| |
| ||
| by Wankerfx | August 8, 2007 | ||
| OCZ PC2-6400 Platinum R.2 Review OCZ PC2-6400 Platinum R.2 Review Price: $96.50 (After MIR) Availability: Chrome Manufacturer's Part Number: P/N - OCZ2P800R21G Warranty: Lifetime Table of Contents 1- Introduction 2- First Impressions 3- Setup 4- Testing 5- Conclusion Introduction: Hardware fanatics are seeking for top-performing memory, but are held back by a budget. OCZ, once again, proves to be the answer to this problem. OCZ entered the memory market in August 2000, making them a veteran to the computer performance world. OCZ focuses on what is best for consumers, rather than drawing their attention to profits. OCZ EL DDR2 PC2-6400 / 800 MHz / Enhanced Latency / Platinum Edition / Revision 2 Part #: P/N - OCZ2P800R21G Capacity: 2X1GB kit Timings: 4-4-4-15 Voltage: 1.9V-2.1V (2.2V with EVP) First Impressions: Attachment 767 The memory is protected in a tough plastic container. The packaging is simple, but it is not too bland. The memory modules are clearly displayed, showing off their magnificent shine. Attachment 768Attachment 769 OCZ has proven to be the dominant memory out of all companies in the visual department, but are they all show and no go? We will see later on in the review. This kit uses XTC technology to dissipate heat quicker. XTC stands for Xtreme Thermal Convection, and effectively implements a honeycomb design on the heatspreader. The honeycomb design is effective, because it helps air circulate across the chips, removing the pockets of heat which are often held in place with normal heatsinks. This keeps the memory cool and potentially increases the overclocking power of these modules. Many overclockers are discouraged from buying ram when they find out that a certain kit of ram does not have Micron chips. This kit uses ProMOS chips and utilizes an unknown PCB. There are some pros and cons about using ProMOS chips; they achieve decent overclocks with low voltage, but their performance suffers when trying to achieve high overclocks. In addition to the amazing visuals of these modules, OCZ incorporates a spectacular warranty. They are known to be the most supportive when it comes to overclocking, enabling enthusiasts to do what they do best. This kit is under EVP (Extended Voltage Protection), which enables consumers to overvolt this set of ram to 2.2V without voiding the warranty. OCZ demonstrates its full support for their consumers; they support overvoltage to a certain extent, and still cover the ram modules under their lifetime warranty. Setup: CPU: Intel E2140 Motherboard: DFI Dark Cooling: Tuniq Tower 120 GPU: Gigabyte 7100GS HDD: Western Digital 250GB PSU: Enermax 535W OS: Windows XP Testing: All results are dual 32M stable. This leaves a good sign of stability, but is not recommended stable for 24/7 use. Multiple tests were performed with various applications, such as: Doom 3 FPS test - Doom 3 is a game full of high quality graphics, it demands fast computer components in order to run smoothly. The Doom 3 test is perfect for measuring the abilities of one's hardware, making it an excellent benchmark for this review. Half-life 2 FPS test - Half-life 2 is another good benchmark program, it is one of the most played games in the twenty-first century, making the results appealing for gamers. Everest Ultimate 3.00.630 read and write tests - Everest is quite common amongst bandwidth junkies; it gives the option to do read and write tests for the ram and measures the results in MB/S. SiSoft Sandra Lite 2007.6.11 INT and Float bandwidth tests - Another great program used to measure bandwidth. (INT and FLOAT, both measured in MB/S) SuperPI 1M - SuperPI is a program used to measure the amount of time it takes to calculate a certain amount of digits in PI. I chose one million digits, because it is very popular in the benchmarking world. CrystalMark 2004R2 - CrystalMark 2004R2 is a program which is used to benchmark memory and several other computer components. It was developed by "hiyohiyo", an enthusiast in Japan, and is widely used amongst several benchmarking forums. A test was made to see if the stock speeds were stable, and then several overclocking tests were performed.The first test was to see if this kit could handle CL3 timings at DDR2-800; it did not boot, not even with several combinations of sub timings and voltages. Moving into the CL4 zone was a relief, as the computer booted without a problem. The following results were attained with timings 4-4-4-12 and the modules accomplished some decent results: Attachment 770Attachment 771 The gaming performance increase should be taken with a grain of salt, seeing as the DFI board cannot run the processor and ram frequency unlinked like the 680i chipset. Processor speed ranges from 2.5ghz-2.6ghz, which affects the FPS of these games, making the results somewhat useless. It was quite surprising when this kit accomplished DDR2-1000 with a relatively low voltage, seeing as these modules use ProMOS chips. Although the game performance was not too impressive, the bandwidth results proved to be of significance. The most notable bandwidth increase is when comparing speed vs voltage. It takes 0.1V to increase the speed from DDR2-900 to DDR2-940, but takes an additional measly 0.05V to increase the frequency to DDR2-1002. The memory is running at DDR2-1002 with 2.25V (slightly above warranty), and it demonstrates a substantial performance increase over DDR2-940 bandwidth-wise. The performance was quite impressive for a budget set of ram, enabling the user to achieve a high frequency without too much voltage. The ram ran cool without a fan, due to its XTC technology, which proves to be effective. An attempt was made to add more than 2.25V into the ram, but the computer would not boot. This also resulted in a corrupted Windows. Lets move onto some CL5, shall we? Maximum stable frequencies were tested with timings 5-5-5-15, but the results were not too impressive. Again, the CPU frequency was different when comparing it to other tests, so please don't pay too much attention to it. The looser timings and the lower frequency took its toll, resulting in lower results in comparison to CL4. The ram with this latency would not accept anything higher than 2.2V, forcing a max stable overclock of DDR2-982. The bandwidth results were average, and the frequency these modules achieved was quite poor for such loose timings. It was surprising to see CL4 running faster and being able to accept more voltage. Any voltage set beyond 2.2V would cause instability for my system with these timings. Overall, the gaming and bandwidth results for this ram was not too bad considering the price of these modules. These sticks are well-suited for those who have CPUs with high multipliers, giving them the ability to run their CPU at high speeds, while maintaining tight timings. A test of SuperPI 1M was conducted for each of the stable speeds: The CPU speed was linked to the ram speed, so the results will vary. Lower timings are still in favour of SuperPI, enabling the modules to calculate one million digits of PI faster than higher timings. These results are really good, considering most ram in this price range would barely be capable of hitting DDR2-900 with tight timings. Here are some of the results that were acquired from CrystalMark testing: This program enables you to test just the ram modules' performance, so that CPU speed does not have anything to do with it. DDR2-1002, 4-4-4-12 completely dominates the other scores. It is worth mentioning that it beats DDR2-982, 5-5-5-15 by an astonishing 674 points, leaving it in the dust. This program shows users that timings win the show, even when frequencies are rather close. Conclusion: This kit has amazing visuals, and it performs really well for the price and the warranty offered. However, these modules do not respond to higher voltages at all. Anything above 2.25V for CL4 and 2.2V for CL5 would cause the test computer to crash upon startup. This kit held its ground and reputation when proving to be stable at DDR2-1002, 4-4-4-12 with only 2.25V. There is no need to run a fan above these modules, seeing as they performed cool through several of the vigorous tests (Thanks to the XTC technology). Another thing I should note is that these modules performed quite well in the CL4 bandwidth testing. This kit should appeal to most enthusiasts on a budget, or to those hardcore FPS gamers, because it demonstrates the ability to deliver higher framerates with an easy overclock. Please note that not all kits of ram are the same, your results may vary from the ones I achieved; it is all the luck of the draw. I also like how OCZ offers a lifetime warranty for these modules, even when operating at 2.2V. With the combination of great performance, great looks and awesome customer service, I would have to say that this kit is of extreme value. Remember, going past 2.2V will void your warranty for these modules, please have fun overclocking, but know your limits. I am awarding this kit the Dam Good Value award, because it demonstrates the performance of an expensive kit, while maintaining a reasonable price and an exceptional warranty. Pros: -Cutting edge looks -Great value -Decent performance Cons: -Does not work well with higher voltages -Cannot achieve DDR2-800+ with CL3 timings. I would like to thank OCZ for sending me this sample to review and for keeping in touch with me. Please feel free to post your comments and questions in this thread: http://www.hardwarecanucks.com/ram/2...html#post16486 |
