Have been running the Gigabyte 8IHXP for several weeks contiunously. Very stable & solid. Works best here with only 2 X256 modules Kingston 1066 rdram in slots 1 & 2, crimms in 3 & 4. (Of course, true 16 bit Samsung 1066 isn't available yet.) Currently, 2.53Ghz@2850 (fsb150), passes all benches, in 3DM, games & apps work fine. other Sys specs: 4 scsi hdds, LeadtekGF4, 2 X 256MB Kingston 1066-32 rdram, cd-rw, dvd-rom, Zalman AIR cool (temps-35-system/44-normal load/55-game load), etc.. Using latest F4 bios, NOT using EasyTune as bios has all OC options you need. (ET doesn't yet support all functions that it supposedly will in the future.) OC to fsb 156 is possible, but I find fsb 150 the limit for practical stability.
Having used Abit Th7II, used to solid performance, this board compares well, is equal to or better in general. The weak point for all 850E boards seems to be the rdram memory that's available for now. Even the Samsung 32 bit rdram is having problems with the Asus 32 bit rdram board. This is probably the best Rambus board out there for now, using the Kingston 1066 rdram (Toshiba), the only real 16 bit 1066 rdram to even consider. (Others are ripoffs. There is some Samsung rdram 800 being sold as 1066. The real stuff ain't here yet, don't be fooled!)
Content with the Gigabyte 8IHXP board, but hoping for more OC performance, when real Samsung 1066 gets here. If that rdram doesn't make for better OC performance, then I'll go for a DDR 400 board on another chipset path, & forget rambus. No more Intel chipsets for rdram announced.
It's hard to believe in the rambus future model now, with all the rdram memory manufacturing problems & delays. Board & memory choices using rdram are very limited!
Very happy I tried this board, but will not consider any others, until good true Samsung rdram memory appears. Since the latest Samsung DDR outperforms rdram 800 & comes close to 1066, even Intel (845PE/GE) will have a DDR-333 chipset after all, super 648-SIS-ddr board is on the rise & may excel, the rambus platform is fading.