The best motherboard for Brisbane 3600+

[buildster]

Hey Gurus,

For under $200, what is the best AM2 motherboard for a Brisbane 3600+? What kind of HTT Bus Speed can be expected?

Read here that some under $100 boards can do quite well. Just wonder whether more money can buy faster speed.

This OC noob appreciates your advice!

[buildster]

If not the best, is there any good MB for Brisbane under $200?

Anybody?

[funnyperson1]

I would stick with the sub $100 choices like the TForce550 and DFI Infinity Ultra II M2.

Imho buying a $60 CPU and spending 3 times that on a mobo makes no sense, especially when the sub $100 boards get very close to maxing out the chip.

If you really want max performance for money, spend the extra money you want to spend on a mobo on getting an F3 Windsor (which can hit 3.4-3.5 on high end air as opposed to the max of 3.0-3.2 with the Brisbanes), or better yet get a low end Core2Duo system that will rock either in performance.

To answer your question, it does seem like boards such as the M2N32SLI and Foxconn 590 seem to hit higher HTTs easier than the midrange boards, but the popular midrange boards do 300+ HTT easily and can do 320+ with the right settings and components.

[SPQQKY]

Yup, on a 570 board, you would probably see a slight cpu and memory performance gain over the 590 boards, but you would see a slight gfx advantage on 590.
Like funny person said, most budget boards will top around 300-320 HTT, 590's can go past that. It all depends on the cpu you end up getting (9.5x multi on 570 would cap you at approx. 3.1GHz +/-) when it could possibly go past that on a 590 board.
Seriously, a 570 or NF4 board in the $90-140 range is all you need for that chip. I had pretty good luck with a brisbane 3600+ on my M2N-SLI D, DFI Infinity NF UltraII-M2 and DFI NF SLI-M2.
A Biostar T series would make a great board with that chip as well.

[SnipingWaste]

I picked up a Gigabyte GA-MA69G-S3H (rev. 1.0) for $86.99 from newegg form my fathers upgrade. The nice thing during some testing was I was able to hit stable in windows running super PI HTT 420. The realy nice thing is it did this from BIOS settings and not from a clock gen program and I only played around in BIOS for 20 min to hit the HTT 420.

[buildster]

Thank you, funnyperson1 and SPQQKY for your suggestions and analysis! Hmm... mid-range boards are good enough Brisbane.

Appreciate SnipingWaste's testing sreens! Would it be nice to set multipler to 8, making it 420x8 = 3360? What stopped you from doing that?

[funnyperson1]

Probably his chip, these seem to max out at 3.1-3.2 on air with the good ones. Although sometimes HTT max speeds with low multipliers are deceiving.

I know with my board I was able to do 350+ HTT with 4X multi, but with 9.5 I couldn't go past 307. Of course then I realized that the 400 memory divider was for some reason holding me back and going up to 533 allowed me to hit 316+ stable, and more where the chip was limiting me.

[SnipingWaste]

Thank you, funnyperson1 and SPQQKY for your suggestions and analysis! Hmm... mid-range boards are good enough Brisbane.

Appreciate SnipingWaste's testing sreens! Would it be nice to set multipler to 8, making it 420x8 = 3360? What stopped you from doing that?

I bought this for my father for a upgrade so I didn't have much time with it. Im going to upgrade my system in a month with the same mother board and Ill do some hard core overclocking then. My fathers setup is now running a 3600+ at 2.4gig now at stock voltage with no temp rise from stock speeds.

[SnipingWaste]

Probably his chip, these seem to max out at 3.1-3.2 on air with the good ones. Although sometimes HTT max speeds with low multipliers are deceiving.

I know with my board I was able to do 350+ HTT with 4X multi, but with 9.5 I couldn't go past 307. Of course then I realized that the 400 memory divider was for some reason holding me back and going up to 533 allowed me to hit 316+ stable, and more where the chip was limiting me.

Don't ever use 4 multi when trying to find max HTT. 4 multi is a bug and the HTT reading will be wrong and will be really lower then what is shown. to test you need to be higher then 4 multi like 5 multi and up.