Asus 680i, P5N32-E Sli Results.

lopri,

you cant compare Allendale to 4 meg conroe scores, they arent comparable regaurdless of the chipset used.

Tony,

you are 100% right.
900 mhz at 1T is definatly a new level to look into for ram manufacturers.

what about making them dual rated?
like PC8500 cas 4-4-4-12 at 2T, then 900, cas 4-4-3-8-1t?
that would be a interesting way to try to start getting 1T back into the market.

I have yet to actually do any game testing, but 1T in super pi is definatly fast. but running ram past 2.2 to 2.3 volts for long periods of time, is definatly not a smart thing, which I definatly agree with.

Quote:

---

Originally Posted by Kunaak

lopri,

you cant compare Allendale to 4 meg conroe scores, they arent comparable regaurdless of the chipset used.

Tony,

you are 100% right.
900 mhz at 1T is definatly a new level to look into for ram manufacturers.

what about making them dual rated?
like PC8500 cas 4-4-4-12 at 2T, then 900, cas 4-4-3-8-1t?
that would be a interesting way to try to start getting 1T back into the market.

I have yet to actually do any game testing, but 1T in super pi is definatly fast. but running ram past 2.2 to 2.3 volts for long periods of time, is definatly not a smart thing, which I definatly agree with.

---

Kunaak

do you think the P5N32-E SLI is a better 680i mobo vs EVGA board?

Like I said, I got rid of P5W-DH so I can't run any more test on it. I just picked up something close from my old pics. On this board (EVGA 680i) I get better dual SP with Allendale than with Conroe. Dual 32M with E6600 @3.6GHz = 19 min-ish, Dual 32M with E6400 @3.8GHz = 17 min-ish.

How high you can run voltage daily depends on the IC also... I doubt anyone is looking elsewhere but Micron chips today, and I think they're pretty resistant to high volts... I'm running mine 2.1v daily, but 2.3v-2.4v doesn't sound really bad if you can keep it cool.

Quote:

---

Originally Posted by saboya

How high you can run voltage daily depends on the IC also... I doubt anyone is looking elsewhere but Micron chips today, and I think they're pretty resistant to high volts... I'm running mine 2.1v daily, but 2.3v-2.4v doesn't sound really bad if you can keep it cool.

---

excellent question

Kunaak..one more question...do you have any sense to vcore flucuations on this mobo

Quote:

---

Originally Posted by lopri

Like I said, I got rid of P5W-DH so I can't run any more test on it. I just picked up something close from my old pics. On this board (EVGA 680i) I get better dual SP with Allendale than with Conroe. Dual 32M with E6600 @3.6GHz = 19 min-ish, Dual 32M with E6400 @3.8GHz = 17 min-ish.

---

I found your thread at AT very interesting . AMD X2> C2d with G80gtx at high res. The offical review I seen of AMD X2 with G80gtx did not show those results.
Yet a C2D @ 3.6 got its ass handed to it by a 3.0ghz X2. Could you show those results here as many of us would be interested in seeing those results.

Quote:

---

Originally Posted by Tony

Edit..533 ram 1:1 with a 6400 3-3-2-4 1t 1.8V was as fast as 5-4-4-15 2T DDR800 1.8V
T

---

Now that`s the most important thing I`ve been looking for.
Thanks Tony :toast:

IXBT in their review showed that this board doesn`t take performance hit with 533 ram and 1:1. Wonder how well it does with higher FSB ?

Kunaak
can you please compare memory bandwidth/latency with async/sync ratios and memory at 533/667 MHz ? :)

Quote:

---

do you think the P5N32-E SLI is a better 680i mobo vs EVGA board?

---

yes and heres why.

BFG, EVGA, ECS, and other 2nd tier companies have only bought default reference Nvidia 680i boards from Nvidia, and simply slapped thier product name on the box.

Companies like MSI, Asus, Gigabyte and DFI are tier1 partners that are taking the reference 680i board, and making it thier own, redesigning it, and making it better to suit thier needs.
Anandtech had a interesting little article on this, in thier review of the 680i chipset if your interested in reading more.

Quote:

---

Kunaak..one more question...do you have any sense to vcore flucuations on this mobo

---

Vcore Droop itself is bad, when at idle you can get 1.4 volts, to drop as low as 1.37 volts.
but thats still better then my P5B Deluxe was doing.
so yep, definatly another board badly in need of a Vdroop mod.

the Vcore itself, that too is horrible...
set 1.6 volts in bios, reboot, and you get 1.47 volts max.

so again, definatly a board in need of a Vcore mod.
this is the worst aspect of this board, the Vcore, is in serious need of some help as its really really bad on this board.

but... no worse then the usual Intel board.
never had a 775 board that didn't need a VDroop mod yet...

Quote:

---

Kunaak
can you please compare memory bandwidth/latency with async/sync ratios and memory at 533/667 MHz ?

---

sure, just give me awhile to think of how to actually test that...

Kunaak,

Do you happen to have a 4MB CPU? I ran SP with E6400, and confirmed your finding. This board indeed is super fast. But there is a catch and I think it's got to do with L2 cache.

DDR2-1140 / 4-4-4
http://img136.imageshack.us/img136/5149/wtfgj5.png

With Xeon 3060 (E6600) @3.6GHz, single 32M barely makes it into 15 mins. Normally Allendale can't make up half the L2 cache, even with 200MHz clock speed advantage. But with this board, it's actually faster than Conroes?!

Unfortunately my board has FSB holes that prevent the E6400 from 3.5~3.7GHz, so I can't do an apple-to-apple comparison. If you have an access to 4MB CPUs, please run tests. I think we can get to the bottom of this.

I really... reaaaaaaally dont want to have to take my E6600 out of this P5B Deluxe to put it in the 680i....

thats a whole lot of insulation and time I had to put into the thing to get it nice and sealed up and air tight, which isnt easy with how little room there is around the CPU socket for sub zero cooling....

I tightened subtimings and got 13 mins 48 secs. :eek: I'm not sure what to make out of this because I could never be under 14 mins with Xeon 3060 on this board, at any frequency. I posted the result in the 680i overclocking thread.

maybe the problem is that the bios just doesnt recognize Xeons correctly....

Quote:

---

Originally Posted by Kunaak

high learning curve will definatly get alot of people posting "my board cant do 500 FSB"... not cause the board can't, but cause they dont take the time to figure out the board. it's definatly a board for the advanced types.

---

How 'bout some tips to point people in the right direction? :)

Quote:

---

Originally Posted by SLi_dog

How 'bout some tips to point people in the right direction? :)

---

Think the user need to spend more time in fine tuning the memory timing to make the system stable. This is my impression to this Mb after going through this thread..

Kunaak, I have a question as well. I've sold my EVGA board and am deciding between the P5N32-E and the striker extreme. I've also sold my ram and picked up a set of corsair c4 dominators pc 6400. In the specs listed on the asus site, the striker is listed in specs and directly supporting SLI memory (Which all the corsair c4 kits are) up to 1200mhz wheras the P5N32 lists it with an asterisk (at the bottom asterisk says specs subject to change).

Does the P5N32 support EPP profiles/sli ready ram? My understanding is that all the 680i boards do but the description on the asus site is ambiguous at best. I'd prefer to save $100 on the motherboard if the P5N will support the profiles.

Thanks

My board arrived DOA today :(

Now I gotta re-assemble th P5B-Dlx...Newegg won't refund it either so I have to wait until it comes back into stock.

Quote:

---

Originally Posted by ACE76

My board arrived DOA today :(

Newegg won't refund it either so I have to wait until it comes back into stock.

---

Check the stock now!!!

If someone with an Allendale can post up 32M at 450x8 4-4-4 2T then I will do same for X6800. I think Lopri might have something here, my X6800 is minutes slower for 32M than it was on 975.....

Regards

Andy

any chance this board would take the striker bios, to get the extra vcore options? from the looks, they're almost identical with the striker having the lcd and extra bling

Quote:

---

Originally Posted by Kunaak

did 14 minutes and 21 seconds in 32M with my E6300 at 7 x 500, 1300 mhz ram, cas 5-4-4-10.

how does that compare to your 32M at the same speed?
cause its far faster then I ever did with my P5B Deluxe.

http://www.xtremesystems.org/forums/...1&d=1163709626

---

Great thread man :toast:

that's a FAST time for sure >>> better than 965

my run with 475MHz 1:1 on RAM was just sub-14m ......i'll be testing the 680i for SPI performance and RAM tweaking...looking forward to it................

http://img95.imageshack.us/img95/133...7187swmqn3.jpg

Can a Scythe Ninja Original be mounted on the P5N32-E (i.e. does it not have those annoying things on the backside that prevent a backplate-having HSF)

I was going to get the eVGA 680i, but after reading that+ the bugs, I'm glad I'm going for the Asus.

I have got to stop reading these threads, they are making me more and more tempted to start a conroe build.

Quote:

---

Originally Posted by DCarez

Can a Scythe Ninja Original be mounted on the P5N32-E (i.e. does it not have those annoying things on the backside that prevent a backplate-having HSF)

I was going to get the eVGA 680i, but after reading that+ the bugs, I'm glad I'm going for the Asus.

---

By annoying things are you referring to caps? Go to this Newegg product page:

http://www.newegg.com/Product/Produc...82E16813131073

One of the pics is the backside of the board but you'll just have to trust Newegg that is a true image of what it looks like. I don't think it's the backside but the heatpipes that might interfere with some hsf.

Kunaak, nice results, any pointers for or against the Striker vs P5n32-e sli ?

Nice!