OCZ ddr booster worth it?

I'm running a gig of Corsair XMS PC3200 rev1.2, which I believe contain winbond CH-6 modules, in the system shown in my sig. Currently, I'm limited by ram and can barely get above 220 fsb (maybe 221-223) stable. From what I've heard, CH-6 doesn't respond to voltage until you get it over 3.0 volts (my motherboard only allows 2.9v or lower). So, I have a few questions. With, let's say, around 3.2 volts, could I actually get something out of my ram? Second, how well does the ddr booster work? And last, would it be worth it considering that I would have to give up dual channel to use it?

you have the nf7s rev2 right? how come you would have to give up dual channel? it says on the DDR booster compatibility page that you should use DIMM slot 3, which means you could use slot 1 and 2 for dual.. not that dual channel makes any difference on athlon xps correct me if im wrong though

Careface*

Careface said:

you have the nf7s rev2 right? how come you would have to give up dual channel? it says on the DDR booster compatibility page that you should use DIMM slot 3, which means you could use slot 1 and 2 for dual.. not that dual channel makes any difference on athlon xps correct me if im wrong though

Careface*

Click to expand...

Dual channel only works when the ram sticks are in dimm 1 and 3 or dimm 2 and 3.

You could mod the booster a bit (thus voiding your warrenty, which in my experience you will need) to fit in slot one, or you could mod the board to move the capacitor out of the way.

But how good is it anyway..?
And would it fit my mobo and still let me use dual channel..?

caincha said:

And would it fit my mobo and still let me use dual channel..?

Click to expand...

Only if you mod the board or the booster (if it's an NF7-S).

caincha said:

But how good is it anyway..?

Click to expand...

It's the only plug-n-play way of overvolting your RAM; ideal for those with BH5, BH6, etc.

Your CH-6 should continue to see gains right up to ~3.8V or so.

That said, if I were in your situation, I would not buy a Booster.

~ A stable 220 MHz FSB isn't shabby on an nForce2 board.
~ Most nForce2 chipsets won't clear ~240 MHz with tight timings, maintaining 3D stability. With tight timings, the AGP controllers commonly tend to go AWOL around there, in my experience.
~ As such, there is a chance that you'll buy a Booster, then find that your board caps at ~225 MHz with tight timings
~ The Abit NF7-S has a VTT (A DDR voltage referenced from VDIMM - it is supposed to be exactly 1/2 of VDIMM) tracking problem; the boards typically do not track VTT properly past ~3.1V or so (VTT won't increase past ~1.55V). As such, to use your booster to its full effect (3.3V-3.5V is probably what you'll be wanting), you will need to voltmod your motherboard.
~ Brings me back to my first point... you buy the booster, voltmod the board, and then discover that your chipset won't clear ~225 MHz with tight timings .
~ There is no guarantee that a Booster will net you any speed increase whatsoever for your ~$40.
~ With the amount of work involved in getting a booster to work properly with your NF7-S, you would save a lot of money and have a lot more fun just doing VDIMM and VTT voltmods to your board, and running an overvolted +3.3VDC rail .


Then again, more than a few NF7-S boards scale FSB right into the mid 250s with full stability (although rarely with tight timings). It's partially luck of the draw. Your RAM definitely has at least 240 Mhz in it however.

Speaking personally, my old NF7-S would not clear a 228 MHz FSB with 3D stability and tight timings, no matter what I did to it. It scaled right to a ~255 MHz FSB for anything but 3D .

Once again speaking from personal opinion here, but for VDIMM overvoltage below ~3.5V, I think that voltmods are far more economical, more educational, and more fun than buying a Booster. For ~3.5V+, a +3.3VDC rail overvolt loses it's practicality.


Just my ten cents .

JoT said:

Only if you mod the board or the booster (if it's an NF7-S).


It's the only plug-n-play way of overvolting your RAM; ideal for those with BH5, BH6, etc.

Click to expand...

Mine is as Abit IC7G as in my sig.
And I asked 'cause I read a review where it said the max you can expect from that booster is stable voltage, but it does not overvolt anything at all...
Anyone ever used that to confirm or deny this?

caincha said:

Mine is as Abit IC7G as in my sig.
And I asked 'cause I read a review where it said the max you can expect from that booster is stable voltage, but it does not overvolt anything at all...
Anyone ever used that to confirm or deny this?

Click to expand...

that's total crap. It will give you stable voltage if the motherboard will track the vtt properly. If you set the booster to 3.5v then that's exactly what you'll get. It does overvolt as it takes the power right from the psu.

Mine is as Abit IC7G as in my sig.
And I asked 'cause I read a review where it said the max you can expect from that booster is stable voltage, but it does not overvolt anything at all...
Anyone ever used that to confirm or deny this?

Click to expand...

In my testing of the device with my DFI nForce4 board, I found that it gives a very stable VDIMM untill around ~3.6V, after which point you'll see a larger and larger voltage droop between VDIMM at idle and VDIMM at load. Most people don't need more than 3.6V however.

The Booster overvolts your VDIMM directly from the +5VDC rail, it works quite well within a certain voltage range. Past a certain point though, the droop gets bad enough that it's sketchy to use. I found that to get 3.8V of VDIMM under memtest86 load (measured with a multimeter) with a Booster, I needed to set it to 4V at post (also measured with a multimeter) - I would not reccomend using a Booster for really high VDIMM overvolts. The overvolt is adjustable through a built-in potentiometer, so you can set the Booster to provide whatever VDIMM you desire - in that respect, it won't overvolt VDIMM unless you want to.

Well don't blame me...
Here's the review I mentioned: http://www.hardcoreware.net/reviews/review-243-1.htm

Well no one seems to have said this, so i'll step in.
Your RAM is most likly not holding you back, or is not the only thing

the NF-7 wont go over 220FSB +/- 5 MHz, with a 133FSB CPU.
You will need to pin MOD the CPU to a 166FSB Chip. and then you will be ok.
Want proff?
My 2500+ Desktop (11x166) had no problems hitting 245FSB stable. Could post at just over 250FSB but i think the RAM i had at the time was my issue. And this is with no Chipset mods for voltage

Then I got a moble 2500+ and 117FSb is my new wall.
Pin mod the CPU to a 166FSB Chip and no problems with the FSB going through the roof again.

jAY said:

Well no one seems to have said this, so i'll step in.
Your RAM is most likly not holding you back, or is not the only thing

the NF-7 wont go over 220FSB +/- 5 MHz, with a 133FSB CPU.
You will need to pin MOD the CPU to a 166FSB Chip. and then you will be ok.
Want proff?
My 2500+ Desktop (11x166) had no problems hitting 245FSB stable. Could post at just over 250FSB but i think the RAM i had at the time was my issue. And this is with no Chipset mods for voltage

Then I got a moble 2500+ and 117FSb is my new wall.
Pin mod the CPU to a 166FSB Chip and no problems with the FSB going through the roof again.

Click to expand...

Yes, I've done the L12 already. Which is why I believe my ram is holding me back.