| Forum Home | ||||
| Press F1 | ||||
| Thread ID: 40640 | 2003-12-13 10:43:00 | SATA and IDE Drives | jonduc (4754) | Press F1 |
| Post ID | Timestamp | Content | User | ||
| 200194 | 2003-12-14 04:42:00 | what motherboard, what drives and what was it tested with?? | tweak'e (174) | ||
| 200195 | 2003-12-14 05:40:00 | Gigabyte GA-8KNXP V2, Dual Segate 120MB drives with 8MB cache set up as a RAID0 array. Does not matter what it was tested with as the programs are relative and two different programs will give you two different results. As long as you use the same program to test both setups then you will get a clear indication. The indication was that the SATA mothered the IDE setup in read access but not as much in write. Can't remember the exact figures off the top of my head now but I tested both to see which one to use to get the best performance. |
Big John (551) | ||
| 200196 | 2003-12-14 06:01:00 | your useing a RAID array. raid setups are a totally different matter. there can be huge differences in chips and setups. you can't compare a raid to an ata system. any chance you can give more details on how the two setups where setup? |
tweak'e (174) | ||
| 200197 | 2003-12-14 06:23:00 | There are three RAID connections on the Gigabyte board. Two use the IDE interface and one uses the Northbridge chip. The board can also remap the SATA ports to IDE ports as well. One of the RAID modes uses the IT8212 controller which is on the IDE side. The second uses a SiL chip which is also SATA (I think, could not find much info on this one so did not touch it). The third is the Hardware SATA which is the Intel ICH5. All options set up as a RAID0 array. |
Big John (551) | ||
| 200198 | 2003-12-14 07:56:00 | as no drive goes quicker than 133, sata won't make a difference. in your case the differences in chipsets, stripe size and drivers can make a huge difference. . this all has been tested and thrashed out every time a new ata standard comes out. its not untill you get drives that go faster than the old standard that you will see any major improvement by useing a faster standard. in your case the "improvements" are related to the chipsets and setup rather than the use od sata. |
tweak'e (174) | ||
| 200199 | 2003-12-15 05:09:00 | However as I said using a SATA drive does indeed make a difference. You have even said it in the sentence. Regardless of where the increase comes from it does make a difference when used in my setup which is what I said. If I attach it to the IDE then slower results. |
Big John (551) | ||
| 200200 | 2003-12-15 08:40:00 | o god this is how urban legands start :( ok lets aee if i can explain it in a bit more detail. i can understand your logic. "if i connect it to SATA it goes faster" therefore SATA must be faster and better. unfortunatly, unless you have a drive that goes faster than 133, its false logic. >if I attach it to the IDE then slower results. this actually makes a lot more sence. think of it this way, going to SATA didn't GAIN any performance (it ran at normal speed) but going to ata133 LOST some performance. you just need to find WHY it didn't run as fast. ata133 is to slow?? no drive goes as fast as 133 so that won't be why.comman reasons drives running slow are driver problems, bios, setup and OS faults. once you fix those faults the speed will be the same as SATA. your benchmarking (ata133) would have shown the sustained transfer rate way lower than what the drive itself is rated at, its usually a good intication there is a problem somewhere. no doubt the SATA transfer rate would have been normal. what would be faster, an ata33 drive connected to an ata66 controller or an ata33 drive connected to an ata133 contoller??? neither in both cases the fastest the drive can go is 33mb/s. connect the same drive to SATA and it will run at 33mb/s. the next thing to concider is that your useing RAID setups. they can vary HUGLY in performance. the differences you noticed can simply be differences in the chips or setup. you can get 4 different SATA RAID cards and they all will do different speeds. if your are going to compare SATA to ata133 then at least use the same brand and with the same settings. also to note- any raid/ata controller built into the motherboard chipset should be faster than any onboard or pci raid/ata controller reguardless of if its SATA or not. sorry to hijack this thread but don't be fooled that converting a ata drive to SATA is going give you an increase in speed. once the drive speeds get close to 133 then SATA will be the only way to go. |
tweak'e (174) | ||
| 200201 | 2003-12-15 10:10:00 | >>but don't be fooled that converting a ata drive to SATA is going give you an increase in speed. once the drive speeds get close to 133 then SATA will be the only way to go. I'd say it currently depends on chipsets and drivers used... at best my nforce2 IDE controllers just cant hold a candle to my SI SATA chipset... consider that the nforce2 southbridge is one extremely complex beast, the less strain on it the better.. otherwise it slows down and becomes a bottleneck.. and this has been reflected in extreme heat output, audio issues and acpi issues documented all over the internet (take a visit to nforcershq.com's forums) I also have hdtach results to prove that my PATA drives running with DSE SATA-PATA convertors have a far more stable performance curve than exhibited on my nforce IDE controllers with any of the driversets I do agree that SATA is at present a new standard, and is prone to overhyping and a lot of myths and misconceptions, but you need to calm down in your delivery of a reality check and back it up with links. Many SATA drives are basically PATA with a bit more cache, a PATA-SATA bridge (all manufacturers excl. seagate) and different connects... the drive manufacturers are only just bringing out proper SATA drives with more of the standard's featureset eg enabled native SATA command queueing.. that's where performance increases are going to start to show through also dont live by the Mb/s speed.. that's only a burst speed... you want to focus on more important things such as rpm, seek time and sustained data rate and some copy/pastes Disc drive data rates have not exceeded ATA100 limits yet, why should I switch to Serial ATA? The maximum internal data rate on an IDE disc drive today is ~72MB/sec. The ATA/100 data transfer rate has not been reached but one of the reasons IDE performance is where it is today is due to the expandable data path PATA has allowed. That data path in PATA has reached its limit. Serial ATA allows disc drives to continue to offer performance and reliability at cost parity to Parallel ATA. In addition, Serial ATA interface requires less voltage meaning better power consumption and management in both desktop and mobile applications. The thinner cable allows for flexible designs and improved airflow in smaller form-factors. Will I see a performance difference in Serial ATA drives? You may see a 1 - 5% performance increase from a PATA drive to a Serial drive but the main performance benefit is in the long run when because of Serial ATA, the hard drive through put will not bottle neck the system performance. In the mean time, system integrators and OEMs will enjoy a big reduction in assembly time and reductions in handling damage due to connector and pin issues. |
whetu (237) | ||
| 200202 | 2003-12-15 10:38:00 | My prediction. SATA will be a dead-end standard. The basic design flaw of the drive hasn't been cured, makin em read faster to make use of the bandwidth and of course increasing the capacity is only going to bring about the same problems we have with our current drives. That being that they are crap. Until they make a device with large capacity and no moving parts available at a reasonable price i simply aint buying into it. |
metla (154) | ||
| 200203 | 2003-12-15 19:44:00 | > o god this is how urban legands start :( > > ok lets aee if i can explain it in a bit more > detail. Okay I will let you have your way and you believe what you care. All I know is that If I had put two Segate 120GB 8MB cache IDE drives setup as a RAID0 on the IDE Buss then it would have been a heck of a lot slower than two Segate 120GB 8MB cache SATA drives on the Northbride. Both RAID0 setups exactly the same strip size et al. I know the speed increase "I" got and dont care how you care to explain it away how SATA is actually not really any faster. To me it is and was worth it. To everyone else you need to try setups and see what your results are to get the best value for money. |
Big John (551) | ||
| 1 2 3 | |||||