Sisoft Sandra 2007 (Memory Bandwidth)
| |
Ram Bandwidth Integer (MB/s) |
Ram Bandwidth Float (MB/s) |
GEIL Black Dragon Quad Channel set
800MHz (5-5-5-15) |
6197 |
6170 |
GEIL Black
Dragon Quad Channel set
800MHz (4-4-4-12) |
6318 |
6344 |
GEIL Black Dragon Quad Channel set
900MHz (5-5-5-15) |
6294 |
6295 |
Teamgroup Elite
2GB PC2-6400 set
800MHz (5-5-5-15) |
6122 |
6095 |
Transcend PC2-6400 Cas5 Value RAM 2x1GB
800MHz (5-5-5-15) |
6085 |
6064 |
Sisoft Sandra 2007 (Memory Latency)
| |
Memory Latency Random Access (ns) |
Speed Factor |
GEIL Black Dragon Quad Channel set
800MHz (5-5-5-15) |
82 |
81.00 |
GEIL Black
Dragon Quad Channel set
800MHz (4-4-4-12) |
74 |
70.20 |
GEIL Black Dragon Quad Channel set
900MHz (5-5-5-15) |
75 |
74.30 |
Teamgroup Elite
2GB PC2-6400 set
800MHz (5-5-5-15) |
81 |
78.20 |
Transcend PC2-6400 Cas5 Value RAM 2x1GB
800MHz (5-5-5-15) |
82 |
80.60 |
Sisoft Sandra is a very popular synthetic benchmark which is ideal for testing the memory bandwidth and latency of any computer. Of course, as Sandra itself warns the user, it is a simple synthetic benchmark which runs a set series of tests and may not tally with real world performance. Higher bandwidth results are better and lower latency results are better.
It would seem that the Black Dragon series receive a distinct performance increase when running at tighter timings, enough to outperform themselves at default settings running at a 12% higher frequency. Although the performance difference is indiscriminate in this test, they also manage to outperform 2GB sets at the same speed and timings in terms of raw bandwidth speed but lose in the latency test.
Everest Ultimate Edition
|
Read (MB/s) |
Write (MB/s) |
Copy (MB/s) |
Latency (ns) |
GEIL Black Dragon Quad Channel set
800MHz (5-5-5-15) |
8233 |
5455 |
5669 |
62.9 |
GEIL Black
Dragon Quad Channel set
800MHz (4-4-4-12) |
8638 |
5458 |
5981 |
57.5 |
GEIL Black Dragon Quad Channel set
900MHz (5-5-5-15) |
8527 |
5456 |
5784 |
62.1 |
Teamgroup Elite
2GB PC2-6400 set
800MHz (5-5-5-15) |
8236 |
5456 |
5651 |
62.4 |
Transcend PC2-6400 Cas5 Value RAM 2x1GB
800MHz (5-5-5-15) |
8194 |
5449 |
5599 |
62.8 |
Everest is also a very popular synthetic benchmark. Despite its similarities with Sandra, Everest is running very simple tests and provides different results which are close to the maximum possible theoretical speed of the system. Higher read, write and copy speeds are better but lower latency readings are better.
Everest clearly shows that the tighter timings can outperform the same modules running at 900MHz, at default settings. The latency test results are much better and the read/copy speeds are significantly higher. It would seem that tighter timings help more than slightly higher speeds, especially when all of the DIMM banks are populated.
PCMark 05
| |
Memory Score |
GEIL Black Dragon Quad Channel set
800MHz (5-5-5-15) |
6119 |
GEIL Black
Dragon Quad Channel set
800MHz (4-4-4-12) |
6261 |
GEIL Black Dragon Quad Channel set
900MHz (5-5-5-15) |
6196 |
Teamgroup Elite
2GB PC2-6400 set
800MHz (5-5-5-15) |
6125 |
Transcend PC2-6400 Cas5 Value RAM 2x1GB
800MHz (5-5-5-15) |
6084 |
PcMark software was developed by FutureMark in an effort to create a benchmark able to test all of the basic components of a system. We ran the memory benchmark 3 times each and noted the average score of the 3 runs. Higher scores are better.
PcMark05 did not alter the performance results we got from our previous tests. Once again the quad channel set performs just as well as other typical dual channel sets. When tweaking and overclocking the set, once again the tighter 4-4-4 timings at the stock speed beat the performance of the same set running 100MHz higher than their rated speed with default timings.
