Print this article

Patriot Memory 2GB DDR3 1333 (PDC32G1333LLK)


Date
: 01/4/08 – 04:29:07 PM

Author
:

Category
: Memory

Page 1 : Index

Manufacturer
: Patriot Memory

Patriot Memory is not a new name in the game. Many of you may recall the days when we looked at some products by
PDP Systems
. After organizational restructuring,
Patriot Memory
was born and we're pleased to welcome them back.

Today we have the opportunity of introducing one of Patriot Memory's DDR3 products. As many of you know, DDR3 is the latest product available on the market and while the top end kits cost upwards of $600, you won't have to worry about spending nearly that much for this kit. Our Patriot Memory PC-10666 7-7-7-20 modules are a great stepping stone for those looking to entertain themselves with DDR3.

Page 2 : Package & Contents

Memory packages have not changed since DDR was introduced and it looks like Patriot Memory has more or less stuck with the simple but effective blister package. From the package, we can see that they have also stuck with the tried, tested and true aluminum heat spreaders. Apart from color, they look exactly like the 2GB PC4000 ELK and 2GB PC5300. To the backside of the package we very clearly see the specifications of the modules: PC3-10666 (1333Mhz), 2GB (2x1GB) Kit, 7-7-7-20 and 1.7V.

If you're new to the game of memory installation or how to install DDR3 modules, the included insert doubles as an installation guide, complete with pictures for your reference. The insert also outlines where to get support, there is no 1-800 number, and the lifetime warranty included in the kit. There is no mention about warranty after you overclock the modules and crank up the voltage supplied.

Removing the modules from the blister package we get our first good look at the sticks. Everything you need to know about the modules is included on the sticker.

Taking a close look at the PCB, you can see production name:
B63URCA 0.71
. These are BrainPower PCBs and Patriot Memory is not the only organization using them. Jody has revealed in his PC3-10666 review that OCZ is also using them. There's very limited information about this PCB so we won't know if it'll be a huge success like other BP products.

The aluminum heatspeader used on the Patriot Memory modules are very low profile and will not restrict your ability to use any CPU cooler.

You can also see from the photo above that the modules are single sided, unlike the OCZ modules, and while the PDC32G1333LLK doesn't have the same bling factor as OCZ, it's all about performance at Overclockers Online.

Page 3 : Specifications

As per Overclockers Online standards, we'll start with a brief look at the specifications. Here's what Patriot Memory has to say about their own lineup:

Click here for a direct link to their product page.

Everything you need to know about this kit and what configurations you can buy them in can be found on this image:

After booting up my machine, here's a look at the SPD table determined by CPU-Z.

Normally at this stage of the review I would be more than happy to rip off the heatspreader to see what's under the hood. However, I don't want to risk damaging my first kit and will hold off until later. My guess is that they are using Elpida or Qimonda modules for this kit as they have been seen in other Patriot Memory products.

Page 4 : Installation

There's not a whole lot to show with DDR3 installation. Once you've seen the process on DDR or DDR2, DDR3 is no different.

There's plenty of room for the largest heatsink you can find that will fit your motherboard and from the view below, you can see that the DIMM is no thicker than the tabs on the memory slots.

Page 5 : Overclocking

This is the section that really makes or breaks a kit of memory. After all, all memory at the same frequency and timings will perform the same. What sets a kit apart from the others is the ability, or lack thereof, to overclock… at least for me it is. The following is a complete list of components used for the DDR3 overclocking and benchmarks. Below that is a list of the hardware used for the comparison DDR2 benchmarks.

DDR3 Bench Setup:

DDR2 Bench Setup:

My stability testing consists of a number of programs including the following list.

• Single 32M runs of SuperPi Mod 1.5

• 2 hours of quad Prime95 version 25.5
• 1000% coverage of dual MemTest Pro in Windows using two 850MB/instance (~1.5hrs)
• Multiple loops of each 3DMark 2006

You might be wondering why I'm not using two instances of 32M SuperPi runs. The reason is simple – regardless of the speed, latency or voltage supplied, one instance of SuperPi
always
fails. This is not the only kit I've heard of that has this problem. Some engineers blame SuperPi, others don't. Take the stability and results with a grain of salt. Every kit on the market will perform differently.

The screenshots at the bottom of the overclocking section outline what this memory was capable of with the outlined testing above. The graph just below is indicative of performance capable and takes into consideration all of the programs above. Some higher frequencies were possible, but often at the cost of Prime95 leading to errors.

The modules actually scaled much better than I had thought. You can see a rather large jump when going from 6-6-5-15 1T to 7-7-7-20 1T while there's marginal benefit when further relaxing the timing to 8-8-8-20 1T. As we loosened the timings and increased the voltage beyond 1.96V, the success of the overclocking process didn't improve all that much more. Here are some screen shots of the best overclocks I achieved.

DDR3-1480Mhz 1T @ 2.00V

DDR3-1680Mhz 1T @ 1.96V

DDR3-1720Mhz 1T @ 1.90V

We will now do some benchmarking at the stock frequencies and timings as well as overclocked and compare the results with a few DDR2 results.

Page 6 : Memory Benchmarks

Comparing the results of a DDR2 machine to a DDR3 machine isn't exactly easy, especially when the two machines are so incredibly different. However, that's the type of situation you might find yourself in when building a whole new machine. You don't exactly expect to carry one part over to the next. With that said, the DDR2 results will only be available for your reference and I will not be commenting on its performance. To kick things off, we'll begin with a few memory specific benchmarks.

SiSoft Sandra Lite XII – Memory Bandwidth

Lavalys Everest Ultimate 07 v4.00 – Memory Bandwidth

ScienceMark v2.00 – Memory Bandwidth

While all three applications came up with different numbers, the ranking is very consistent. At the highest frequency, the Buffalo DDR2 and Patriot Memory DDR3-1700 go toe to toe. At any frequency less, the DDR2-1200 has the advantage. It's quite interesting to see how Patriot Memory could have specified these as 6-6-5-15 as it was absolutely rock-hard stable at this timing.

Lavalys Everest / SiSoft Sandra – Memory Latency

The latency is pretty much all over the place, the lowest latency at the highest speed with the loosest timing. That said, DDR2 isn't even in the same league as DDR3. DDR2 pulls ahead by a few nanoseconds.

Page 7 : System Benchmarks

FutureMark PCMark 05

It's not even fair to compare the DDR3 to the DDR2 results because of the system differences, so I won't. Focusing on the DDR3 results, we see the high frequency looser timing (1700-8-8-8-20) doesn't give that many more advantages over the tighter timed but lower clocked 1512-7-7-7-20. A mere 3 point advantage isn't a whole lot to be excited about.

WinRAR 500MB Benchmark

Only through sheer luck did the results show the WinRar timings completely independent of memory type and CPU speed. What we do see is that memory performance has a significant effect on the result. FOr the same speed, increasing the memory frequency from 1512 to 1700 shaved off 16 seconds from the time.

SuperPi Mod v/1.5 – 1M

SuperPi Mod v/1.5 – 8M

SuperPi Mod v/1.5 – 32M

The ultimate memory performance application. As Jody once mentioned, he who has the most bandwidth rules SuperPi… okay so he didn't exactly say it like that. Basically we again see consistent performance coming out Super Pi results as we've seen in the bandwidth reports. If we extrapolate the results, it's clear that DDR3 can compete with DDR2 in SuperPi performance.

Page 8 : 3-D & Gaming Benchmarks

FutureMark 3DMark 01/03/05/06

If you wanted boring results, this page is where it's at. There are negligible differences in performance once the CPU speeds are the same. Gaming benchmarks are heavily reliant on video cards, so once you reach your GPU bottleneck, your CPU can only carry you so far.

Gaming Benchmarks – Crysis Demo (CPU) / World in Conflict

Once again, it's a dead draw with World in Conflict and Crysis. The system is limited to GPU performance and we can clearly see that here.

Overall with gaming, DDR3 helps in giving you that high CPU frequency so it doesn't become the bottleneck, but like all systems out there, if the video card is the bottleneck, the world's fastest CPU won't help you very much.

Page 9 : Conclusion

We've looked at a few numbers and the Patriot Memory DDR3-1333 has a good future ahead of itself. The bandwidth performance is amazing. If you buy Patriot Memory's higher binned gear, one can hope you can expect higher performance.

The Patriot Memory DDR3-1333 had a good amount of headroom for the stock settings. If you're dying to enter the DDR3 game but don't have a lot of cash to your name, this wouldn't be a bad place to start. DDR3 is still very expensive regardless of whom you buy from, but building a new machine around DDR3 will delay future upgrading… at least for a little while.

Advantages

• Stay on the bleeding edge of technology

• Capable of running tighter timings and faster
• Performance on par or better than DDR2

Disadvantages

• DDR3 is expensive. $295 for 2GB

• Won't run dual 32M Super Pi at any setting

Overclockers Online would like to thank Patriot Memory for making this review possible.