Gaming Platform Performance:
-
Gigabyte GA-X38-DS5 - 7900 GTX
-
Gigabyte GA-X38T-DQ6 - 7900 GTX
-
Asus Maximus Formula - 7900 GTX
-
Abit IP35 Pro - 7900 GTX
Frames Per Second - higher is better
-
Gigabyte GA-X38-DS5 - 7900 GTX
-
Gigabyte GA-X38-DS5 - X1900 CF
-
Gigabyte GA-X38T-DQ6 - 7900 GTX
-
Gigabyte GA-X38T-DQ6 - X1900 CF
-
Asus Maximus Formula - 7900 GTX
-
Asus Maximus Formula - X1900 CF
-
Abit IP35 Pro - 7900 GTX
-
Abit IP35 Pro - X1900 CF
-
64.5
-
135.4
-
65.0
-
137.6
-
64.5
-
126.1
-
64.2
-
71.5
Frames Per Second - higher is better
-
Gigabyte GA-X38-DS5 - 7900 GTX
-
Gigabyte GA-X38-DS5 - X1900 CF
-
Gigabyte GA-X38T-DQ6 - 7900 GTX
-
Gigabyte GA-X38T-DQ6 - X1900 CF
-
Asus Maximus Formula - 7900 GTX
-
Asus Maximus Formula - X1900 CF
-
Abit IP35 Pro - 7900 GTX
-
Abit IP35 Pro - X1900 CF
-
84.4
-
109.0
-
84.0
-
115.8
-
85.1
-
114.1
-
79.8
-
117.2
Frames Per Second - higher is better
-
Gigabyte GA-X38-DS5 - 7900 GTX
-
Gigabyte GA-X38-DS5 - X1900 CF
-
Gigabyte GA-X38T-DQ6 - 7900 GTX
-
Gigabyte GA-X38T-DQ6 - X1900 CF
-
Asus Maximus Formula - 7900 GTX
-
Asus Maximus Formula - X1900 CF
-
Abit IP35 Pro - 7900 GTX
-
Abit IP35 Pro - X1900 CF
-
73.9
-
92.9
-
74.6
-
94.5
-
73.1
-
93.0
-
72.6
-
81.6
Frames Per Second - higher is better
Performance on Gigabyte's boards is very good, if a little sporadic at times. It performs slightly better in
Call of Duty 2 and
World In Conflict, and is also consistently fast in
Enemy Territory: Quake Wars, but falls slightly short in
Team Fortress 2.
TF2 is more CPU and memory dependent, so that could be a reason for the DS5 dropping a few FPS with CrossFire enabled, but it's still just as fast with a single card.
Overall, the X38 chipset is faster in
CoD2 and
ET:QW by offering up more bandwidth from the dual X16s than the x16/x4 on the P35. We couldn't get World in Conflict to finish the timedemo on any of the boards we tested it with when we ran X1900s in CrossFire using the latest Catalyst 7.10 drivers, it simply crashed out all the time. Using a single card or without enabling CrossFire the test was fine however. We also tried a pair of HD2600 Pros using internal CrossFire (not old external dongle types) and found the problem persisted.
Want to comment? Please log in.