I normally make a major upgrade around this time of year. Although I have been mostly happy with my A8N-SLI Deluxe, I figured I would replace both motherboard and CPU. I have been keeping my eyes open for the A8N32-SLI Deluxe (as many were), but decided I wanted to proceed with the upgrade this week while I am off work. Of course, two days after I decide to order an A8N-SLI Premium, the retailer gets the A8N32 for $40 (CDN) more. Now the dillema. Do I keep the Premium, or get the A8N32. Now I'm talking more than $40 because the e-tailer charges a 15% restocking fee, and I'll have to pay shipping. I'm probably better off just selling it privately, but I'll still lose as I'm sure I won't get the full amount I paid (even though the board is still new and sealed).
What I find frustrating is that Anandtech is the only site that I have found that shows the A8N32-SLI performing noticeably better than the regular nforce4 SLI. To make things worse, the only subsequent motherboard article they have is for the DFI Expert board, and they chose to go with a whole new review format for that article ( :| ). There are no figures there to compare to the A8N32-SLI review.
What makes the figures even more suspicious is the significant performance improvements for single cards at high settings. I understand that the newer chipset could be tuned slightly better, but any performance improvements there would be most evident at lower settings. At the tested settings in games like FEAR, Far Cry, and Quake 4, even the 7800GTX gets pushed pretty damn hard, and I honestly have trouble accepting that such huge differences are possible without there being some other variable.
I know that Wesley says the tests were rerun, but did anyone else double-check the results? Did anyone else double-check all system settings, particularly the performance settings within the Nvidia control panel? Did you reinstall everything from scratch on both systems, or just restart the systems and rerun the tests?
I'm also curious if Nvidia has done something in the 6.82 driver (only available for the NF4-SLI x16) that could also be tweaked in an updated driver for the regular NF4. Even then, I'd be curious as to what exactly such a tweak was doing to cause just a noticeable improvement.
What I find frustrating is that Anandtech is the only site that I have found that shows the A8N32-SLI performing noticeably better than the regular nforce4 SLI. To make things worse, the only subsequent motherboard article they have is for the DFI Expert board, and they chose to go with a whole new review format for that article ( :| ). There are no figures there to compare to the A8N32-SLI review.
What makes the figures even more suspicious is the significant performance improvements for single cards at high settings. I understand that the newer chipset could be tuned slightly better, but any performance improvements there would be most evident at lower settings. At the tested settings in games like FEAR, Far Cry, and Quake 4, even the 7800GTX gets pushed pretty damn hard, and I honestly have trouble accepting that such huge differences are possible without there being some other variable.
I know that Wesley says the tests were rerun, but did anyone else double-check the results? Did anyone else double-check all system settings, particularly the performance settings within the Nvidia control panel? Did you reinstall everything from scratch on both systems, or just restart the systems and rerun the tests?
I'm also curious if Nvidia has done something in the 6.82 driver (only available for the NF4-SLI x16) that could also be tweaked in an updated driver for the regular NF4. Even then, I'd be curious as to what exactly such a tweak was doing to cause just a noticeable improvement.
Facebook
Twitter