PDA

View Full Version : A few simple questions (i think simple.)


Lee
December 3rd, 2005, 06:15 PM
1) What's the difference between ATA and SATA?

2) What exactly is RAID?

3) How noticeable are different ram timings?

4) On a scale of, say, 1-10, how good is my ram? (see attached pics)

Mafia Leader
December 3rd, 2005, 07:04 PM
1. SATA is faster than ATA, uses smaller wires, few other things I can't remember at this state of drowsiness.

2. RAID is using two or more physical drives and some kickin' rad hardware to fake the separate drives being one drive. There are various methods to RAID, some of which are scarier than others, but all do the same thing.

3. Depends how sensitive and elitist you are.

4. Personally, I'd say about 5, but that's only because it seems to be so horrifically generic.

Agent Law
December 3rd, 2005, 07:35 PM
RAM timing will make a big difference when you're really pushing your rig. Otherwise, it's not something the average gamer should be too concerned about.

siddy
December 3rd, 2005, 10:57 PM
1) What's the difference between ATA and SATA?

2) What exactly is RAID?

3) How noticeable are different ram timings?

4) On a scale of, say, 1-10, how good is my ram? (see attached pics)


1) Different connectors. ATA runs at about 100mb/s SATA runs at 150, SATA2 at 300mb/s. However, these are theoretical maximums and you'd never see that limit at all

2) RAID is a Redundant Array of Inexpensive Drives. There are many different types of RAID, RAID 0, 1, 0+1, etc, etc. They work differently but basically it's using two hard drives for different reasons.

RAID 0 is for speed...you use two hard drives to spread out information between the two...both hard drives are used at the same time to find that information, so techincally having 100mb/s x2.

RAID 1 is for backing up info. The 2nd hard drive copies everything on the first, so if one HD goes, all the info is there on the 2nd, so you don't lose everything.

3) RAM timings are most noticable in gaming, and benchmarking. For the typical user, timings won't make the biggest difference in the world, but it will be noticable.

4) On a scale of 1-10, 10 being the lowest, 1 the highest, i'd put it at a 7 or so. I'm not sure of the entire specs of your system, but you're probably running your RAM asynchronious to your processor, so your computer isn't running as well as it could.

StandingCow
December 3rd, 2005, 11:16 PM
how bout mine?

Lee
December 4th, 2005, 12:13 AM
4) On a scale of 1-10, 10 being the lowest, 1 the highest, i'd put it at a 7 or so. I'm not sure of the entire specs of your system, but you're probably running your RAM asynchronious to your processor, so your computer isn't running as well as it could.

could you elaborate as to how i could get it running better?

siddy
December 4th, 2005, 12:28 PM
how bout mine?

not quite sure myself, as you have 2gb rather than 1, so your timings are a bit looser, but i think that you have something decent there cow.

could you elaborate as to how i could get it running better?

well, not knowing your system, I'll give you a broad outline.

You want your RAM to run at the same frequency as your processor, so theoretically there is delay for the two communicating between eachother.

Athlon 64/Semperon systems need PC3200 (DDR400) speeds to communicate synchronously to eachother.

Unfortunately tho, I don't know ALL too much about RAM, and I'm sure there are some great tutorials on the internet.

But really, if your timings are running asynchronously with Athlon 64/Semperon systems, you're only looking at a performance loss of about 5% or so, due to the processor's integrated memory controller.

Lee
December 4th, 2005, 04:50 PM
gotcha.

Paluch
December 5th, 2005, 03:15 AM
Cow dont tell me you have 4 512mb sticks? Becuase if you do you shouldnt be worried about your high timings. You should be worried about the loss of bandwidth.

The optimal set up on all DDR P4 and A64 rigs is either 2x512 or 2x1024. The problem with having 4 sticks is that the chipsets/memory controllers have to sync all 4 sticks which causes performance to be less than 2 sticks. The result is higher timmings and lower bandwith.

Paluch
December 5th, 2005, 03:22 AM
Lee your timings are good. Although the frequency that they are running at isnt.

Even if you set the 4:5 divider instead of the 1:1 you wont see much of a difference. I wouldnt change a thing

siddy
December 5th, 2005, 01:10 PM
Cow dont tell me you have 4 512mb sticks? Becuase if you do you shouldnt be worried about your high timings. You should be worried about the loss of bandwidth.

The optimal set up on all DDR P4 and A64 rigs is either 2x512 or 2x1024. The problem with having 4 sticks is that the chipsets/memory controllers have to sync all 4 sticks which causes performance to be less than 2 sticks. The result is higher timmings and lower bandwith.

good point. Also (with A64 systems) you can run RAM at 1T timings, which usually yeilds an extra 5-7% system boost. With 4 sticks, 1T is impossible.

StandingCow
December 5th, 2005, 01:57 PM
Cow dont tell me you have 4 512mb sticks? Becuase if you do you shouldnt be worried about your high timings. You should be worried about the loss of bandwidth.

The optimal set up on all DDR P4 and A64 rigs is either 2x512 or 2x1024. The problem with having 4 sticks is that the chipsets/memory controllers have to sync all 4 sticks which causes performance to be less than 2 sticks. The result is higher timmings and lower bandwith.

Nope, its 2 1 gig sticks.

The loss in performance is very slight even if I did have the 4 512 sticks (I read all about it on toms hardware) and the only time you notice it is in benchmarks.