8RDA6+ Pro Romsip Mod

[elzhi]

http://home.comcast.net/~teckzilla/epox.PNG
hi,

What i did is bend the two capacitors back just a little to fit my booster in the
memory slot to allow dual channel crazy but worth it to me

The most volt i put to the memory is 3.9volts ,24/7 use i have it at 3.4volts

I got a bad batch of bh-5 there not stable over 250 even when i had my Dfi board,
Im sure if you have a good set of bh-5 you can hit 250 easy

PS.ocz booster is very picky about PSU my brand new Enermax460 doesnt work with it
but my Antec350 works fine

[MietmaN]

@ Runnel :

Are all those results you're posting the Everest's read test results or something else? I'm asking, because you've only mentioned the benchmark once - just in case ...

I wrote they depend on the multi - try to do the write test as well - as far as I know the write results increase with the multi, but I'm not sure about that. In my case they've increased much more than the read ones while using Merlin's bios, in comparison to the stock bios.

Am I wrong or you were able to run your mem in 2x512MB dual channel using the non-modded bios with CPC On? If you were indeed, than I think you should try out Merlin's CPC On thingy.

As to daily use I run at 11x200 V-core 1.50v, 2x512MB dual channel 2-3-3-11 - higher OCing is just for fun!
Whatever you choose I think you will not notice any serious difference.

@ Naibot :

It's the nForce2 itself, that's holding back your FSB. Dual channel is the first problem and 2x512MB of memory is the second and even more serious one. If you can push your system up to stable FSB235 with 2x512MB in dual channel mode, than you should be glad with it.

Try the CPC Off bios, but don't count on any noticeable performance increase unless you reach much higher FSB :

http://www.insanetek.com/index.php?page=cpcnf2

Mark that elzhi has 2x256MB configuration, so he must have less problems with high FSB speeds.

@ mrapples :

I'm almost sure it's your cpu - try to give him more voltage or simply get a better one.

@ all :

This thread has been really getting more and more interesting

[mrapples]

ill try some more volts, but i doubt that its the proc, its a 2600 with an iqyha stepping, and i honestly have never seen them do this poorly

btw, the 12v rails on my psu is 22A

[Runnel]

@ MietMaN

thanks for replying so quickly. answering your questions:

- yes, all are Everest's read test results.
- write-test with multi 9,5 and 11 I can do tonight - with the September stock-bios. using Merlin's bios again I'll wait for the hopefully fixed standby.
- yes: 2 x 512 MB MDT DS dualchannel (put in the sockets of the same color). yes: using the non-modded bios.
- "with CPC On?" Sorry, where is the switch? In "Advanced Chipset Features" I have "Memory Auto Precharge" 'enable'. "CPU Interface" I have 'optimal'. I found no difference to 'aggressiv'. So where is "CPC On"? - yes, next time I will try out 'Merlin's CPC On thingy'. (I heard that this is only for expensive memory, therefore I first tried the CPC Off version.)

Yeah the nice thing about this nForce boards is the possibility to adapt the fsb automatically. In daily life I switch with speedfan between fsb 133 and 233 (idle / load). multi 9.5, Vcore 1.47 v. 80 mm-cpu fan: 1200 rpm; 3 x 80 mm case fans, 2 @ 1050 rpm, one @ 1200 rpm. cpu-temp: 36° and almost no noise. beautiful!

The only pitty about the 8RDA6+ Pro is the fact that the board is not supporting the automatic fan speed control by speedfan.

[Wolf2000me]

There is no switch in the bios user interface to enable it. The only way to change the cpc setting is by flashing to the corresponding Merlin bios.
I do notice a difference by setting the cpu interface to agressive but i heard it makes the most difference when you're using tCAS 2.

[MietmaN]

Originally Posted by Runnel

(...)
- yes: 2 x 512 MB MDT DS dualchannel (put in the sockets of the same color). yes: using the non-modded bios.
- "with CPC On?" Sorry, where is the switch? In "Advanced Chipset Features" I have "Memory Auto Precharge" 'enable'. "CPU Interface" I have 'optimal'. I found no difference to 'aggressiv'. So where is "CPC On"? - yes, next time I will try out 'Merlin's CPC On thingy'. (I heard that this is only for expensive memory, therefore I first tried the CPC Off version.)

Hi, CPC On (1T) and CPC Off (2T) is sort of a hidden memory timing, which used to be selectable in bios on older VIA chipsets (Socket A) and you can find it also on AMD64 chipsets. When set to 2T it allows to reach much higher FSB (well, at least in theory) with tighter memory timings - is especially useful while using dual channel and 1GB (or more) of total RAM memory. A serious con is, that it decreases the overall memory performance, especially in 3D applications. In stock bioses it's always On. Here's the best explanation I could find :

http://www.insanetek.com/index.php?page=cpcnf2

The CPU interface does seem to afflict system performance in my case (again, 3D applications and benchmarks) but less than with Abit NF7 I've had. On my long dead Epox 8RDA3+ rev.2.1 it also allowed me to reach higher FSB when switched to 'Optimal', but again with performance loss. The 'Optimal' setting might have allowed you to do FSB233 with 2x512MB dual channel - I personally prefer lower FSB and the 'Aggressive' setting.

Yeah the nice thing about this nForce boards is the possibility to adapt the fsb automatically. In daily life I switch with speedfan between fsb 133 and 233 (idle / load). multi 9.5, Vcore 1.47 v. 80 mm-cpu fan: 1200 rpm; 3 x 80 mm case fans, 2 @ 1050 rpm, one @ 1200 rpm. cpu-temp: 36° and almost no noise. beautiful!

@ mrapples :
Well, it depends also on the week of production and on individual CPU you have - e.g. for AMD XP-M 2500+ IQYHA 0408 was one of the best weeks (the 8th week of 2004), but there was another one I can't recall, that was simply a tragedy - 2.4GHz with voltages about 1.9v.

Your 12v line seems to be fine with 22A, providing the PSU isn't one of those Codegen (Code****) or Megabyte ones.

[Runnel]

So, now I tested the memory performance til the end, my friend.

To compare with multi 9.5 (@ 1.475 V) with multi 11 (@ 1,675 V Vcore), the other settings I didn't touch (2,5/4/4/10 - optimal) (Epox-September Bios ((sorry for wasting so much space with the stock bios in Merlin's thread)).

To make it more easy to compare I each time start with the cpu-speed, followed by multi and fsb. EverestHome: read (MB/s)/write (MB/s)/delay (ns)

1580 MHz:
9.5 x 166: 2382/905/126,3 (+21%/+14%)
11 x 143: 1954/792/146,3
1710 MHz:
9.5 x 180: 2577/977/117.2
11 x 156: 2128/866/133,6
1900 MHz:
9.5 x 200: 2864/1087/105,2 (+21%/+14%)
11 x 173: 2356/ 957/120,8
1995 MHz:
9.5 x 210: 3007/1144/100,8
11 x 181: 2459/1006/115
2086 MHz:
9.5 x 220: 3136/1196/ 95,7
11 x 190: 2593/1053/109,7
2181 MHz:
9.5 x 230: 3285/1250/ 91,6 (+21%/+14%)
11 x 198: 2707/1095/105,2
2287 MHz:
9.5 x 240: 3450/1310/ 87,5
11 x 208: 2843/1147/100,2
2327 MHz:
9.5 x 245: 3515/BSOD (@ 1,475 V)
11 x 212: 2886/1166/98,6
2416 MHz:
11 x 220: 2995/1216/95
2504 MHz:
11 x 228: 3102/1261/91,7
2525 MHz:
11 x 230: 3130/1274/91,1 ... 9.5 x 220= 2080 MHz: read 3136
... for the same write-performance: 9.5 x 235 = 2234 MHz: write: 1283 MB/s.

Conclusion: The memory read/write performance with multi 9.5 is about 21%/14% better compared to multi 11.

What does that tell us?

[Wolf2000me]

Originally Posted by mrapples

ill try some more volts, but i doubt that its the proc, its a 2600 with an iqyha stepping, and i honestly have never seen them do this poorly

btw, the 12v rails on my psu is 22A

Your OS doesn't do anything out of the ordinary when you try a stable setting? Indeed 22A for the 12v rail should really do the trick. I'd say check your memory modules, try with one and try asynchronous (lower of course). If it's still the same deal you may indeed be out of luck with your cpu.

Runnel, i find this very strange. I'll test it out right now.
btw, are you quoting the doors, my friend?

edit:

I normally run at 12x12.5 @ 2.5-3-3-11, now i changed the multiplier to 9.5, I had to do that in the bios, changing multipliers on the fly makes my system hang and resets my bios o_O. Here are my results:

12x
3297 int
3028 float

9.5x
3247 int
2990 float

I'd say those are rather normal results when you lower your multiplier.
Of course i have the cpc1 bios.
What happens when you try single channel and/or one stick of memory? What you're experiencing is not normal. Perhaps someone with a cpc2 bios could test the same?

[mrapples]

nah, im in linux and it runs just fine at my stable settings

also, i only have one stick of 512

what if my psu wasnt stable enough, could that cause this?

[MietmaN]

Originally Posted by Runnel

So, now I tested the memory performance til the end, my friend.

Indeed, you did, well, I'm impressed

But I think the conclusion is somehow false : you can't say your memory performance with 9.5x multi is 21%/14% better than with multi x11. Why? Just because you've done all the tests at different FSB to keep the CPU at the same speed, but not the memory, which was the thing you tested. CPU FSB = Memory FSB, unless you run in asymetric mode (makes no sense on nFroce2). The right conclusion in my opinion is, that :

stock Epox September bios is simply optimized for lower multipliers (probably 9-10x).

Your results with 11x230 match those of 9.5x220 - you used the same timings, so here is the difference and here you can find the right %%%%%%

While testing the influence of the multiplier on memory performance the FSB shlould remain the same - this makes the theoretical maximal memory bandwidth the same for all possible multis.

What does that tell us?

As above - this stock bios is optimizied for multi 9.5x (probably the most, but I can't tell without tests). Here, I bet you would be surprised, are different tests with different multis on different modified bioses for ABIT NF7:

D26 Manta Rays XT by TicTac :

http://forum.hslab.pl/viewtopic.php?...=asc&start=405

(3rd post on the page, Mavers, seems to be some problems with his server, plz be patient)

D26 TaiPan (one of them, there've been many ) by Merlin

http://forum.hslab.pl/viewtopic.php?...=asc&start=390

(4th post, Mavers, as above - takes time)

D26 Discovery by Merlin and D26 Black MantaRay by TicTac

http://forum.hslab.pl/viewtopic.php?...=asc&start=375

(Mavers as usual - the first post and then another one down the page)

Interesting, isn't it?

@ Wolf2000me : have a look at those tests above. You're running the Merlin's bios now, so I see it is well optimized for many multis - 9.5x and 11x are almost identical. Good to know - I'm gonna tell Merlin - he should be happy with this fact.
This is what all the modders have been trying to achieve - similar bandwidth for a wide selection of multis. Usually, the bandwidth dropps while lowering multis, but I see that Epox has prepared something special for us

Btw. Am I writting a compendium or something? I'll try do save those tests above to my HDD an upload them on my server - it takes ages untli they've loaded.

[MietmaN]

Different bioses, different multipliers, different memory performance

by mavers - hslab.pl

D26 Black Mantaray by TicTac

D26 MantaRays XT by TicTac

D26 Discovery by Merlin

D26 TaiPan by Merlin

[Wolf2000me]

Originally Posted by MietmaN

Interesting, isn't it?

@ Wolf2000me : have a look at those tests above. You're running the Merlin's bios now, so I see it is well optimized for many multis - 9.5x and 11x are almost identical. Good to know - I'm gonna tell Merlin - he should be happy with this fact.
This is what all the modders has been trying to achieve - similar bandwidth for a wide selection of multis. Usually, the bandwidth dropps while lowering multis, but I see that Epox has prepared something special for us

Interesting to say the least
I didn't know of all this until now. Thanks MietmaN
If there's anything you want me to test, i'll gladly make some time for you

Originally Posted by mrapples

nah, im in linux and it runs just fine at my stable settings

also, i only have one stick of 512

what if my psu wasnt stable enough, could that cause this?

Yes, it could cause it. It's even the most likely cause. If you're running a lot of hardware, like HDs and new graphics cards you can check it. Try to run with as little hardware plugged in as possible and check if your OC does slightly better. If that's the case then you know the cause. But from my experience i expect your cpu not to be of high quality. Of course if you have a different psu you could test with that as well but it likely won't be a better one or you'd be using that of course

edit: I'm going to look for some winbond UTT or BH-5 memory and buy me a booster. It'll likely be on Twinmos speedpremium memory in the near future. So perhaps no hassle with P/N numbers and stuff like that

[mrapples]

i unplugged 2 cd drives and a couple of fans, the rest (hdd, vid card, etc) had to stay plugged in, and i could barely oc higher, like 30mhz higher

[MietmaN]

@ Wolf2000me :

I'm not sure, if I'm asking too much

If Merlin's gonna make something really special for us, not only fix some minor issues (the cold boot issue and S3 state power saving), some tests are necessary. It would be good, if you could make similar tests like mavers from hslab.pl once did : let's say FSB200 with the most aggressive timings (but stable) and multipliers 8, 8.5, 9, 9.5 and so on with the 12.5x multi at the end. At this FSB your CPU is able to handle all of them and we'll see, how well this bios is optimized and if there's antyhing that could be eventually fixed.

The tests : Sandra 2005 memory tests and Everest Home (1.51 is the latest one) read and write tests (plus delay time). I would do this myself, if I were in Poland now - shouldn't take you more than a few hours, plz.

@ mrapples : 2 x CD-rom drive and some fans makes something like 10-15 Watts - it ain't much. What's your system like - could you list everyting?

[mrapples]

6 fans, 1 sata hdd, dvd/cdrw, cdrom, sound blaster audigy platinum(w/ 5.25in bay controller), fx-5200, 10/100 pci nic, smartfan2 hsf, 1 stick of 512mb ddr

those are all the things sucking power

[MietmaN]

Originally Posted by mrapples

6 fans, 1 sata hdd, dvd/cdrw, cdrom, sound blaster audigy platinum(w/ 5.25in bay controller), fx-5200, 10/100 pci nic, smartfan2 hsf, 1 stick of 512mb ddr

those are all the things sucking power

Not many, not many and your video card isn't one of those power devouring like mine. Now I'm almost sure it's your CPU and no other thing.

[Pandaz3]

new guy here... I have a 8RDA6+Pro with a Mobile 2600 and 2x512 Mushkin PC3200...... because of the memory I decided to try Melin's T2 version of th Bios and could not be happier. I do have a minor glitch that I have seen posted. Cold Boot seem to require that I switch the PSU off, push the front panel power switch a few time to fully discharge the PSU, turn the PSU back on and hold the keyboard 'Insert' key down while powering up.... then all is well, but it won't cold boot normally. It will reboot fine, say you have just installed the latest driver for something and need to re-boot to activate it. It will reboot normally for me.

I am running at 12x200 right now, but I need to get a replacement monitor as its on the blink and I can't "Lend' my A64's monitor to that system all the time!

My first EPoX and now that I have the Monitor woes sorted, I really like it.

[Barix]

I saw that a problem in the Merlin's bios is that S3 and hiberanation are broken ..

Because I really need this functions (pratically I never shutdown windows, I only use suspend to disk and I don't want to abandon this way) I want to ask to Merlin if in the next 'release' he can check if this functions continues working..

By the way thanks for the great work you're doing, Merlin!!