DDRII memory is becoming the new RAM that overclockers are using. The higher clock speeds of up to 1GHz (QDR) mean that the memory gives plenty of headroom for those overclocking their PIV processors. This is also where much of the improvements in memory come from currently, as DDRI is pretty much at a clock speed standstill.
As we have mentioned previously Crucial is one of the biggest memory providers. The have plenty of OEM contracts that put Crucial memory into plenty of different computer systems. With DDRII they seem to be one of the leaders in production of this memory.
We've already looked at one kit of Crucial DDRII memory, in the form of the Ballistix PC5300 1GB kit. The RAM itself wasn't the most overclockable, but did well, and with the lower price Crucial has on most of its RAM it did well. Lets see what was sent this time.
Crucial PC6400 4-4-4-15 1GB Kit
Crucial sent the next speed bump, in the form of the 1GB PC6400 kit. Does this RAM look any different from the PC5300 previously reviewed. Lets take a look at the pictures.
What do you get with this RAM? Simply put nothing apart from the actual RAM. In all looks it is identical to the package of the PC2-5300 RAM we looked at in the past. The only difference is that there is no offer for the free software that the other package included.
Installation is the same as most sticks of RAM, simple if you've done it before and fairly easy if you haven't. The only reason you may have trouble is by putting it in backwards, but lining up the pins solves that very easily.
Overclocking
This RAM is rated at 800MHz QDR, so it should allow for a decently high clock speed. However since it is among the higher end of the RAM speed scale, it might not have as much headroom. But as always your results may very well vary. So lets see the results.
The maximum this RAM could get was 800MHz, and even that wasn't perfectly stable, as it occasionally didn't boot at this speed. Whether this was a fault of the RAM or of the motherboard, I can't say for sure, but if I find out I will update the review.
System Setup
| CPU: |
Intel PIV 2.8E 775LGA |
| CPU Clock Speed: |
2.8GHz |
2.8GHz |
2.94GHz |
2.8GHz |
| Motherboard: |
ECS 915P-A
|
Asus P5LD2-Deluxe |
| Memory: |
Corsair TwinX PC4000 (2*512MB) |
Crucial Ballistix PC2-5300 (2*512MB) |
Crucial Ballistix PC2-6400 (2*512MB) |
| Memory Timings: |
3-4-4-7-1 |
4-3-3-12-1 |
3-3-3-12-1 |
4-4-4-12-1 |
| Memory Speed: |
400MHz |
533MHz |
533MHz |
700MHz |
800MHz |
| Hard Drives (Master) : |
80GB Western Digital 7200RPM SE 8MB Buffer
|
| Video Card: |
Asus Extreme AX800XL-2DTV |
| Operating System: |
Windows XP Pro SP2 Direct X 9c |
| Drivers: |
Catalyst 4.7 |
| Cooler: |
Swiftech H20-120 REV. 3 Liquid Kit |
| Case: |
CoolerGuys Windtunnel IV |
| Power Supply: |
RaidMax 400Watt Power Supply |
| Direct X Benchmarks: |
Unreal Tournament 2004 (CPU timedemo) |
|
Half Life 2 - Anandtech Canals Demo |
| Other Benchmarks |
XMPEG 5.03 |
VirtualDub 1.6.4 |
DivX 5.21 |
|
ScienceMark 2 Build 171102 |
SiSoft Sandra 2005 SR2a |
|
TMPGEnc Plus 2.59.47.155 |
Sysmark 2004 |
Just like the previous RAM review, we used two different motherboards, the ECS to show difference between DDR and DDRII, and the Asus to show difference in memory speeds.
All of our tests can show at least some part of memory bandwidth limitations, but they are there to show what difference an increase in bandwidth can make on various programs. The synthetic tests give us an indication of how much difference there will be at the maximum. Our video tests show if encoding in two popular formats makes a difference. So lets see how this faster RAM does.
Synthetic Tests
First we will look at the synthetic as a guide. The two programs we are using ScienceMark and SiSoft Sandra give two different types of results. ScienceMark gives a result that has no special processor optimizations, and SiSoft gives the best case result with any optimization possible. So how fast does ScienceMark say the RAM is?
| ScienceMark Results |
Bandwidth (MB/s) |
| ECS 915P-A DDRI: |
4010.42 |
| ECS 915P-A DDRII: |
4312.75 |
| Asus P5LD2-Deluxe DDRII: |
4388.34 |
| Asus P5LD2-Deluxe DDRII (700MHz): |
4553.80 |
| Asus P5LD2-Deluxe DDRII (800MHz): |
4546.66 |
What does this test show? Both the 700 and 800MHz RAM speeds are very close, though the 700MHz test has an advantage of 10MHz in FSB speed and lower latency. Increasing the bandwidth for the RAM offers a 3.6% increase in used bandwidth, which doesn't bode well for the less memory specific tests. Lets see what SiSoft says.
| SiSoft Sandra Results |
Integer Bandwidth (MB/s) |
Floating Point Bandwidth (MB/s) |
| ECS 915P-A DDRI: |
4656 |
4658 |
| ECS 915P-A DDRII: |
4871 |
4897 |
| Asus P5LD2-Deluxe DDRII: |
4980 |
4983 |
| Asus P5LD2-Deluxe DDRII (700MHz): |
5149 |
5151 |
| Asus P5LD2-Deluxe DDRII (800MHz): |
4965 |
4965 |
This test offers no difference between stock 533MHz DDRII and the 800MHz DDRII. Increasing the available bandwidth of the FSB does give a 3.3% increase on a 5% FSB increase. So which result will prove most like the rest of the results, or will we see something totally different? Lets first look at the video tests.
Video Tests
First we will look at converting from a MPEG-2 file to a DivX AVI file. The video in question is a 8555 frame video from the original Hitchhikers Guide to the Galaxy, which matches to most analog video that a Media Center PC can record these days. So does the additional bandwidth of 2.7 GB/s, albeit with slower timings, show any improvement?
Looking at these results shows what? Increasing the memory bandwidth gives a 2% increase, just above our margin for error. This results in an extra 1fps speedup when moving to an 800MHz memory bus. Overclocking the processor by 5% gives a larger difference, of 6.5%. Considering the fact that the memory bus didn't cause more than a 2% increase on its own here, the test is more CPU dependent at this time. What about a program that doesn't use any specific optimizations for encoding a video, will it show a marked improvement like that of ScienceMark?
This program doesn't show any real difference between DDRI at 400MHz and DDRII at 800MHz with similar latencies. Overclocking the processor is the only way that there is any improvement in this test. So we can see in this case that an optimized program can show very moderate improvements over a non-optimized one. What if we encode in MPEG-2, does this show the same improvements as the XMpeg test?
NEXT
