Asus A8V

[Sierra_abx]

Originally Posted by xgman

Are you going to try the vmod? I haven't given up on that yet.

I'm interested in that VDimm device that Bigtoe has been showing off. One of my biggest priorities right now is getting more juice to my BH-5 sticks. As you know, the board only goes up to 2.8v and that just isn't enough.

By the way, Anandtech just posted a review of the Corsair XL. Link. The reviewer hit 225 at 2-2-2, and 250 at 2.5-3-3. I'm wondering how OCZ's 3500EB and 3700EB will perform with this board?

[Sierra_abx]

Just some basic tips on memory settings in the BIOS. Disable 2T command rate. Also, the divider settings are a little confusing at first. In order to run CPU/memory in sync, use 2:1 DDR400. Don't use 1:1 DDR200.

221-222 HTT at 2-2-2-5 remains the best settings for me. The system feels lightning quick. Below is the highest FSB (HTT) I've managed to get so far. Running 5:3 divider and timings at 2-3-2-5.

[xgman]

I don't think using asny sort of dividers makes any sense for us. like you have shown if you can get 2-2-2-5 or close to that at 230fsb. You can't beat that with 9x270 with 5:3. At least I wouldn't think so. I still need more voltage. I may try the vdim mod later a bit more carefully. I think I shorted something on the "read" chip.

[Sierra_abx]

Good luck, xgman. And you're right. Dividers are not the way to go, especially with the FX. Very high HTT makes for impressive screenshots, but you aren't really gaining anything. Bigtoe suggested running a high TRAS. I tried 2-2-2-11 and 2-2-2-9. Times were about a second slower in Pifast than 2-2-2-5.

[xgman]

That was the old way to run the NF2. 2-2-2-11.

[Blue078]

Oh so this is were you guys are hanging out now!

Looking good guys..nice to see some people that understand that lower FSB, with tighter timings and higher multi's work well.

[Sierra_abx]

One important thing needs to be mentioned; the question of whether the AGP/PCI lock works on this board. Anandtech's Wesley Fink (aka Prometheus) had these comments about the Hothardware review of the A8V:

"I am really shocked the reviewer barely mentioned the missing PCI/AGP lock on the Asus A8V Deluxe and the problems with the premier VIA K8T800 PRO feature - the addition of a PCI/AGP lock. I have had an Asus A8V for several weeks and I have been working closely with Asus to fix this problem. The board also has had serious stability issues with overclocking an FX53, which have been improved if not fixed with the latest hardware upgrade and BIOS revision.

I have been able to reach 265 FSB on the 2nd revision of the A8V with the latest BIOS that improves FX53 compatibility and OC problems. The nF3-250 Ultra closes the gap with VIA as you can see in my chipset launch review at AnandTech on June 2, and the AGP/PCI lock does work NOW on every nF3-250 board I have tested - Socket 754 or Socket 939. The same can not be said for VIA, although the situation is definitely improving.

The revised Asus adds 2 resistors that enable the PCI/AGP lock. My review will not go up until Asus has fixed the last remaining problem - a serious stability issue with just the FX53 CPU. The 3800+ has no such problems on the A8V. This is shaping into an excellent Socket 939 board, and Asus has made giant leaps in resolving the issues with the A8V Deluxe."
-----------------------------------------------------------
Bigtoe has a modded BIOS that addresses the FX53 problem. Link. You have to register at Bleedinedge.com in order to gain access to the BIOS.

[xgman]

Got frustrated and took the IHS off. Dropped temps 5 degrees. It was a little hairy though trying to get that thing off. Also tried some Mushkin level2 BH5's. Didn't make much difference. Vdim mod still blocked over 1.875, so I took it off. Something really odd going on when I up the vdim voltage over 1.875 or so. Usually won't post no matter which ram is in, and I know the BH5 takes more than that with no sweat.

[xgman]

Originally Posted by sierra_bound

One important thing needs to be mentioned; the question of whether the AGP/PCI lock works on this board. Anandtech's Wesley Fink (aka Prometheus) had these comments about the Hothardware review of the A8V:

"I am really shocked the reviewer barely mentioned the missing PCI/AGP lock on the Asus A8V Deluxe and the problems with the premier VIA K8T800 PRO feature - the addition of a PCI/AGP lock. I have had an Asus A8V for several weeks and I have been working closely with Asus to fix this problem. The board also has had serious stability issues with overclocking an FX53, which have been improved if not fixed with the latest hardware upgrade and BIOS revision.

I have been able to reach 265 FSB on the 2nd revision of the A8V with the latest BIOS that improves FX53 compatibility and OC problems. The nF3-250 Ultra closes the gap with VIA as you can see in my chipset launch review at AnandTech on June 2, and the AGP/PCI lock does work NOW on every nF3-250 board I have tested - Socket 754 or Socket 939. The same can not be said for VIA, although the situation is definitely improving.

The revised Asus adds 2 resistors that enable the PCI/AGP lock. My review will not go up until Asus has fixed the last remaining problem - a serious stability issue with just the FX53 CPU. The 3800+ has no such problems on the A8V. This is shaping into an excellent Socket 939 board, and Asus has made giant leaps in resolving the issues with the A8V Deluxe."
.

Well that ****** me off and gives me hope, but how the hell do we know which boards are effected. It sounds like all current boards have this defect. What do you think? In all likelyhood if there is a NF3 250 939 board that kicks ***, I may try it. I imagine the Abit AV8 is in a similar state as the Asus. Guinea pig again.

[xgman]

How can we get a straight answer on which boards are effected?

[Sierra_abx]

Hmmm, that's not good about the VDimm. By the way, I left you a PM over at XS. As for the lock issue, we'd have to know which resistors are involved, so we can check our boards. My guess is that our boards don't have those two resistors. But so far, my RAM is acting the way it should. Even when using a divider, the memory maxes out around 225 which is what I expected.

Yes, I enjoy being a guinea pig. lol. By the way, I think the X800XT I pre-ordered at Best Buy is going to ship soon. My order status changed today. A new video card will cheer me up.

[bmg]

Originally Posted by sierra_bound

The revised Asus adds 2 resistors that enable the PCI/AGP lock. My review will not go up until Asus has fixed the last remaining problem - a serious stability issue with just the FX53 CPU. The 3800+ has no such problems on the A8V. This is shaping into an excellent Socket 939 board, and Asus has made giant leaps in resolving the issues with the A8V Deluxe."
-----------------------------------------------------------
Bigtoe has a modded BIOS that addresses the FX53 problem.

Is the updated bios all that's needed to solve the FX53 problem or do hardware changes need to be made?

[Sierra_abx]

Hey there, bmg, and welcome to the thread. After I flashed to Bigtoe's latest BIOS, everything seemed to be fine. I haven't had any stability problems. Knock on wood. I don't know what to tell you about the resistors because I don't know which two were added on later boards. By the way, I just got home and received an email from Best Buy. My X800XT has been shipped. Intend to do some serious benching this weekend.

Edit: well, maybe I won't get my card by this weekend. I checked my UPS tracking page. Card is being shipped ground with expected arrival 6/21. I should have upgraded the shipping. But too late now.

[cmay119]

Well well, sierra bound I see you've got everything up and running now. This A8V which I personally have not heard all that much about seems like a very nice board. Tell me, how does this board compare to the SK8V that you were using before? Any major improvements in any areas? I hope you like it, and congrats. :wave2:

[Sierra_abx]

Hi cmay119. The main difference between the two boards is that the A8V does not required registered RAM. There is an issue with the AGP/PCI lock. But a BIOS update seems to have fixed that. I'm 99.9% certain the lock is working now on my board. The next issue is getting more VDimm. Xgman says he can't get above 2.875v, even with a vmod. This could be problematic since my BH-5 sticks need more voltage.