IDE vs SATA

[oenone]

Before I go ahead and buy my hard drive , I have one quick question. Will there be a performance increase if I went with SATA once I have actually opened an application? (Faster rendering times, or a higher frame rate in games) Thanks, and sorry for the silly question ;)

[Catweazle]

You will only experience a performance increase in circumstances where you are accessing data on the drive. Games will not benefit other than with quicker load times. Applications such as rendering will only benefit if the data files being worked on (and the resultant calculations) are too large to be held in system memory.

[oenone]

Thanks, I was just making sure. I see so many people going for more expensive SATA hard drives with about 10 thousand RPM. My question is, why spend so much money if all it does is allow you to interact with your hard drive a few miliseconds faster? (Servers are a different matter, I am talking about SATA for a pc.)

[rcrevolution]

was my post deleted?

[Catweazle]

There are no deleted posts in this topic. Posts which get deleted are still viewable to me ;)

For some tasks, the faster drive operation can make a lot of difference. Video editing is perhaps the most commonly used such task.

[oenone]

Ok thanks, I don't think I'll be doing any video editing with this computer.

[Thong_Ispector]

Cat...
How do the SATA compare to EIDE on a Raid Controller
Raid striped 133's get you 266 +\- which I use for my video work...
I have not had any experience with the SATA's yet...

[Zachery]

SATA's are alot faster overall but unless your motherboard has SATA capability and move alot of massive files you will not see any huge increase in preformance.

[Coconut Monkey]

SATA (150Mb/s max theoretical transfer) is not much faster than ATA133. However, the standard does offer smaller cables, easier physical installation (no jumpers), and some degree of future proofing (should IDE ever be phased out - which won't be happening any time soon).

Throwing two drives into a RAID 0 configuration does not net you twice the speed, as numerous benchmarks can attest. A combination of overheads, access times, the fact that most RAID controllers in consumer boards are software based and lack of data redundancy in RAID 0 means that RAID 0 isn't really worth it for general usage and gaming (unless you really like gaining several precious seconds in loading screens). RAID 1 can be handy if you're paranoid about data loss, but you are sacrificing an entire drive for mirroring.

[Thong_Ispector]

Those benchmarks were interesting...

However I would rather have a benchmarking program to compare what a drive does on my own systems.

In real life, I see a huge difference in a system setup using the Raid configuration over a single drive, much more than the 10% - 15% that test shows.

Even at that modest 10% figure a 133 drive gets (146) close to the SATA (150)

[Thong_Ispector]

I just ran Cli Benchmark III SMP 0.7.16...

I ran it on two different PC's , both P4 2.4 with 512 meg and maxtor drives, both running 98se.

The only real difference is one is setup with raid (Mine) the other is not..(Wife)
Ha, Ha...

It found that average read speed DOUBLED on the Raid

It found the average write speed improved by over 50%

These findings are fairly consistent with the "Feel" of the system...
Side by side you would swear that the raid was running a much faster processor.

This is only a 150k download for anyone running an ATA setup on a similar processor. I would like to know if there is an SATA in my future.

[alc6379]

I just ran Cli Benchmark III SMP 0.7.16...

I ran it on two different PC's , both P4 2.4 with 512 meg and maxtor drives, both running 98se.

The only real difference is one is setup with raid (Mine) the other is not..(Wife)
Ha, Ha...

It found that average read speed DOUBLED on the Raid

It found the average write speed improved by over 50%

These findings are fairly consistent with the "Feel" of the system...
Side by side you would swear that the raid was running a much faster processor.

This is only a 150k download for anyone running an ATA setup on a similar processor. I would like to know if there is an SATA in my future.

But see, you're still talking theoretically when running a benchmark. If that system was a dedicated video editing solution, then yes, RAID would be an excellent choice. But, on a desktop system, unless you're just interested in the extra drive space, I don't think you're going to notice a real impact on system speed. I've got two 7200RPM ATA133 drives in RAID, and I've not noticed that much of a difference.

[Thong_Ispector]

I can agree benchmarks are never a good way to judge a system.
Every report or benchmark gives RAID a 10% to 15% advantage over non-raid.

That puts it right there with the SATA setup

Two systems side by side is the best way to evaluate performance...

My system is used for video editing...plus some games...

There is a noticeable difference between these two systems.

How long has it been since you ran your system with a single or non-raid drive...

There is also a big difference in Raid controllers, software based vs dedicated processor etc...

I hope they make some major advances before I need to upgrade again...
I would like to see a clear winner before spending my money...

[rcrevolution]

can you run a primary SATA HDD and a slave ATA/133? on the same system?

and i have a DFi LANPARTy mobo does that support everything SATA has too offer?

[Coconut Monkey]

can you run a primary SATA HDD and a slave ATA/133? on the same system?

Yes.

and i have a DFi LANPARTy mobo does that support everything SATA has too offer?

Does it have SATA connectors (and thus a SATA controller)? Yes? Then you're SATA ready.

Most new boards come with SATA and RAID support.

[blindxp]

Having ran RAID 0 for years now, I see that some of you are seeing the same thing I saw, little performance gain with your two drives RAIDed. Now, try three. Big jump in performance, a noticeable one at that. Now try 4 drives in RAID 0. Sure you've just dropped your chances of making it through the next day with all your data intact to 1 in 4, but the performance screams with a real RAID controller card. No software RAID though, go for a real card and you'll see real performance. All my experience is with good ol' IDE 133s. I recently had a hard drive bite the bullet, in the middle of a backup of course, so I'm back to two drives and I can notice how much slower everything is. So my point is, if you're going to RAID, plan on 3+ drives. This is for a home PC where graphic design (large 600+MB files) and gaming (many small files) are used daily. I like the 16k cluster size, it hits the sweet spot for me.

[alsoada]

This is so dead but got in on a search and I thought maybe I could bring up the dead. :mrgreen:

I have a 80Gb SATA and I want more space, or mainly to put Windows alone in a new smaller drive and everything else in the old 80Gb one with out having to pay too much money for a new drive.

I have an old and busted 20Gb Maxtor EIDE HD that I wanted to see if I could get fixed up or maybe simply buy a new small SATA. I mainly use the pc for internet brosing, some small picture editing and gaming. any thoughts on this? Would I lose a lot of performance runing my OS on a EIDE? would a SATA allow it?

thanks

[jbennet]

You will only experience a performance increase in circumstances where you are accessing data on the drive. Games will not benefit other than with quicker load times. Applications such as rendering will only benefit if the data files being worked on (and the resultant calculations) are too large to be held in system memory.

ive found battlefield 2 runs faster with sata as part of its anticheat system is to find out the hash of every file in its install dir. Also it makes reading from swap faster