Archive 1997–2024

ASUS Striker II Formula under the Microscope

Initial Thoughts

Initial BIOS releases from ASUS were certainly a little rough around the edges, with each release completely changing how the board operates. The 1101 BIOS release is pretty much dead-on for dual-core CPU overclocking and general day-to-day use. During our time with the board, we were subjected to five different BIOS revisions, each with their own sweet-spots for performance and idiosyncrasies. Performance with the 1101 BIOS is generally good, and stability and peripheral device operation is excellent as we will see in the comparison article.

NVIDIA's 780i is on a par with the Intel X38/X48 DDR2 boards when it comes to dual-core CPU overclocking in the performance zone. ASUS has managed to supply clean power to the memory banks that allows us solid performance and 8GB memory overclocking potential, something we have not managed on other boards with the same settings.

It will be interesting to see how the reference boards measure up to the strengths of the Striker II Formula. For now, we would not hesitate to recommend this board for a dual-core SLI-based gaming system. Although 790i may be robbing this platform of the limelight, the real-world performance figures of both platforms are not enough to convince gamers that they need to make a jump over to the new DDR3-based chipset unless they have money burning in their pockets.

For the quad-core CPUs, we see the 780i coping with the 45nm Penryns in the same way we experienced when the 680i was first introduced to 65nm quad-core processors. As we expected, refinements to the GTL reference voltage circuitry has improved things a little, but the basic characteristics of the 680i SPP remain intact. The stable FSB limits of the 780i boards are still lower than we'd like - especially compared to Intel boards.

It's also interesting to note that the board behaves quite differently in its reboot cycles and failed overclock recovery when a quad-core CPU is used for testing. These effects generally begin to manifest as FSB speeds approach 400FSB, while speeds over this may require a CMOS reset to get the board up and running again. Then you can return to the prior settings before the overzealous overclocking attempt.

Our experiences using the EVGA 780i board so far have shown similar tendencies in certain ways, so the fault is down to the BIOS base code or the SPP itself. "Memory hole" issues are almost non-existent, with the Penryn quad-cores being the only ones to show any real flaws in the current BIOS. We still have some testing left to do, and our conclusion hinges upon the capabilities of the EVGA board when subjected to similar forms of testing. (The failure of the original board is naturally a concern already, however.) Of course, we are seeing the revised 750i and 650i boards hold their own in SLI performance in the midrange sector, so our final roundup of NVIDIA's current products for Intel owners should be interesting.

Originally published at https://www.anandtech.com/show/2484 (Wayback Machine).