[Overclockers.com]

Patriot Sector 5 DDR3-2400 4GB Memory Kit Review
by Jeremy Vaughan (hokiealumnus)

Quite a few memory manufacturers have taken advantage of the very strong IMCs (integrated memory controllers) on Intel's i7 CPUs. Patriot has thrown its hat in the ring with kits as fast as DDR3-2500. Today we're going to be digging in to their second-highest clocked kit coming in at DDR3-2400 with timings of 9-11-9-27, model # PVV34G2400C9K.

... Return to the article to continue reading.

Digg | Reddit

Discuss this article below and please click on the social media links above to support our content. If you are interested in contributing to the front page (www.overclockers.com), please feel free to contact splat, mdcomp, or hokiealumnus. For the latest updates, follow Overclockers.com on Twitter (@Overclockerscom).

[redduc900]

Nice review as usual Jeremy, as was your review of the G.Skill PiS DDR3-2400. Just curious as to what all of the final voltages were at DDR3-2562; 213 BCLK (PLL, VTT, etc.)?

[hokiealumnus]

Thank you very much! Final voltages in BIOS for DDR3-2562 were:

PLL - 2.0V
VTT - 1.45V (to help with bclk, just in case)
vDIMM - 1.72V

vCore was bumped somewhere in the vicinity of 1.4V just to keep from getting instability while overclocking the CPU. It's much more than ample for a SP1M run at those speeds. Crashes were from the RAM and nothing else, guaranteed.

[redduc900]

These are my current BIOS settings using the G.Skill PiS DDR3-2400 at a 215 BCLK x 18, DDR3-2580; MCH Strap: 1600...

PCH Voltage: 1.05V
CPU PLL Voltage: 1.80V
CPU VTT Voltage: 1.425V
DRAM Voltage: 1.68V

CPU PWM: 849
VTT PWM: 634
DDR PWM: 240

Vcore set to 1.40V w/ Vdroop enabled = 1.32-1.33V under load; 1.38V no load.

[hokiealumnus]

Huh; looks like it takes a bit of vDIMM (.04V) to get those few extra MHz. The only change to the above ...

Originally Posted by hokiealumnus

Thank you very much! Final voltages in BIOS for DDR3-2562 were:

PLL - 2.0V
VTT - 1.45V (to help with bclk, just in case)
vDIMM - 1.72V

vCore was bumped somewhere in the vicinity of 1.4V just to keep from getting instability while overclocking the CPU. It's much more than ample for a SP1M run at those speeds. Crashes were from the RAM and nothing else, guaranteed.

...for the G.Skill kit for DDR3-2626 was that PLL was at default 1.8V like yours. All else was the same. I also didn't mention it, but PCH has always been at default 1.05V. I've never seen any help from increasing it.

Nice system you have there, by the way.

[MattNo5ss]

Great job as always

I think I found a typo though:

These sticks were set to their rated speed using the 10x multiplier at 200BCLK with timings 9-11-9-27.

Shouldn't it be the 12x multiplier

[hokiealumnus]

Aye, you did find a typo and it's fixed now. Thank you.

[MIAHALLEN]

Nice job Jeremy

Would you be willing to try the stability test again with stock PLL to see if it actually made a difference?

[hokiealumnus]

Thanks! Sure, no problem. It might be tomorrow but I'll give it a re-run and report back.

[hokiealumnus]

Well allllrighty then. No need to run the R.S.T. Pro 3 on that setting!

I started the test. Five minutes into it (swarm test, which isn't too strenuous and can pass at unstable clocks), my wife comes up to use the computer. So I restart back into Windows....except I don't. Stuck on the screen with the little circle saying "Welcome". Froze up and required a hard restart. Thus, I guarantee it won't pass the R.S.T. test.

Looks like these need every single setting set to be stable!

[Brolloks]

Jeremy, you really want to run SPI 32M to test for RAM stability, it is the best recognized, and readily available 2D benchmark for such.

[hokiealumnus]

Completely agreed, but trust me when I say the R.S.T. Pro 3 is better. It just hadn't made it to the part of the test where it ferrets out the applicable instability. The Swarm test is five minutes into about an hour long test and isn't the hardest one by any stretch.

[Brolloks]

Fore sure the RST is better, no doubt...unfortunately it is not known by most users out there so they have no reference

[hokiealumnus]

I hear you. If you're saying it should be run in addition to the RST, it already is. All stock benches are run three times - including SP32M.

[Brolloks]

I meant running SPI 32M on your OC settings, last section.

[hokiealumnus]

Oh, gotcha. Well, I'm not fishing out stability...just seeing how high it can run the same benches the other kits could. SP32M is definitely tougher and maybe should have made it previously. I'm open to adding it but no longer have two of the kits to re-run it on (both are allocated to others' systems for reviewing other components and physically not here).

It's definitely safe to say SP32M would never complete in any of the reviews for the high MHz test. It might have for the low CL at reduced Mhz.

[Brolloks]

From a bencher and also a daily users perspective if RAM passes SPI 1 M it meants really nothing, it will probably crash when opening a game or next time you boot up, so RAM OC only starts to count when it passes at least SPI 32M.

So what I'm trying to say for a RAM review SPi 32M is the hurdle to pass otherwise your OC does not really carry any weight, hope this helps and do not see it as negative commentary.

[hokiealumnus]

Not at all, I understand what you're saying and know from experience SP1M is not to be used for stability. I think we are just looking at it from two different perspectives. I completely understand SP1M takes almost nothing to pass, which was sort-of the point - to see the absolute maximum MHz attainable to run a simple bench at the given settings.

The maximum you can reach for SP32M is of course going to be somewhere between stock and the OC I obtained. I'm not trying to carry weight, but see the maximum the sticks are capable of. Put some cold on them and throw some more voltage and at that frequency, they will probably pass SP32M. Definitely not at the settings I'm working from and definitely not on air. Nowhere do I indicate nor am I trying to pretend this is stable. Stability wasn't the point.

I definitely see where you're coming from though and will strongly consider adding it in the future.

[MIAHALLEN]

Originally Posted by hokiealumnus

Well allllrighty then. No need to run the R.S.T. Pro 3 on that setting!

I started the test. Five minutes into it (swarm test, which isn't too strenuous and can pass at unstable clocks), my wife comes up to use the computer. So I restart back into Windows....except I don't. Stuck on the screen with the little circle saying "Welcome". Froze up and required a hard restart. Thus, I guarantee it won't pass the R.S.T. test.

Looks like these need every single setting set to be stable!

Thanks Jeremy

These GTX3 DIMMs I have can't do squat with my current CPUs
Waiting for a new Lynnfield & Gulftown to arrive to push them a bit more