Kingston Hyper-X blu DDR3 1600 – Review

Kingston is one of the largest, if not the largest provider of memory products for various electronic devices. It has its claws extended in almost all memory sub-territories. Be it Flash memory (flash drives, memory card for phones and solid state drives) or random access memory –Kingston is everywhere.

The company caters to all market segments when it comes to desktop memory. From the ubiquitous Value Ram to the high end Hyper-X.

To offer more at the high end, Kingston started offering a variant of its Hyper-X RAM, called Hyper-X blu. This is an entry level memory kit mean for gamers and enthusiasts. It pairs the performance characteristics of the Hyper-X family with the price of its value line.

Today we have, under the knife, a 4GB kit (2×2) of DDR3-1600 MHz RAM under the Hyper-X blu banner.

MEMORY FACTS

PACKAGING

The modules come in a very Kingston like box. The outer cardboard shell holds the side-by-side plastic containers. Each container has one of the two, 2GB memory sticks.

Both modules feature a blue Hyper-X styled heat sink with the memory spec label.

MEMORY MODULES

Each memory module has a speed of 1600 MHz, rated to work at CAS latency of 9. Kingston is probably being conservative with the CAS latency timing on this kit as it worked at a latency setting of 8 at 1600MHz without changing any other parameter.

CPU-Z shows the timing characteristics of this memory kit

As you can see, the 1600 MHz profile is tied to Intel’s XMP specification. The rest of the timings are defined by JEDEC specifications. As has already been mentioned that, though the CAS latency is listed as 9, the modules work at 8 without changing any other parameters (voltage or timings).

Also note that the memory requires 1.65 Volts to work at DDR3-1600 speed.

TESTING

Intel is changing the rules of desktop memory marketing with the advent of its Sandybridge line of processors (the new Core i series). With these processors it won’t be able to overclock by raising the base clock. This is because the clock generator will reside on the processor, rather than the motherboard. With the base clock, for all practical purposes, locked, Intel will offer multiplier unlocked version of its processors. This, like the current practice, will be indicated by a “K” suffix at the end of the processor name. Their will be a price premium over unlocked variants, but it will not be as much as between a stock and extreme edition processors.

With the base clock locked the allure of buying high speed memory to overclock a processor diminishes greatly.

Memory speed does not have as much an impact on real world performance as processor speed has.

How will memory makers react to this remains to be seen, but Intel has set new ground rules.

With this in mind, most of the tests done are at in-spec speeds. Though an overclock test was done to see how far the memory can be pushed, it was not a variable that weighed in heavily on over all memory performance.

TEST SYSTEM

The test system contains one synthetic and two real world tests (The winrar benchmark can be considered to be semi-synthetic).

1. SANDRA –LATENCY & MEMORY BANDWIDTH

There is a significant difference in performance, especially between 1066 and 1600 MHz. But keep in mind this is purely synthetic benchmark

2. WINRAR
Winrar has a built-in benchmark tool. This can be used to gauge the program’s performance with varying memory speeds

Again a significant performance difference is seen with this test as well.

3. MAFIA II
Mafia II is the long awaited sequel to one of the greatest third person shooter –Mafia. It was selected because of all the games tested this was the one that showed the most “difference” in performance as memory timing were changed. The built-in benchmark tool as used. Just to make sure that the lack of user generated content does not make a difference, the first level was played while benchmarking with FRAPS. (The difference between the two was not statistically significant).

The game was run at 1080p, with all in game settings turned to their maximum. As the testing was done on an ATi video card equipped system, dedicated Phys-X and APEX effects were disabled.

In a dramatic turn of events, the game churns out its best performance at 1333 MHz, rather than the rated 1600 MHz. It does better at 1066 MHz as compared to 1600 MHz settings as well. This is due to a balance between timings and memory speed. As has been said this was the most “visible” difference seen when testing modern games (released within the last year). Even then the difference is performance is a measly 6%, which is a difference only because it is so and not because it makes any difference in game play.

OVER CLOCKING & IMPRESSIONS

The memory over clocked to an admirable 2000 MHz at relaxed memory timings (CL 9). The synthetic benchmarks did show an improvement but the gaming benchmark preferred a balanced setup (between timings and speed).

In a changing scenario, where memory speed will take a backstage to capacity and probably timings, the 1600 MHz Hyper-X blu kit represents the best that old guard has to offer. There is nothing wrong with the kit and it does what it is supposed to do: i.e. performance at a reasonable price.

It will be interesting to see how Kingston and other memory makers respond to Intel’s move of locking the base clock. I have not factored AMD processors in this, as they are not the king of performance any more. There is some optimism with regards to their Bulldozer architecture, but even as Intel *leaks* decent Sandybridge performance figures, AMD has nothing concrete to offer. Also AMD already offer black edition (multiplier unlocked) processors, which take away the memory speed factor altogether as well.

Also of note is the fact that most games prefer quicker memory as compared to faster memory.

PROS

• Performs better than stated specs (works at CL 8 at 1600MHz)
• Overclocks well
• Lifetime warranty

CONS

• Faster vs Quicker –You’ll need to decide what you want