OCZ Gold PC4000 GX XTC 2GB Dual Channel Kit
Saying OCZ is not a great memory company would be saying overclocking is pointless and no fun. Are you nuts? I should just slap you in the head if you think that way and still call yourself a PC enthusiast. The main reason why OCZ is a top notch memory provider is because they listen to the cries and whines of consumers. They provide the solutions to our demands. When the 2GB frenzy began, OCZ was one of the first, if not the first, to provide a 2GB solution that was rated faster than PC3200. I was skeptical about the Gold PC4000 that used Samsung UCCC ICs. The memory sounded fantastic, but the timings were rather poor at 3-4-4-8. I have to be honest and say that I was accustomed to TCCD's rather tighter timings at insanely high speeds. 250MHz with latencies of 3-4-4-8 was just a shrug to me. I expected better, especially from OCZ. After spending time experimenting and learning various 2GB solutions, I looked in the mirror and told myself not to look down on OCZ anymore. High clocking 2GB solutions are very hard to find and the trade-off with looser timings for the larger capacity is more than understandable. That was their first answer to the demand of enthusiast 2GB kits. The second answer came with the release of the Platinum PC4000 EB. These modules featured Infineon CE-6 ICs and were capable of running 250MHz+ with timings of 3-3-2. While many of you will think that those timings are nothing to brag about, the truth is that those timings are exceptional. I would say running the CE-6 at 265MHz-270MHz at 3-3-2-8 game stable is equivalent to TCCD running at 300MHz 2.5-3-3-6. Yes, that is how good those CE-6 are, even if it is picky about certain integrated memory controllers or have game stability issues at high speeds. Now we get to the third solution from OCZ--heatspreaders. Enthusiasts have been complaining about heatspreaders for decades now and companies never really respond to them. One of the reasons is that heatspreaders are considered to be a commodity more than a necessity. From my point of view, I see heatspreaders as a way for companies to distinguish themselves. What enthusiasts complain about is the heat trapping issue. OCZ has taken the time and developed their XTC (Xtreme Thermal Convection) heatspreaders. These new heatspreaders are designed with honeycomb openings to allow heat to escape quickly and work even better with airflow. Is it just a gimmick or will it be the next best thing since sliced bread? That will be hard to tell, but it can only mean one thing--overclocking the memory will no longer be affected by trapped heat.
OCZ Gold PC4000 GX XTC
The package of the
Gold PC4000 GX XTC is the same as all the other OCZ memory kits--a clear
clamshell with an orange insert. The new XTC heatspreaders no longer use
metallic clips to hold the heatspreaders in place. The only thing used to
keep the heatspreaders on the ICs is a layer of thermal tape.
Unfortunately, these new XTC heatspreaders are made of aluminum and not copper.
At first glance, the XTC heatspreaders look to be a single piece of metal, but
they are actually a two-piece part. The frame with the "Z" logo is one
piece while the honeycombs behind is another piece.
The prices on
Samsung UCCC based memory have dropped drastically, making it more affordable
for all types of enthusiasts. OCZ continues to model their UCCC based
memory under their Gold line with their Gamer eXtreme label. It would only
make sense since it is less expensive than before and still allows good
overclocking potential for increased performance. The particular sample I
received came with a date code of 528 on the UCCC ICs. From the Corsair
XMS PC4000PT review, we have seen that date codes do matter when it comes to
UCCC overclocking headroom. We will find out how these 528 chips fair
later on in the overclocking section. OCZ has tightened the timings on the
Gold PC4000 from 3-4-4-8 to 3-4-3-8, running at speed of 250MHz with a voltage
rating of 2.75v. It is not a huge difference, but a good difference
nonetheless. The PCB used is none other than the Brainpower 808.
Test System and Overclocking
Test System:
Overclocking: Keep in mind that during overclocking, the latencies are as follows: CAS-TRCD-TRP-TRAS The CPU multiplier has been kept at 9x to prevent any CPU bottlenecks during the overclocking process. My San Diego 3700+ has a good integrated memory controller and has been double checked to make sure that it is not a source of bottleneck. The CPU has been tested and verified to be stable to at least 9x300HTT = 2.7GHz. Thus, all benchmarks will be conducted with a 9x multiplier. Going through the standard steps of overclocking, I began by dropping the timings as low as the memory would allow. There is not much to expect from UCCC in the low latency category. The lowest they like is 2.5-3-3. Maintaining a voltage of 2.8v, which is what UCCC typically likes, the memory pushed to a decent 225MHz before coming to a halt. I loosened TRCD to 4 and managed to get another 5MHz out of the memory, resulting with 230MHz. Looser timings of 2.5-4-4-8 and 3-3-3-8 provided no help and neither did lowered and increased voltages. I relaxed the timings to the rated 3-4-3-8 and came to a decent 265MHz before hitting a brick wall. No matter what I did beyond that point was futile. The memory would not budge at all. This is just another proof to show that UCCC date codes play a very important role when it comes to overclocking headroom. The best UCCC so far is still the original Gold PC4000 with a date code of 525. Overclocking Results:
Results
A series of memory intensive benchmarks will be run to test the performance of our sample. The other tests will be based on common benchmark programs, such as 3DMark and PCMark. Real life performance differences will be tested through current games, such as Far Cry and UT2004. Memory bandwidth benchmarks will be based off of SiSoft Sandra and Everest Home Edition. Super PI and PiFast will also be used to test memory performance, as the benchmark benefits largely from FSB increases. SiSoft Sandra Memory Bandwidth
SiSoft Sandra is the most widely used memory bandwidth test. It can tell you the increase in performance from just lowering the latency timings. Everest Home Edition
Much like Sandra, Everest Home Edition tests the memory's ability to read and write data in large chunks. The results show that the lower latencies seem to help just a little with performance. 3DMark2001SE
3DMark is likely the overclocker's favorite benchmark. Unlike 3DMark2003, which stresses mainly the video card, 3DMark2001SE tests all of the system's main components. Increasing the CPU speed, memory speed, and video card speeds will result in higher results in the final score.
Results Continued
Far Cry
Since Far
Cry is a game, the performance results will be very similar to 3dMark2001. I
ran the benchmark at 640x480, so the benchmark would not be too GPU-dependent.
The tests under 1024x768 and 1280x1024 are an approximation of what kind of
frames you'll be getting. While the results are similar to 3dMark2001 in
performance, this benchmark will give more significant results, because Far Cry
is a lot more demanding.
Unreal
Tournament 2004
Unreal
Tournament is another popular game engine to test overall system performance.
Here, we can see the same results as Far Cry and 3dMark2001. Like Far Cry,
UT2004 was also run at 640x480 to limit GPU dependency. The other resolutions
provide an approximation of actual game play performance.
PCMark2002
PCMark is
another benchmark created by Futuremark. It tests the memory compression and
decompression speeds in small to large chunks of data.
Super PI
Since the
value of PI is an infinite value, Super PI is a speed test to find the digit
you're looking for. In this case, we chose the millionth digit. Super PI
benefits greatly from higher FSB, as well as low latencies.
PiFast
PiFast is
similar to Super PI. The results can be greatly affected by a slight adjustment
to the memory megahertz and latency.
Conclusion and Thoughts Many of you might question why the heatspreaders were not tested for performance results. The reason is because there are too many variables to allow proper data collection. The CPU cooler will release airflow towards the area of the memory DIMMs and affect the cooling performance of the heatspreaders. The position and location of the temperature probe becomes another factor if not placed correctly. The list goes on and on and is much more complicated than testing a CPU heatsink. All I can say is that the heatspreaders were not even warm to the touch during full load. The previous Gold PC4000 with the copper heatspreaders were quite hot during full load tests. That alone should give you a small idea of what to expect. While the Gold PC4000 GX XTC did not overclock exceptionally well like before, it is in no way OCZ's fault. Their original Gold PC4000 was able to overclock very high to 280MHz so there is no reason why they cannot do it again. The PCB used is the same, but the ICs used are different. The date codes of Samsung UCCC ICs play a huge role when it comes to the maximum overclocking potential. I guess date 528 is just not as good as the older 525. However, OCZ has done a great job with the new implementation of their XTC heatspreaders and reaching out to the gaming market. Everyone is entitled to his or her opinions, but I believe that the XTC heatspreaders are very sexy. The use of those heatspreaders will be great for enthusiasts who want their parts to run nice and cool. The XTC also looks great inside the system. Compared to typical heatspreaders, the shiny mirror finish and honeycomb vents will definitely stand out in a windowed case. The price of the OCZ Gold PC4000 GX XTC is also nothing to complain about. It is very affordable and will easily fit within your gaming system budget. Great looks, affordable price, and good overclocking potential? I'm sold.
Pros:
Cons:
We
would like to thank OCZ for
providing us the sample. If you have any questions or comments, feel free
to hit us up in the
forums. You can also check out more of our latest reviews on the
front page.
|
