Mushkin 2GB XP2-8000 Redline Test Report

[Eldonko]

Mushkin 2GB XP2-8000 Redline Test Results

Hey guys, this is still a work in progress so keep an eye for updates.

This RAM has been around a while but I havent saw a whole lot of tests for it for some reason. This stuff uses highly sought after D9GMH B6-3 chips and in my testing proves to be some of the top clockers currently on the market.

That being said, here is your Mushkin XP2-8000 Redline (part 996525) test report.

First I’ll start with a few words on the test platform and photos of the RAM and board. I’m using an E6400 Allendale setup with a s965p DQ6. The XP2-8000 usually has the red spreaders like other Redline, but this set got the black. I like the black better personally.




Test bed
Gigabyte GA-965P-DQ6 (unmodded)
E6400
MSI GeForce 7800GTX @ stock
Mushkin XP-650
Windows XP, untweaked
CPU Cooling - Custom H20 and stock cooler on the NB







Testing methodology:
I wanted to test this mem with a stock DQ6 to show the potential without volt mods. Vdimm on this board goes to 2.37v set in the BIOS, higher than that the board will not boot and vmod is necessary. I plan of doing the vdimm mod soon, when I get some time and will post tests with more volts.

I did a cpu-z validation in most cases, check the validation # on the RAM speed CPU-Z tab if you enjoy checking that type of stuff.

Benches used:
3dmark 2001 and 2006
Super PI v1.5 (two instances)
Orthos
Sandra 07

Tools:
MemSet
EasyTune 5
CoreTemp

Most of the tests are divided by voltage and timings, basically I tested 4-4-3, 3-4-3, and 3-3-3 at 2.3 and 2.37v. For my cas3 tests, I set cas3 in the BIOS and memset, sandra and cpuz all confirm cas3. People been saying cas3 is sometimes screwy, maybe it is who knows. However, I can run 3-4-3 @ 550Mhz no problem and 3-4-3 is slower running dual 32M but its faster running single 32M, so it might be working but I cant tell for sure

Now for some test results..

[Eldonko]

3-3-3 – 2.3v (bios) – 410 Mhz

Dual 32M



3DMark 2001 and 2006



Sandra



Orthos




3-3-3 – 2.37v (bios) – 422Mhz

Dual 32M



3DMark 2001 and 2006



Orthos




3-4-3 – 2.30v (bios) – 535Mhz

Dual 32M



3DMark 2006



Dual P95





3-4-3 – 2.37v (bios) – 550Mhz

Dual 32M



3DMark 2001 & 2006



Sandra 07



Orthos

[Eldonko]

4-4-3 – 2.30v (bios) – 538Mhz

Dual 32M



3DMark 2001 & 2006




4-4-3 – 2.37v (bios) – 553Mhz

Dual 32M



3DMark 2001 & 2006




4-5-4 – 2.30v (bios) – xxxMhz

Testing in progress

[Eldonko]

Overclocking Summary:
2.3v
3-3-3 - 410Mhz
3-4-3 - 535Mhz*
4-4-3 - 538Mhz

2.37v
3-3-3 - 422Mhz
3-4-3 - 550Mhz*
4-4-3 - 553Mhz

*Unable to confirm that cas3 working properly, investigating further now.

[jmax_oc]

I think that your 3-4-3-12 is a 4-4-3-12.
Never seen suck resultas at 2.4V...

If you make 550MHz 3-4-3-x, you can easily do more than 600MHz 4-4-3-x.


Despite my remark, your test is very clean

[Eldonko]

Yeah you are probably right, I figured that when 3-4-3 would clock so well. No idea how I can use "real" cas3 though..

[Kunaak]

Unfortunatly Gigabyte 965 boards are among the worst there is for reviewing ram with, cause there are numerous ways to get very misleading results.

with all that we seen from the Gigabyte 965 boards, we put a self imposed ban on using them in future reviews on the frontpage to avoid any possible misleading info that can come up in a review.

my only suggestion for you, when using Gigabyte 965 boards, is to use Memset, and manually set cas 3 or whatever your trying.

unfortunatly, Gigabyte boards are severly good at overclocking, and just as good at giving misleading results when reviewing/testing ram.

I have found very few ways to eliminate some of these problems.
like running 4:5 usually will report true cas, while 1:1 will often report crazy cas.... like my 5400 UL has booted a few times at cas 4-2-2-9 on my DS9... at almost 1.1 ghz...

[Eldonko]

Can you be more specific on how giga boards show misleading results for RAM Kunaak. If you read ^^ I stated that I dont think cas3 is working properly for 3-4-3, what are the other problems?

[Kabauterman]

this aren't no REDLINES the REDLINE are Red
y your RAMs have Black HS?

[Duonger]

this aren't no REDLINES the REDLINE are Red
y your RAMs have Black HS?

we ran out of redlines heatsinks. so we are shipping them with black till the reds come in.

[fhpchris]

Be happy they didnt send you Purple, they have done it before

[DeToNaToR.cl]

Really nice sticks

[95blackz26]

my set is right on par with the results posted above. ive had them for a little while. pretty good memory if you ask me.

i have a p5b deluxe with core2duo 6400

[charlie]

WOW!!

4-4-3 @ 2.4v/550mHz is just amazing... AMAZING!!

[sierra_bound]

Actually 550 @ 4-4-3 and that voltage is not that hard with the better RAM out there.

When I saw 550 @ 3-4-3, my eyes nearly popped out. Even if it's a bug, still looks like decent memory.

Thanks for doing the tests, Eldonko.

[Eldonko]

The vdimm measures low too, so 2.37 is more like 2.31-2.33 DMM. Although 3-4-3 doesnt do much over 4-4-3 (see comparo below). It's just the platform though, there would be much more of a performance increase because of tight timings for AM2. C2D is all about speed, gunna take some getting used to heh heh.


CAS 4



CAS 3

[steelballzz]

nice ram u got there....

[[XC] Fifty]

Great test thank you. I have 4 gigs of this and can not wait to test it out.

[Eldonko]

Thanks GibbonThree, hope ya get some good sticks. I am just about finished reviewing xp2-8500 as well (the latest release from Mushkin) and will post shortly.