Patriot Sector 5 DDR3-2400 4GB Memory Kit Review

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.

Before moving on with this review though, Reefa_Madness brought to our attention this wonderful mea culpa on DDR3 RAM from Anandtech. It’s a great read and I would highly recommend taking the time if you have any interest in how RAM really works.

Packaging, Specifications & First Impressions

This kit comes in a nice looking, red & black box surrounding the sticks that are encased in snug fitting plastic. The presentation is nice, as memory kits go. Certainly better to look at than many kits I’ve seen.

There they are! One thing separating these from some kits is that the PCB is black in addition to the heat spreaders, which is a nice aesthetic touch.

The heatspreaders on these aren’t as tall as the G.Skill Pi series kit we’ve already looked at and it’s nowhere near the Kingston HyperX kit. Based on that alone, these would be a better choice if you needed lower profile RAM for things like heatsink or fan clearance issues.

As you can see on the labeling, this kit is rated at DDR3-2400 (PC3-19200) with timings of 9-11-9-27. That tRAS (27) is one tighter than the referenced G.Skills. One last photo of them and we’re done.

All in all a solid package in appearance and specifications. Enough gandering around though. It’s time to take these things to task!

Stability Testing

Before performance testing, it’s always good to look at stability. Ultra-X has generously supplied their RAM stress tester, the R.S.T. Pro 3 PCI Express. There are only a couple review sites that have one of these and we are proud to be one of them. It’s not your standard Memtest++ bootable CD. This is a standalone, bootable piece of hardware that plugs into a PCIe x1 slot and, man, does it ever beat on some RAM. By the time the tests finish, the sticks are quite toasty.

Several manufacturers use this to test their memory before it goes out the door to reduce RMAs. If it passes this test, it is stable and then some. Manufacturers generally run the full memory test for three passes. We’re overclockers though and tend to abuse sticks a little more than other people, so for our purposes, doubling the manufacturer’s requirement to six passes sounds good.

These sticks were set to their rated speed using the 12x multiplier at 200BCLK with timings 9-11-9-27. vDIMM was 1.65V as specified and VTT was set at 1.4V. This is where things got a bit complicated for this Patriot kit. Prior to sticking the R.S.T. in there I always run Intel Burn Test (IBT). Unfortunately, they wouldn’t pass.

That’s not the end of the line though. Sets rated at speeds like these may need a little finesse. The next step is to look at the XMP profile and try those settings. EVGA’s sub-timings aren’t as extensive as some (ASUS, for one), so I just set what was available and hoped for the best. To check out the XMP settings, just pop open CPU Tweaker and click “SPD”.

Inserted as best you can into this BIOS translates to this.

They passed IBT no problem at these settings. Unfortunately after putting the R.S.T. in there, the XMP settings didn’t pan out so well.

Back to the drawing board, as it were. While speaking with Patriot when trying to straighten out the first set of RAM (which turned out to be defective; testing you’re seeing here is on a replacement set), they pointed out a settings section of their forums, which is a great customer service and invaluable if you’re having trouble stabilizing your RAM. Their screenshots are from an ASUS board, so again, we set what we can with fingers crossed – with the exception of B2B CAS Delay, which causes funky issues on this board (in multiple BIOSes as well). I also set vPLL to 2.0 V as recommended.

Let’s boot back to the R.S.T. Pro 3 and see what we get.

Whew, that’s a relief. Stable and happy through all six passes.

Performance at Rated Speed

To establish a baseline for performance all tests were run on a 4GB set of G.Skill Trident DDR3-2000 with timings 9-9-9-24. Also included in the results are the G.Skill Pi DDR3-2400 and Kingston HyperX DDR3-2000 kits reviewed earlier. Every benchmark was run three times and the best score is recorded below.

Obvious anomalies, if they existed, were discarded and the benchmark was run again. The only tests that this happened on were the Everest read/write/copy tests, where they’re known to glitch about 10,000MB/s over every other run. This only happens rarely (in fact it hasn’t happened since the first kit was tested), but it does happen.

All tests were run with the CPU at 4.0 GHz (200 BCLK x 20) and RAM set to its stable speed and timings. The rest of the system in question consists of:

• CPU: Intel i7 870, kindly supplied by Intel.
• Motherboard: EVGA P55 FTW
• GPU: HIS HD4890 Turbo
• PSU: Corsair TX650
• OS for stock testing: Windows 7 Professional x64.
• OS for overclocking: Vista Enterprise x86.

3D Performance

Like the previous Kingston review, we’re going to stick with the most recent 3D benchmarks from Futuremark – 3DMark06 and 3DMark Vantage- because they showed the largest differences in testing. Note these graphs do not start at zero, they both have a range of 5,000 points in order to accentuate the difference in scores.

Ouch. 3DMark06 really likes tighter timings to MHz, especially when you get as loose as the stable clocks required for this kit. To keep things in perspective, the difference between the highest and lowest scores, while totaling 611 points is still only -2.9%.

3DMark Vantage again shows a preference for tighter timings. The difference is less substantial in this case, at 216 points & -1.6%, but goes to show raw MHz isn’t everything when it comes to 3D benching.

Rendering & Compression Performance

For some “real world” testing we’ll look at Cinebench R10 64-bit & Cinebench R11.5 64-bit to test graphics rendering and WinRAR for compression performance.

Cinebench R10

Cinebench R11.5

Honestly, I can’t figure this bench out. Neither of them seem to give a rat’s rear end about timings or MHz. Differences in scoring seem to have more to do with the individual run than anything having to do with the RAM. Looks like it might be time to start looking for a new rendering bench for memory reviews.

WinRAR

Compressing files shows a substantive difference between the DDR3-2400 kits. With its loosened timings, this kit acts more like the baseline DDR3-2000 (which operated at 9-9-9-24) than the high MHz beast it should be.

Synthetic Number Crunching Performance

Here we’ll have a look at how each kit performed when running benchmarker favorites SuperPi 1M, SuperPi 32M, WPrime32M & WPrime 1024M.

SuperPi 1M

SuperPi 32M

SuperPi 1M lost a little bit of ground, even from the base kit. 32M performed better than the base, but not as good as the other DDR3-2400 kit. While the percentage difference is relatively small, almost five seconds is a large difference when you’re at the top of the SuperPi 32M rankings on hwbot.

WPrime 32M

WPrime 1024M

As in the other reviews, WPrime32M seems to prefer timings to MHz. WPrime 1024M still doesn’t seem to care all that much. You’re talking about a test that runs for over three minutes and best-to-worst difference is a mere .164 seconds. Timings, MHz…WPrime 1024M doesn’t care!

Synthetic Bandwidth and Latency Results

Everest & MaxxMEM both are great for highlighting differences between memory kits. Even if differences shown here aren’t massive in every day use, it’s valuable to see what they are for evaluation purposes. As we saw with the WinRAR testing, there is definitely a difference to be had and it’s nice to know where that’s coming from.

Everest

Everest Read

Everest Write

Everest Copy

Everest Latency

While not quite as strong as the G.Skill 2400 kit with the tighter sub-timings, Patriot still packs a punch going through the everest suite.

MaxxMEM

MaxxMEM Read

MaxxMEM Write

MaxxMEM Copy

MaxxMEM Latency

MaxxMEM continues to show similar gains & losses compared to Everest, but on a slightly smaller scale. Writes were surprising, as was latency – both of which lost a little ground to the baseline kit.

Of course, some people like to see the Everest quick bench, so here they are side by side.

Overclocking

As usual, we’ll seek out two maximums for RAM overclocking – tightest timings at one-multiplier-below-default and voltage of 1.68 V and highest MHz at default timings at 1.72 V. Benches for the former include SuperPi1M, WPrime 32M & MaxxMEM. One multiplier below default puts us at DDR3-2000. The G.Skill Pi kit managed 6-9-8-24. With the stability challenges above, expectations were low. Let’s see how they did.

Not only did they match the G.Skill kit, they did it one better by taking tRP down one notch for overall timings of 6-9-7-24! No tRFC tricks here either folks; both were set at 88. This was quite the pleasant surprise.

With stronger-than-expected performance exploring timings, there may be some hope yet for MHz! For this next test, we’re looking for the absolute maximum MHz at default stable timings when running SuperPi 1M.

Well then. Again, a better than expected result at 1281Mhz / DDR3-2562 (validation). It’s definitely not as strong as the G.Skill kit, which hit DDR3-2626 (at tighter sub-timings), but not bad by any stretch. While in the BIOS the next time, I figured it couldn’t hurt to find out what happened if the sub-timings were set at the same levels.

Now we’re cooking! Same MHz, tighter timings. Neither would go one MHz more but with a few hundredths of a second shaved off at no expense, there are nothing but good things here!

Final Thoughts & Conclusions

This kit turned out to be somewhat of an enigma. While it takes almost an act of god to stabilize at stock settings, they actually overclock decently. Certainly high enough to do anything you would in many overclocking situations. Even more impressive than the MHz reached were the timings. DDR3-2000 at CL 6 is nothing to sneeze at.

Sure, it’s a bit of a pain at first for 24/7 operation, but with the patronage Patriot provides in the form of easy-to-use screenshots it’s just a matter of using a few extra settings. The G.Skill kit obviously uses somewhat stronger ICs in their offering; stability was effortless and just a matter of setting the four rated timings. They also clock better and come with a fan. But, and it’s a big but (teehee), there is one absolutely critical factor when rating these two against each other – price.

Right now at Newegg the G.Skill DDR3-2400 kit is selling for $224.99. They’re worth it to some, but that’s a steep premium. This Patriot kit performs very close to the level of that one and it turns out they are selling for a mere $159.99 after rebate. At that price, it’s difficult to argue against Patriot’s offering. That’s what some DDR3-2000 kits sell for and this one would almost certainly out-perform most, if not all of them. However, it would be more difficult to choose this kit if it were at the regular price of $209.99.

When taking the current price into consideration though, I wouldn’t hesitate to jump on this kit. Strong Mhz, tight timings at DDR3-2000 and a price that can’t be beat combine to make this kit Overclockers Approved.

Thank you to Patriot for supplying this kit for review.

–Jeremy Vaughan (hokiealumnus)