Question for RavenSEAL

[Malak]

You should PM him if you have a question for him.

 

[OCGuy]

I have figured out what ATOT's Xmas present is this year: Multiple threads full of fail.

Enjoy!

 

[Mike Gayner]

lol

 

[alfa147x]

Callout thread!
Ban!

 

[TridenT]

callout.

permaban NOW

 

[Fayd]

I'm not calling him out. I specifically asked him why he chose that set up. 5GB seems like a very deliberate number so he had to pick it for a reason right? I'm the newbie. I don't really have a clue.

could be 2x2gb and 2x512mb.

 

[Dumac]

could be 2x2gb and 2x512mb.

I was going to guess this.

He probably used 2x512mb back in the day, and eventually got 2x2gb

 

[RavenSEAL]

LOL, i just randomly found this thread after google searching my nickname

The answer is, I had an extra stick left over lol. So i thought i might as well use it, cost me $30 after all, can't hurt right?

I have 1x 2GB and 3x 1 GB as of right now. 2 of the 1GB modules are paired.

 

[Ichinisan]

..., can't hurt right?

Wrong. You're no longer running in dual-channel mode unless your mobo supports some fancy asynchronous configuration.

 

[MagnusTheBrewer]

Wrong. You're no longer running in dual-channel mode unless your mobo supports some fancy asynchronous configuration.

Which is worthless unless you are benchmarking.

 

[Ichinisan]

Which is worthless unless you are benchmarking.

Almost anything that's CPU-limited should be affected by memory bandwidth...even between high-speed caches on the CPU die. When encoding video, for example, most of the CPU cycles are wasted; waiting for more data from the system memory. This will be true until the interface to system memory matches the core clock of your CPU (not happening with any architecture we can conceive at this time). If I understand correctly, dual-channel gives double bandwidth...which is very significant.

 

[RavenSEAL]

Hey, unless ya'll wanna send me a check for $30...I really don't think it matters, my games play just fine

 

[MagnusTheBrewer]

Almost anything that's CPU-limited should be affected by memory bandwidth...even between high-speed caches on the CPU die. This will be true until the interface to system memory matches the core clock of your CPU.

Affected yes, noticeable in real world applications? Nope.

 

[BoomerD]

Just delete the thread. I didn't mean to make a fuss!

Damn...if you're gonna hang around here, you need to learn to stick up for yourself and tell the assholes to shut the fuck up...

MOST OF US here aren't too bad...once we get to know you...
BTW, howzabout some more pics? :biggrin:

 

[olds]

Damn...if you're gonna hang around here, you need to learn to stick up for yourself and tell the assholes to shut the fuck up...

MOST OF US here aren't too bad...once we get to know you...
BTW, howzabout some more pics? :biggrin:

I'd hit it too.

 

[guyver01]

yanno what's funny?

Ricky happens to be ZaneJohnson's name too...

could this be ZJ's 'dual' account?

 

[MagnusTheBrewer]

If I understand correctly, dual-channel gives double bandwidth...which is significant.

There are many sub systems with greater bandwidth than can be currently utilized. My favorite lately is SATA 3.0 which is marketed as a must have and adds exactly zero performance gains over SATA 1.5

 

[Ichinisan]

[moved to next page]

 

[BoomerD]

I'd hit it too.

heh-heh...but we're old enough to know it's ALL hittable... I don't know about you, but the WORST I ever had was OUTSTANDING!

 

[olds]

heh-heh...but we're old enough to know it's ALL hittable... I don't know about you, but the WORST I ever had was OUTSTANDING!

:biggrin:

 

[Ichinisan]

Affected yes, noticeable in real world applications? Nope.

I had edited that post.

Video encoding is sometimes a real-world application, no?

There are many sub systems with greater bandwidth than can be currently utilized. My favorite lately is SATA 3.0 which is marketed as a must have and adds exactly zero performance gains over SATA 1.5

This is the opposite situation. The bandwidth between system memory and the CPU is a very limiting factor for CPU performance..the most limiting factor, in fact. Any less than the CPU's FSB / width is significant lost performance. Of cource, most people never push the CPU anyway.

 

[MagnusTheBrewer]

I had edited that post.

Video encoding is sometimes a real-world application, no?

The problem is architecture, CPU and file size play a MUCH greater role in video encoding than dual channel memory. Which is why, aside from folks who like to benchmark, there currently is no real world advantage.

 

[lothar]

Which is worthless unless you are benchmarking.

Running 5GB over 4GB seems just as worthless.
I would rather run 4GB dual channel than 5GB single channel.

 

[RavenSEAL]

I had edited that post.

Video encoding is sometimes a real-world application, no?

I don't encode videos lol

 

[Ichinisan]

If you have any significant amount of video to encode, and you can cut the speed of an encoding job nearly in half, I think it's significant (no matter which CPU you're using).