Speedfan on P5B?

korsch · Post by **korsch** » Mon Aug 21, 2006 2:55 am

I am upgrading my system and trying to decide on a motherboard to go with an e6400 or e6600, primarily between the Asus P5WDH and P5B Deluxe.

Software fan control is one important factor. Speedfan seems to controls the P5W very well. But, according to what I read, on the P5B Speedfan does not currently read the motherboard and/or CPU temperature sensors.

Can someone confirm what if anything Speedfan can do with this board? If true about the limitations, I wonder if a Speedfan update is likely to enable proper monitoring of temps and control fans in the near future. I've looked but never found a discussion forum at the Speedfan site.

cmthomson · Post by **cmthomson** » Mon Aug 21, 2006 11:03 am

Another thing to consider: the P5B does not let you control DRAM parameters. Check the BIOS pages in the manual (download from ASUS) before buying.

I'm happy with my P5W DH. SpeedFan works great, and can directly control up to five 3-pin fans in three PWM groups. The CPU header can run a 4-pin or 3-pin fan directly and has DC or PWM options.

The two boards are about the same price, although availability of the P5W DH is poor.

smilingcrow · Post by **smilingcrow** » Tue Aug 22, 2006 3:58 pm

Likewise, I found the fan control on the P5W DH Deluxe to be exemplary. Going back to a lesser setup is a hardship

korsch · Post by **korsch** » Tue Aug 22, 2006 8:57 pm

I found a similar thread at the xtreme forums with a revealing post that confirms the problem with the P5B but implies future Speedfan support is likely. It's here if anyone else is interested:

http://www.xtremesystems.org/forums/sho ... p?t=112452

It's encouraging that the P5B will probably have full if not better support of Speedfan, but I'm not sure how patient I am to wait for this. The only advantage of the P5B for my purposes is that it seems, thus far, to be the best motherboard for overclocking e6400 and probably e6600 chips. On the downside for me, there is the lack of manual software fan control, support for only 2 as opposed to 4 PATA devices, and the possibility that a PCI slot will be blocked by third party video cooling due to the slot layout (looks not to be a problem on the P5W).

The decision would be much easier if I didn't suffer from split-personality-computing, trying to overclock and suppress noise at the same time, with the two sides ever warring in my head and the two objectives at cross-purposes in my computer design.

The P5W probably overclocks enough for me while offering the better feature set for my needs. I'm almost decided...

smilingcrow · Post by **smilingcrow** » Wed Aug 23, 2006 2:29 am

korsch wrote:The only advantage of the P5B for my purposes is that it seems, thus far, to be the best motherboard for overclocking e6400 and probably e6600 chips.

The decision would be much easier if I didn't suffer from split-personality-computing, trying to overclock and suppress noise at the same time, with the two sides ever warring in my head and the two objectives at cross-purposes in my computer design.
The P5W probably overclocks enough for me while offering the better feature set for my needs. I'm almost decided...

I found that the P5W DH could over-clock an E6400 to the point where the heat produced was becoming an issue to cool quietly. This was at 3.2 GHz with a Ninja & Nexus fan, details here.
Your noise threshold may be higher than mine though.

I think the P5B suffers from the same issue as all P965 based boards in that it runs the memory no slower than the FSB. This means that for a large over-clock you need fast RAM. The P5W DH can run the RAM slower than the FSB speed so you can push it further when using DDR2-667.

korsch · Post by **korsch** » Wed Aug 23, 2006 7:36 pm

Thanks, that's very useful info.

ijozic · Post by **ijozic** » Thu Aug 24, 2006 4:57 am

It doesn't seem to be an issue for you, but if you don't have a good audio card, the one on P5W (Realtek ALC882M) seems much better than the one on P5B (ADI Soundmax). I have this ADI solution on M2N-SLI and it's crap, at least with the current drivers. Some MP3s are great, some are rather distorsive and 3D Audio in America's Army is crap. Enough to drop it off in favour of CL Live.

korsch · Post by **korsch** » Thu Aug 24, 2006 1:57 pm

ijozic wrote:It doesn't seem to be an issue for you, but if you don't have a good audio card, the one on P5W (Realtek ALC882M) seems much better than the one on P5B (ADI Soundmax). I have this ADI solution on M2N-SLI and it's crap, at least with the current drivers. Some MP3s are great, some are rather distorsive and 3D Audio in America's Army is crap. Enough to drop it off in favour of CL Live.

I am trying to consider all the pros/cons so this is a good point. I do still have a Turtle Beach Santa Cruz card that might be moved to the new system but either way the better sound is another plus on the P5W side.

If the bloody results I keep seeing for the P5B with e6400s were not so bloody good this would be much easier for me.

Anvar · Post by **Anvar** » Fri Aug 25, 2006 1:48 am

Just to up the ante on the P5B, ASUS just released a new BIOS, and the guys at Anandtech are loving it.

Seems like this is the one I'll get.

Mariner · Post by **Mariner** » Fri Aug 25, 2006 1:54 am

Incidentally, I've just seen this article at Anandtech which talks about a new BIOS for the P5B which allows unlocking of the core multiplier.

Of course, Anandtech use this ability to see how high they can clock the chips but I'm sure this could also be useful for an undervolting/underclocking combo to help silence a PC.

Edit: Beaten to it by Anvar!

Anvar · Post by **Anvar** » Fri Aug 25, 2006 2:00 am

Yay me

smilingcrow · Post by **smilingcrow** » Fri Aug 25, 2006 6:20 am

Mariner wrote:Incidentally, I've just seen this article at Anandtech which talks about a new BIOS for the P5B which allows unlocking of the core multiplier.

Of course, Anandtech use this ability to see how high they can clock the chips but I'm sure this could also be useful for an undervolting/underclocking combo to help silence a PC.

You can set the multiplier on a C2D using Windows utilities such as RMClock on probably all motherboards, provided they support EIST. Almost all boards do support EIST and you donâ€™t really want a board without it anyway.

The advantage of being able to set the multiplier in the BIOS is for situations where you want to run with a very high FSB. In this case it can be impossible for the system to load Windows if using the stock multiplier.
e.g. with a E6600 which has a multi of 9:

500 * 6 = 3.0 GHz
500 * 7 = 3.5 GHz
500 * 9 = 4.5 GHz

The first two are workable, the third is not, at least with air cooling.
If you can load Windows at 4.5 GHz, you can then lower the multi using RMClock for hopefully a stable system.

cmthomson · Post by **cmthomson** » Fri Aug 25, 2006 2:35 pm

A bunch of folks on the overclock forums are in a mad race to see who can get the highest FSB frequency. This is of course not a very useful indicator of overall system performance. Typically they have to choose asynchronous and slow memory speeds to run higher front side bus speeds, and this is counterproductive.

Also, lowering the multiplier to get a higher FSB clock is really just running in place.

From an SPCR perspective, the most important parameter is vCore, which determines the CPU power consumption, and thus the fan noise. This voltage is dictated more by the CPU core clock rate than by anything else.

cmthomson · Post by **cmthomson** » Fri Aug 25, 2006 2:50 pm

Anvar wrote:Just to up the ante on the P5B, ASUS just released a new BIOS, and the guys at Anandtech are loving it.

Seems like this is the one I'll get.

Before you get too excited about this, read the last page of that post:
http://www.anandtech.com/mb/showdoc.aspx?i=2822&p=4

The conclusion is that with this BIOS, the 965P-based P5B is now almost as good as the 975X-based P5W DH.

Also look closely at that tabe of multipliers and frequencies. Every one of the lower multipliers decreased the CPU core speed compared to the original locked multiplier! The only advantage to the lower multipliers and higher FSB speeds (other than forum bragging rights) was the ability to better match high-performance memory that the emasculated 965P normally precludes.

Get a P5W. It's definitely worth the few extra dollars.

cmthomson · Post by **cmthomson** » Fri Aug 25, 2006 2:55 pm

Mariner wrote:Of course, Anandtech use this ability to see how high they can clock the chips but I'm sure this could also be useful for an undervolting/underclocking combo to help silence a PC.

Of course, if that's your primary concern, get a 6300, run it at 6x, and save a bunch of money too!

NoiseFreeGuy · Post by **NoiseFreeGuy** » Mon Aug 28, 2006 7:25 pm

cmthomson wrote: Get a P5W. It's definitely worth the few extra dollars.

But the P5W doesn't have 'AI Nap' which is way cool, nor an LPT port; two features the P5B has which are important to me.

korsch · Post by **korsch** » Wed Aug 30, 2006 12:37 am

cmthomson wrote: The conclusion is that with this BIOS, the 965P-based P5B is now almost as good as the 975X-based P5W DH.

Anandtech's conclusion, true, but elsewhere opinions differ. At xtreme forums, it seems that others are having far more luck overclocking the retail e6400 and e6600 chips on the P5B to 3600-4000mhz+ with minimal effort and sometimes minimal over-voltages. Whereas most people seem to be stuck in he 3200mhz-3400mhz range with the P5W even with extreme voltages. Can the P5W make up 400+mhz in overall performance with any clock-for-clock advantages it might have (lower latencies etc.)?

If indeed a new BIOS comes out for that addresses what people seem to think are the factors limiting FSB on the P5W, including memory timing options and the lower multipliers, then maybe this overclocking gap will close.

I still weigh the other advantages of the P5W a little greater at the moment, but I have another week or so until my other parts arrive to decide.

cmthomson · Post by **cmthomson** » Wed Aug 30, 2006 7:15 am

korsch wrote:
cmthomson wrote: The conclusion is that with this BIOS, the 965P-based P5B is now almost as good as the 975X-based P5W DH.
Anandtech's conclusion, true, but elsewhere opinions differ. At xtreme forums, it seems that others are having far more luck overclocking the retail e6400 and e6600 chips on the P5B to 3600-4000mhz+ with minimal effort and sometimes minimal over-voltages. Whereas most people seem to be stuck in he 3200mhz-3400mhz range with the P5W even with extreme voltages. Can the P5W make up 400+mhz in overall performance with any clock-for-clock advantages it might have (lower latencies etc.)?

If indeed a new BIOS comes out for that addresses what people seem to think are the factors limiting FSB on the P5W, including memory timing options and the lower multipliers, then maybe this overclocking gap will close.

I still weigh the other advantages of the P5W a little greater at the moment, but I have another week or so until my other parts arrive to decide.

On a bunch of the OC threads I've read, the goal seems to be to post the highest FSB clock, or highest percentage of overclock, or highest voltage, or highest fan RPM, or whatever. Other than superpi, many of them don't even run benchmarks! And a lot of them quote numbers with the comment "it will POST but won't boot Windows". Wow, that's useful.

One thread I've found interesting is this one:
http://www.hardforum.com/showthread.php ... ge=1&pp=20
where only stable configurations (defined as running Orthos for 4+ hours) are accepted. P5W and DS3 boards way outnumber P5Bs. All the chips max out at around 3.6 GHz under this criterion.

I dialed my 6600 back to 3.1 GHz so I could lower the voltage and turn the fans way down. The highest stable CPU speed I got was just under 3.6 GHz at ridiculous voltages and fan speeds.

Note that to successfully overclock a P5W DH, you need to disable Hyper Path 3. This straps the 975X to 1066 (instead of the overclocked 800) and allows much higher FSB clocks.

noac · Post by **noac** » Wed Aug 30, 2006 1:26 pm

I don't know if anyone has posted this link, Im a bit tired:

http://www.legionhardware.com/document.php?id=580&p=0

It compares the two mobos in question. Thought it could be helpful.

korsch · Post by **korsch** » Wed Aug 30, 2006 1:30 pm