like any benchmark. so stop bashing it. it works, it lets compare, what else do you expect?

it's definitely not the worst test. it's a 1-number info about the system which part is how fast. and it's good at that. saying no is just stupid microsoft-bashing.

 

The WEI hard drive benchmark is probably the worst hard drive benchmark. I've seen faster faster hard drives get lower ratings than slower hard drives. The OCZ Core V1 probably gets a very high WEI rating. Please ignore it.

I won't stop bashing it

PS. I'm not bashing Microsoft. I actually like MS.

 

sure. do so. spread wrong knowledge. we love it at this forum

 

If you provide me with evidence that WEI HDD score is indeed a benchmark "that needs to be taken seriously" I'll change my mind. Sofar you haven't and my conclusion remains: WEI HDD score is severely limited and there are several free benchmarking tools that provide a much better indication of real performance.

Just saying I'm wrong and implying that you are speaking for several people doesn't help.

 

sure thing.

then again, you do the same right now. there is no need to counterarguments your points: other benchmarks show more detailed results (but, being much in the ssd topics, i see that a lot of those fail to deliver accurate, or user important results themselfes, once tech changes).

well, my evidence is that WEI so far delivered what was needed for me: showing how weak the hdd was, which directly correlated with the user experiences on how slow the system was (which was always vistas fault, for them, obviously ).

it's a quick'n'dirty point-the-user-at-it thing: "there's your problem". and fixing it always helped (being it gpu, hdd, cpu, ram, what ever).

and thats what it's for. it's not for geeks to bench each inch of their system.

but so far it worked fine as a quick look.

you state, it doesn't, as you had wrong results. i counterargue no, the results where fine here always, and it might have been your fault. but one can argue around that till eternety, and it's useless.

i just counterfight the generic "WEI SUCKS" statements spread in that forum. it doesn't. nothing is perfect, so isn't WEI, but it works for what it's supposed to.

 

I'm not sure if you do, but if you have a slow hard drive along with your Intel drive(don't know how it is on other SSDs, probably the same), your system will feel slower.

I once plugged in a 5400RPM notebook drive and my system felt significantly slower. My data storage drive is a 7200RPM desktop drive, not a crappy laptop one.

I think one day I'll reassign this G1 to a data drive and future G2 to the main OS drive, and ditch the 7200RPM. For me I think the fact that I don't have to waste time formatting and still be much faster than a fresh platter HDD even at degraded state is enough of a reason to get an SSD.

That and less disk repairs I have to do with tools like Scandisk.

 

I'll agree that WEI is good for a quick and dirty look. That's it.

 

and that's what it's made for. (and that the system can verify it's own hw performance to adjust accordingly)

 

IntelUser,

Those are great and interesting points you make. I will definitely keep them in mind.

One question does come to mind; is the data storage drive a SATA1 or SATA2 drive? Or, maybe more importantly, is the enclosure SATA1 or SATA2? Just trying to think of a logical reason as to why it would slow the system down significantly enough for you to notice. Besides that's it a mechanical HD, of course.

 

Yes, I was hoping for the 'feel' of the 7K200 myself, but either way, definitely an upgrade from the Scorpio Blue 500GB I was using.

Cheers!

 

Thank you, ok lets see, what does this tell us:




I think I can measure with the 2 youve posted, the differences arnt much.

Well, because I havent got a SSD.. never even dared to look at one cause of the price.
But holyfudgestickles...... never thought that it would make such a difference.