My SB750, HD4850 x-fire, 9850, H20 Build...

[SmaKKed]

Nice results so far fellas, anyone going to show some gaming benchies?

[oldpueblo]

I will take some pics if you still need them...

Naw, I think I've got it based on your earlier posts. I think I'm just dealing with a fluke, I'll try reinstalling and other odd tricks to see what's up.

[Oliverda]

Naw, I think I've got it based on your earlier posts. I think I'm just dealing with a fluke, I'll try reinstalling and other odd tricks to see what's up.

http://my.ocworkbench.com/2008/gigab...290%20copy.jpg , http://my.ocworkbench.com/2008/gigab...291%20copy.jpg

Have you pressed the ctrl+F1 in the BIOS?

[Rammsteiner]

Nice going charged!

You said the Phenom first only did 2.8Ghz and was hardly stable. Were you actually able to clock it to say 2.7Ghz and let it be 100% stable (so no idle freezing either?).

My CPU couldnt boot really above 2.8Ghz. Sometimes I was lucky and managed up to 2.9Ghz, but when pressing enter to log on everything froze (no matter what Vcore, usual SB600/Phenom problem).

I actually got even at 2.6Ghz idle freezes, then gave up and let everything run at stock.

For the Asus M3A78-T board, i think I read somewhere the BIOS wasn't 100% yet, if I didnt read it, I most certainly know it anyhow.

[Particle]

I'm going to pop the 9850 out of my box again later and give it another go I think. I'll see what it can do at 1.400V and with ACC set to Auto. I hadn't tried that.

[keithlm]

I'm going to pop the 9850 out of my box again later and give it another go I think. I'll see what it can do at 1.400V and with ACC set to Auto. I hadn't tried that.

Try uninstalling the AOD 2.1.2 and rebooting.


Then make sure ALL directories have nothing in them for the AMD/AOD stuff.

DIRECTORIES:
MAIN: C:\Program Files\AMD\AOD

C:\Users\USERNAME\AppData\Local\AMD\OverDrive in vista
(Or C:\My Documents\..... in XP.)

After those are clean you can then reload AOD 2.1.2 and hopefully it will work better. (I had that version crash until I did the above. Now it works fine.)

[Particle]

Perhaps I should clarify. I haven't had any AOD related crashes or problems. It is just that ACC (despite being enabled and showing so) wasn't getting me anything.

[keithlm]

Perhaps I should clarify. I haven't had any AOD related crashes or problems. It is just that ACC (despite being enabled and showing so) wasn't getting me anything.

Oops. Sorry... thought you had issues with that also.

Keep us posted on how it works... if you retry your 9850....


BTW: I think the problem with AOD on my original install before uninstalling and reinstalling was that the AOD service somehow did not get installed. (So anyone having problems... check and make sure you have that service.)

[bingo13]

What BIOS are you on? 0304 pretty much fixed all the issues we found, still a few performance items left to improve but overall the board is much better behaved. ACC was not working correctly in 0204.

Well, so far my experience with the ASUS M3A78-T has been pretty much pants. ACC offers me no improvement over no ACC. I'm still not stable above 2800MHz no matter where I adjust ACC to. I do like how this board's bios doesn't have to be cleared before each boot though. That's an improvement over my K9A2, but of course I still don't have all the fancy add-in cards on this board yet.

[Particle]

Interesting, bingo. I noticed that there weren't any BIOS updates available on the website, so I'm assuming 0304 is a beta somewhere? I'll fire up the ASUS again and give a look to see what BIOS revision it has currently.

[Particle]

Whaddaya know? My board has 0204. Where can I get the update?

[Oliverda]

Whaddaya know? My board has 0204. Where can I get the update?

Have you tried the ASUS Update Utility?

[Daveburt714]

What BIOS are you on? 0304 pretty much fixed all the issues we found, still a few performance items left to improve but overall the board is much better behaved. ACC was not working correctly in 0204.

I'd be very interested in this bios too Bingo... I've Googled but can't seem to find it anywhere.

Could you get us a link?
Thanks.... Dave

[charged3800z24]

Nice going charged!

You said the Phenom first only did 2.8Ghz and was hardly stable. Were you actually able to clock it to say 2.7Ghz and let it be 100% stable (so no idle freezing either?).

My CPU couldnt boot really above 2.8Ghz. Sometimes I was lucky and managed up to 2.9Ghz, but when pressing enter to log on everything froze (no matter what Vcore, usual SB600/Phenom problem).

I actually got even at 2.6Ghz idle freezes, then gave up and let everything run at stock.

For the Asus M3A78-T board, i think I read somewhere the BIOS wasn't 100% yet, if I didnt read it, I most certainly know it anyhow.

My best with the 9600BE was 2.6ghz 100% stable with stock v-core. My 9600 hated voltage..2.7 was worse then 2.8 actually lol...

[charged3800z24]

I am going to do som Cinebench, 3mark06, vantage, an anything else I can, tomorrow night when I get home.

[SmaKKed]

I am going to do som Cinebench, 3mark06, vantage, an anything else I can, tomorrow night when I get home.

Maybe some game benchies if you have any aswell

[charged3800z24]

[Boschwanza]

Cinebench 32 Bit realy sucks with the phenom. Bad Scaling etc.

Try 64 bit and try to get NB as fast as possible and you will get an additional extra boost in performance by near as 10 %

[charged3800z24]

Cinebench 32 Bit realy sucks with the phenom. Bad Scaling etc.

Try 64 bit and try to get NB as fast as possible and you will get an additional extra boost in performance by near as 10 %

I will load 64bit tonight if time permits.. But I don't see me achieving the same OC on the 64 bit... I already dicovered this on my other board.

Edit: I need a spell checker lol.. I type to fast and never proof read..wow

[Aquineas]

I hate to sound like a conspiracy theorist, but with regards to benchnmarks, I am now at the point where I want to know what compiler a commercial program was compiled with.

[charged3800z24]

I will post some other benchies if any one has any the want to see..... I will run 3.3 benchies tonight if I have time. I will throw in the 64bit ones after I get it loaded.. I might try hitting 3.4ghz..but it might take 1.5v eek

[Boschwanza]

I hate to sound like a conspiracy theorist, but with regards to benchnmarks, I am now at the point where I want to know what compiler a commercial program was compiled with.

Cinebench is highly Intel optimated thats not new, at least for 64 Bit operating System Cinebench is forced to use SSE2 on every processor. Thats one explanation the Phenom does much more better with 64 Bit in Cinebench except the better 64 bit implemetation on AMD CPUs

For example (Cinebench)
32 to 64 Bit on a Intel system gives you an extra boost about 10 %
32 to 64 Bit on a AMD system gives you an extra boost about 20 %.

[informal]

Plus Cinebench doesn't use vector SSE at all,but instead scalar SSE which has zero improvement going from K8 to K10(and why would they improve scalar SSE??).The main point in using SSE optimized code is to try to vectorize the apps so that they can use the full potential of SIMD.
Also i wonder what SSE path is used with non intel chips in Cinebench.Penryn gen. uses SSE4.1 optimized path,while Conroe is on SSE3/SSSE3 i believe.
look at Agner's post at aces' here for more on the complier CPUID rigging:

The CPUID instruction in modern PC microprocessors works in an awkward way. Originally, the CPUID gave 4 bits for family number and 4 bits for model number. This means that the maximum numbers are family 15 and model 15. When these numbers were exhausted, they added 4 more bits for the model number. The new 4 bits are concatenated with the old 4 bits to make an 8-bit number, so the maximum value for model number is 255, or FF hexadecimal.
It would be logical to do the same with the family number, but instead they have added 8 more bits called "extended family". The new 8 bits are not concatenated with the old 4 bits to make a 12-bit number. Instead they have specified that we must calculate the sum of the old 4-bit "family" number and the new 8-bit "extended family" number. This means that the same family number can be specified in more than one way - and I think I know why!

Here's my theory:
Intel have made a compiler to support the ever growing extensions to the instruction set. The Intel compiler puts a CPU-dispatcher into your code to check whether the CPU supports the SSE2, SSE3, SSE4 or whatever instruction set. The compiled program can contain more than one version of critical parts of the code, and the CPU-dispatcher automatically chooses the version that fits the available instruction set. So far so good. The problem is that the CPU-dispatcher makes its choice based on family numbers and not only based on the feature bits that tell whether SSE2 etc. is supported. And it will not recognize unknown family numbers. The consequence is that any future Intel CPU with a family number different from 6 or 15 will not be recognized and will run with all SSE instruction sets disabled or will not run at all. There are lots of software on the market that is compiled with the Intel compiler. All this software would fail to run efficiently on a new Intel CPU with a family number different from 6 unless it is recompiled. The CPU-dispatcher checks only the old 4-bit family number. They can make the old family number = 6 in order to fool the CPU-dispatcher and then make the extended family number = e.g. 10 to make the sum = 16 or whatever number the marketing department dictates.

So the awkward implementation of the CPUID instruction is to make up for a serious blunder made by the people who designed the Intel compiler.

Funny that AMD have accepted this scheme, but they probably had no choice. BTW, the CPU-dispatcher in the Intel compiler also checks the brand name in the CPU and disables all SSE extensions if the brand is anything but Intel. See
http://www.agner.org/optimize/optimizing_cpp.pdf for how to circumvent this and make the code compiled with the Intel compiler work on AMD processors........

I just tried Intel C++ compiler version 10.1 with option /QxO as you suggested. It generates the following versions of code for common mathematical functions: SSE2, SSE3, SSE4.1 and non-Intel SSE2. It doesn't work on any CPU prior to SSE2. This is the only compiler option that makes it run reasonably on an AMD, but why are there two different SSE2 versions, one for Intel and one for AMD? When I hack the CPU-dispatcher and makes it believe that it is an Intel, it runs 50 - 100 % faster. This means that the Intel-SSE2 version is faster than the AMD-SSE2 version when running on an AMD processor!

There are also options that work on any processor. For example /QaxB. This options runs non-vectorized SSE2 code on Intel processors and old 8087 code on AMD processors. I measured this to be 5-10 times slower than the /QxO option on an AMD Opteron.

As you can see above,AMD is fighting an uphill battle here,it won't be easy at all.Especially when apps are complied with intel's compiler with above mentioned rigging in mind.

[LightSpeed]

performance rigging?

didnt know, honestly. its a shame a company such as intel would do such a thing

or is it the software companies?

[i found nemo]

maybe intel has influence on software companies. would not surprize me, tbh.