Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Mr.PS

macrumors 6502a
Original poster
Jan 8, 2008
528
9
The picture says it all, and no - I, nor the readers of this thread do not want to be bored with lectures about raid and increased failure statistics... We all know the risks, backups are a must, with or without raid.

the array is compromised of two 750gb Samsung Spinpoint f1 32mb cache hdd's.

 
well, no...the picture doesn't say it all.

how about a comparison of the 2 samsungs in RAID0 vs the samsungs by themselves. all you did was compare them to the slow stock drive that everyone knows is slow.
 
well, no...the picture doesn't say it all.

how about a comparison of the 2 samsungs in RAID0 vs the samsungs by themselves. all you did was compare them to the slow stock drive that everyone knows is slow.

Well my comparison was stock vs samsung array. I didn't have time to bench mark them individually, although it would be a good benchmark. This shows how much faster raided sammys are over the stock hdd. I think this is a good benchmark and eye opener as well.
 
yes, but you're not actually showing what the benefit of the RAID is. everyone knows the stock drive is a dog. comparing single drive to single drive is an effective benchmark and showing unRAID'd vs RAID'd drives (of the same make) is an effective benchmark.

There's no real way (from your test) to see how much of the improvement is the difference of the drive or the addition of RAID. It's like comparing a Supercharged Mustang to a stock Civic. Of course the Mustang is faster, but you don't know how much is the car and how much is the supercharger.
 
yes, but you're not actually showing what the benefit of the RAID is. everyone knows the stock drive is a dog. comparing single drive to single drive is an effective benchmark and showing unRAID'd vs RAID'd drives (of the same make) is an effective benchmark.

There's no real way (from your test) to see how much of the improvement is the difference of the drive or the addition of RAID. It's like comparing a Supercharged Mustang to a stock Civic. Of course the Mustang is faster, but you don't know how much is the car and how much is the supercharger.

That I agree with, this benchmark will be most beneficial to those looking to run a Sammy raid over the stock drive. That was my intended purpose. That's what I went for at first, I never considered a single sammy, just a raided sammy setup.

lol @ sammys. Reminds me of this burger joint around here .
 
Can anyone else with a Samsung F1 post the single drive benchmarks?

After seeing how well it did in RAID 0, I may return my unopened F1 1TB to newegg and get 3x750gb for a raid 5.

Damn, or maybe 2x750's raid 0 and keep the 1tb for backups.

Maybe I should quit worrying about raids and get upgrade one of my old, viewsonic lcds... decisions decisions.
 
Can anyone else with a Samsung F1 post the single drive benchmarks?

After seeing how well it did in RAID 0, I may return my unopened F1 1TB to newegg and get 3x750gb for a raid 5.

Damn, or maybe 2x750's raid 0 and keep the 1tb for backups.

Maybe I should quit worrying about raids and get upgrade one of my old, viewsonic lcds... decisions decisions.

I say keep the 1tb for backups/time machine. That's what I'm using the stock 320gb drive for. I would like to see a single sammy benchmark as well too!

Do you have a RAID-card or just the software RAID?

Software raid...
 
The "pre-raid" test looks incomplete - you're missing the random uncached read (256K blocks) score, which would cause the overall score to be artificially lowered too. Do you have a full set of data for comparison?
 
The "pre-raid" test looks incomplete - you're missing the random uncached read (256K blocks) score, which would cause the overall score to be artificially lowered too. Do you have a full set of data for comparison?

Actually, I noticed that too, I let it run and it appeared to finish, thus allowing me to save the result. I'd really have no reason to forge any of these numbers. The stock drive was more then enough, I just wanted some more juice for video/audio and loading of plugins/programs, etc..
 
Ok here is the Samsung F1 750gb in a single configuration: (I do have 10gb ram and its in a Mac Pro 8 core 2.8ghz)
 

Attachments

  • Picture 3.png
    Picture 3.png
    45.7 KB · Views: 112
Well my comparison was stock vs samsung array. I didn't have time to bench mark them individually, although it would be a good benchmark. This shows how much faster raided sammys are over the stock hdd. I think this is a good benchmark and eye opener as well.

Thanks for sharing this info, Mr. PS, it's very helpful. So, how did you migrate the OS and files from the 320GB drive over to the RAID?

Thanks!
 
Ok here is the Samsung F1 750gb in a single configuration: (I do have 10gb ram and its in a Mac Pro 8 core 2.8ghz)

That's quite an improvement :eek:
I got a 56.28 score on my HD test.
I'm waiting on my Samsung F1 750gb it should be here some time today.
 
Here's my benchmark with one of the 1TB Samsung HD's. Not too different from that of the 750GB F1's.

2.8X8/8800/10GB
 

Attachments

  • Picture 2.png
    Picture 2.png
    96.1 KB · Views: 105
That's quite an improvement :eek:
I got a 56.28 score on my HD test.
I'm waiting on my Samsung F1 750gb it should be here some time today.

Indeed :D I was thoroughly impressed with the big improvement over the stock drive and I can only thank people on here after all the discussions about them.
 
What about the Accesstime?
IMHO it is more important for fast Program Startup.
RAID 0 slows a little bit the Accestime.
Does NCQ work in a RAID 0 Configuration?

Greetz
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.