Mushkin HP3 10666 DDR3 Memory Review

ccokeman - 2007-09-30 15:58:47 in Memory
Category: Memory
Reviewed by: ccokeman   
Reviewed on: October 1, 2007
Mushkin
Mushlin
Price: TBA

Introduction:

Moore's Law states that process improvements will happen roughly every eighteen to twenty four months for electronic components. Well, it's that time again! With DDR2 on the way out, DDR3 is going to be used on the next wave of motherboards. To ride that wave, Mushkin has jumped into the waters with its new HP3-10666 modules that feature enhanced latencies of 6-7-6-18, operation at 1.6 volts and Frostbyte technology to cool the modules. This memory kit consists of two 1 gigabyte modules as a matched set with the signature blue heatspreaders of Mushkin's HP line of modules. The modules are specified to run at a speed of 1333MHz. Let's see if the increased memory speeds positively impact system performance.

"Founded in 1994, Mushkin is best known for producing “Enhanced” memory modules. Located at the base of the Rocky Mountains in Denver, Colorado Mushkin provides performance enhanced computer products to users worldwide. Exceptional quality, enhanced performance and unparalleled customer support are what make Mushkin products the best in the industry."

 

Closer Look:

The memory comes shipped to the consumer in a clamshell enclosure. The modules are indexed into the enclosure and are held tightly in place to prevent any damage during shipment. Inside the enclosure are the two modules as well as the Mushkin product card. The product card shows the company logo on the front and installation and troubleshooting tips on the rear.

 

 

These new modules feature reduced latencies of 6-7-6-18. 1.6v to 1.7v is the voltage required by these modules to operate at the frequency specified. The blue heatspreaders identify these modules as the HP series of modules.

 

 

These modules feature Mushkin's patented Frostbyte technology to keep the RAM cool. The heatspreaders are kept in contact with the memory modules by thermal tape. The tape does not appear to have come loose from any of the modules. Something I have seen on lower grade modules, but not on these.

 

 

Now that you have seen the modules, the next logical step is to get them installed!

 

Installation:

Installing these modules is no different than any other module. The first thing to do would be to power down the system. Press down on the module release lever and pull the modules free. To install the new modules, index the module correctly into the DIMM slot and push down the module on both ends to lock it into place. Verify that the clips are in and secure the modules. When adding new hardware to your system, it is a good idea to clear the CMOS to make sure that all of the settings will be reset to factory defaults. This way, you have no issues on system restart. Now you are ready to hook your system back up, power it up and change the settings in the BIOS. Reboot and enjoy!

 

 

 

 

Specifications:

 

Part number 996574

Frequency:
1333MHz
Latency:
6-7-6-18
Voltage:
1.7-1.8V
Module:
128Mx64
Chip:
Unbuffered
Pins:
240
Chip:
64Mx8

 

Features:

 

Testing:

To test this memory I will use a series of applications and benchmarks to measure the performance of the modules. Since DDR3 is a newer product and we have no other DDR3 modules to test against, I will use some of the higher rated DDR2 modules for comparison. This comparison will be at DDR 800. The DDR3 modules will then be shown at rated speed and timings. All timings and voltages will be set to factory defaults so that there are no variables or overclocking taking place in the first part of our testing. The software that will be used to benchmark this memory is listed below.

 

Benchmarks:

Testing Setup:

 

CPU-Z: This application shows us the settings that we have chosen in the BIOS. Items shown in this application include CPU speed and bus settings, motherboard manufacturer, BIOS revisions, memory timings and SPD chip information.

 

Task Manager: We use this utility to show physical memory, kernel memory, page file usage and processor usage (%).

 

 

Testing:

PcMark05: With this benchmark, I will be running the system suite as well as the memory test suite. The measure for the system suite will be the total score. The measure for memory performance is the total memory score. A comparison will be made of the performance at DDR2 800 and DDR3 800 and 1333 speeds.

 

 

 

 

 

 

 

SiSoftware Sandra XII: In this program, I will be doing the following benchmarks: Cache and Memory, Memory Bandwidth, and Memory Latency. Again, I'll be comparing the sticks at DDR2 800 speed and then the DDR3 will be shown at its native FSB of 1333. Higher is better in all tests except for Memory Latency, where lower is better.

 

 

 

 

 

At stock speeds, the DDR3 memory really does shine. It beats the other modules hands down in all the tests but the latency test and overall PcMark05 score. The latency test was where I was expecting it to not fare as well as the other modules tested.

Testing:

CacheMem: This benchmark is another tool we use to gauge memory read-write performance.

 

 

 

 

 

 

 

 

Higher is Better

 

Far Cry: For this game test, we will use the following settings and run the Hardware OC 1.8 benchmark utility. The measure for this benchmark will be in FPS.

 

 

Higher is Better

 

At DDR 800 speeds, the HP3 falls behind the other two sets of memory in the Cachemem part of the testing. However, it does outperform the other modules by a large margin once at rated speed and timings. The gaming benchmark showed no appreciable gain or loss of frames per second at DDR 800 speeds or above.

Overclocking:

After seeing the performance of the modules at both DDR 800 speeds, as well as the rated DDR3 1333 speeds, I was anxious to see just how far of an overclock I could achieve with these modules. The maximum voltage these modules are rated for is 1.7v. During our testing, we chose to increase the voltage beyond that point to show what the modules can do with a little more juice. The maximum voltage used in our testing is 1.9v. The highest speed that I could achieve on the memory was 765 FSB. To do so, the voltage was lowered to 1.8 volts, as any higher did not allow any increase in memory speed or reduction of the timings. The timings that the HP3-10666 would run at with the maximum overclock are 8-8-7-15. With that being said, I will re-run our benchmark suite at that speed and use the DDR2 Mushkin modules as a comparison running at DDR2-1200 (600 FSB) speeds with the latencies set to 5-5-5-15.  It will be interesting to see how the comparison of these modules will go.

As an enthusiast community, we tend to push our hardware to the maximum all the time. At OverclockersClub.com, we do not condone running your hardware outside of the parameters set by the manufacturer and will not be responsible for any damage to your hardware while trying to duplicate the results of our testing. With that out of the way, let the testing begin.

 

CPU-Z:  765 MHz at 8-8-7-15 latencies.

 

PcMark05: Here are the results of our comparison.

 

Higher is Better

 

Man, did the DDR3 modules show some muscle in the PcMark05 tests. The memory score is huge, so much so that it actually surpassed the overall score.

 

 

Sandra XII:

 

 

 

 

 

 

 

 

 

 

Amazing performance, to say the least. Higher performance was expected, but not quite this high.

 

 

Overclocking:

 

CacheMem:

 

 

 

 

 


 

Higher is Better

 

Far Cry: We will once again use the Hardware OC 1.7 benchmark utility. The settings used on this benchmark are listed below.

 

 

Higher is Better

Conclusion:

These modules did what they were supposed to and more. Rated at 1333MHz, I was able to squeeze 1530MHz out of them without resorting to insane voltages. That is a 98MHz increase over the rated speeds of 667MHz. Performance at this speed was amazing. The bandwidth that was available was huge, easily beating the competition in just about every benchmark. At stock speeds it fell short in only two benchmarks, but even so, it handily beating the DDR2 competition. At stock speeds and voltages the memory ran with the timings set to 5-7-5-15 1T. Yes, you heard that right, 1T. In fact, it ran 1T up to the limit of its performance, something wholly unexpected. When you couple that with the fact that this two gigabyte set delivers the goods and can run with any processor on the market 1:1 and above, it makes this kit a great choice when migrating to the next generation of hardware. With a lifetime warranty, high performance and tight timings, Mushkin, as always, gives you more than expected.

 

Pros:

 

Cons: