- Joined
- Feb 4, 2007
- Messages
- 131
What is faster? Intel Matrix or JMicron Array? I setup my raid w/ 3 WD 250GB HDD's and I went with the Intel Raid and I don't really see a huge speed difference.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Raid-0 Question
- Thread starter jfespinoza
- Start date
Oh god, not again.
I will assume you went with a Raid0.
A Raid 0 array will increase the sequential read speed of the array. It will not improve the seek time of the array. This will lead to an increase in real world speed only in situations that require large file transfers, but only if the sectors are stored sequentially. If the heads have to move much at all, you will see little if any "REAL" world performance increase.
That said, how are you quantifying your claim that you have no speed increase. If it just does not feel faster, then I would say that is normal. If you are benchmarking your array, then you should see a dramatic increase in your sequential read score.
This argument gets beat to death fairly regularly around here. Do some searches on the forums, and you will see the endless arguing about Raids performance gains, or lack thereof.
Some people swear up and down by Raid arrays and wont live without them. Others, like me, have tried them and been unimpressed. They have some benefits, but I feel they are not worth the hassle and expense.
Don
I will assume you went with a Raid0.
A Raid 0 array will increase the sequential read speed of the array. It will not improve the seek time of the array. This will lead to an increase in real world speed only in situations that require large file transfers, but only if the sectors are stored sequentially. If the heads have to move much at all, you will see little if any "REAL" world performance increase.
That said, how are you quantifying your claim that you have no speed increase. If it just does not feel faster, then I would say that is normal. If you are benchmarking your array, then you should see a dramatic increase in your sequential read score.
This argument gets beat to death fairly regularly around here. Do some searches on the forums, and you will see the endless arguing about Raids performance gains, or lack thereof.
Some people swear up and down by Raid arrays and wont live without them. Others, like me, have tried them and been unimpressed. They have some benefits, but I feel they are not worth the hassle and expense.
Don
The JMicron is used for the esata on that board.
I noticed a huge difference between a single 36GB raptor, and two of them in Raid 0.
I've built computers for friends with a single huge 7200RPM drive and I DO notice a difference.
You nay-sayers can say all you want. Wasn't there just another thread with people bashing these drives? Pay the extra money and pull out the floppy drive, the performance is worth it. And plus, this is the most unlikely thing to ever change in your system. You buy two 150GB 10k RPM drives and your good for years... and I mean years. Got another system with two 74GB 8mb cache in raid 0 for 4-5 years now? still going stong.
I noticed a huge difference between a single 36GB raptor, and two of them in Raid 0.
I've built computers for friends with a single huge 7200RPM drive and I DO notice a difference.
You nay-sayers can say all you want. Wasn't there just another thread with people bashing these drives? Pay the extra money and pull out the floppy drive, the performance is worth it. And plus, this is the most unlikely thing to ever change in your system. You buy two 150GB 10k RPM drives and your good for years... and I mean years. Got another system with two 74GB 8mb cache in raid 0 for 4-5 years now? still going stong.
I believe the Intel Matrix will allow you to run a Raid 0 and Raid 1 arrays on the same drives. Kind of a neat trick. Link
One person's observation.
Your results.
In theory, 3 drives in Raid 0 should be another 50% faster than 2 drives. In some benchmarks, they will be. But in real world use, you won't notice alot of performance gains with Raid 0 because it does nothing to improve seek times.
If you already have the drives, run them in Raid 0 and get what performance you can. But remember, you have now tripled your chance of catastrophic array failure. If any of the 3 drives fails, all your data goes bye bye. Just make sure you have a backup of any irreplaceable data.
Look, I'm just saying don't throw alot of money at a Raid setup expecting something dramatic. You will prolly be disappointed. just like the OP.
I have an Asus P5K, and the Jmicron controller on my board also controls the PATA interface and 1 of the 5 Sata ports.
Don
One person's observation.
Your results.
In theory, 3 drives in Raid 0 should be another 50% faster than 2 drives. In some benchmarks, they will be. But in real world use, you won't notice alot of performance gains with Raid 0 because it does nothing to improve seek times.
If you already have the drives, run them in Raid 0 and get what performance you can. But remember, you have now tripled your chance of catastrophic array failure. If any of the 3 drives fails, all your data goes bye bye. Just make sure you have a backup of any irreplaceable data.
Look, I'm just saying don't throw alot of money at a Raid setup expecting something dramatic. You will prolly be disappointed. just like the OP.
I have an Asus P5K, and the Jmicron controller on my board also controls the PATA interface and 1 of the 5 Sata ports.
Don