Printable View
Mine works if set to .85 from windows (I run 2.2 @.85 as my lowest setting in crystalcpuid), but it wont post with it set in the bios. I havent tried higher than 1.55 since my chip didnt clock well with the 601 bios.Quote:
Originally Posted by justapost
That's ASUSs' overclocking tool I linked Achim. ;)
Reboot to see if Smart Guardian works.
Manual says Ai Gear helps to reduce noise and power consumtion. I expect it allows switching p-states, it has a three gear shift. :rolleyes:Quote:
Originally Posted by KTE
Currently reinstalling windows cuz it won't work in cnq anymore (other os on different disc works). Will try smart guardian again when i finished the setup.
EDIT: Tried it on a fresh install (cnq works now). It requires to disable manual overclocking settings and does nothing, which is better than freezing the system. Tried it with cnq on and off in the bios everything else had to be auto. No frequencie voltage or power consumption change, not even the fanspeed changes.
Smart guardian does not work even after a reboot, That whould have been a nice way to change the vcore.
@jakalwarrior: Do you get an CPU-Overvoltage Bios warning when you set 0,85V in the bios. Have not tried crystalcpu for anything beside msr editing. Looking at your sig, you must have good cooling to run your 4000+ with such low voltages.:)
For me is the Asus M3A history!:down:
Next week I want to test a brandnew E8200 Wolfdale + MSI P35 Neo2-FR ...
I think a better choice at the moment;)
I will give mine another couple of weeks and then if no voltage options exist it will go, and possibly the Phenom as well.
My old 5000be and mobo were not much slower and used less power.
pfff.. viral marketing trolls again... at least you could go beyond 10 post to tell us that Penryn+P35 is the better choice over Phenom+AMD 7-series chipset... you know it's all about credibility that you lack ;)Quote:
Originally Posted by planet456235
Just picked up the m3a32-mvp deluxe busy testing, so far very similar in performance to the m3a.
Uhm in that price segment i'd prefere the DFi board, but i'm just curious because i never had one. Have you compared the memory bandwith of the two mobos?Quote:
Originally Posted by Thesavage
Guess I stay with 770 and grab the Biostar 770 A2+ if asus can not realize a vcore option in the bios or i'll be patient enough to wait for a 780G mATX mobo with sb700.
I'll build an intel rig with an qx6850 for a co-worker in a week or so. I'm curious how it's resposiveness and power consumption will compare to my phenom rig if I downclock that thing to 2,2GHz.
No memory bandwidth test yet just testing games so far, seems the wireless lan and vista dont like each other. Will keep you posted soon though.BTW this is my second m3a32 deluxe the first one would not start.;)Quote:
Originally Posted by justapost
Did some research how vcore mods are done. So far I located and identified the voltage regulator chip. It's an L6740L and here is the pin-layout.
http://www.abload.de/img/l6740lgyh.jpg
The specs can be found here.
Great, thanks. So you bought the M3A for the time m3a32 was in rma?Quote:
Originally Posted by Thesavage
I have seen no write ups on the Biostar board, might be safer to wait for the 780's. The only good performance reviews I have seen go to this board:Quote:
Originally Posted by justapost
http://www.newegg.com/Product/Produc...82E16813130136
I plan to get one of those for a machine at work (together with an 6400+ maybe).Quote:
Originally Posted by jpierce555
About vcore modding. According to this article, the next step is to identify the resistances connected to pin #4 (FB) and pin #33 (NB_FB) and shade those.
The chip is located next to the mobo's power plug, close to the cpu. Guess I must remove the heat spread connector to inspect this pins. :shakes:
You mean the cpu cooler? Not sure what you mean by heat spread connector.Quote:
Originally Posted by justapost
Just be careful and try a 2hb pencil first. I have a 2hb and a 6hb, used the 6 on my g-card. I would rather you be the experimenter LOL.
Sorry, yep I mean the adapter for the cpu cooler.Quote:
Originally Posted by jpierce555
I think I identiefied the nb resistance but for the cpu I'd need a camera. Used a normal pencil and shaded the nb resistance. AMD Power Monitor did not show a difference but I think that tool reports the vid's. So I try prime95 with lower nb voltages now to see if it worked.
EDIT: It did not work. Must wait till room mate comes home with a camera.
Well I emailed Asus engineering again this weekend since they never replied last time. This time I asked them flat out if they expected the voltage issue to be fixed.
This is really frustrating.
Did you notice disabling the TLB fix via AOD does not bring the same performance increase as disabling it via the Crystal CPU?
Found this in the specs. I think I shaded Rf,nb and not Rfb,nb.
http://www.abload.de/img/l6740l_29uu.jpg
Hope that pic will help identifying the right resistances once I have a few pictures.
I plan to mail em the bu list for 601 once I checked a few things I found with 501.Quote:
Originally Posted by jpierce555
Makes me look into vmodding, so i don't mind atm. :)Quote:
Originally Posted by jpierce555
I noticed a difference between yellow and red mode but aod's red mode.Quote:
Originally Posted by jpierce555
To apply the fix with crystal cpu one must modify the two msr registers on all cores, otherwise only the first core will use the tlb. That might have an impact on benchmarks.
Hey ya'll I have the thread over on Legit, was wondering if any of you have ran water cooling on the NB? I just spent the week fabricating an external water cooling box and the cpu is water cooled (x2 6400 heat pump), the zalman NB water block is laying up on the PSU waiting and ready (after heavy mods to the now not so universal WB lol).
So far it's been a really good board with the X2, but had to wait on NB water cooling for a solution to the NB water block rocking on that tiny chip. Neoprene rubber hand cut is what I'm planning unless you have a better idea.
Flipped 450 FPS on a volt modded 8800gt under Lightsmark on the board. The only better I've seen is an 8800 gtx max oced under water and it crunched a 555 fps.
Lemme know if you watered the NB (safe/not safe) and if you used a gasket and if you did what material.
Much Appreciated!
BTW nice thread/forum
Hi Methious,Quote:
Originally Posted by Methious
Linked your thread on the first post, glad you joined.
Can't help with nb mounting but I'd be interested in a good solution. I googled for nb waterblock mounting on the M2A-VM. This mobo uses the same nb cooling but did not find anything usefull.
I remeber somewhere in this forum was a recommendation about nb cooling on the m2a-vm, just have to find it.
I have two reliable sources, one an engineer that designs water systems for a living, and Bio-Hazard, if you Google "Monster Rig" or Bio-Hazard you'll get about 2 pages on him. One says 3M double sided tape with a window cut into the tape for the NB, the other says neoprene foam cut large enough not to cover the NB chip, but small enough the WB sets on it. (Watch upward pressure from the neoprene on the water block, so the fit has to be close)
I Googled the heck out of M3A NB Water Cooling, no luck. I finally came up with a Zalman universal I got on sale, it's not universal now, you have to Dremel off the the loops used for through the board installs, then where they used to be grind a little to fit it between the loops. Then due to poor capacitor placing cut the end of the hold down clip off right in front of the hole used for the loop hold down hook. Shame I was looking forward to using that water block on later builds. If you go that route be real careful they pull out of the board pretty easy.
If you want greater stability over clocking you can try video ram heat sinks thermal taped to the MOSFETS. Helped me break the 360 FSB wall with a 80mm fan blowing on the mosfets from the blow hole. Don't know if it's going to help Phenom users.
Thanks I joined because you all are doing some great testing and I had been cruising the forum looking for tips. I don't want to mislead any one about my water skills, I'm new to it, first water build after 25 years of messing with this stuff. But I found out that I had access to two people who had massive experience so my info comes from there.
I got the 6400+, don't know if I'd go that direction again. It's the hottest chip I've ever run, even under water, Swiftech MCP350 pump, Dell XPS 710 Rad, Swiftech Water Block, runs 30C idle, 40-42C load. Under a ThermalRight 120 Ultra, load ran 46C with upward creep, nMedia PC IceTank ran 55C load with a lot of creep. Runs like a dream though if you don't mind lack of OC overhead (probably not going there at work :cool: ) Normal non-100% working load it runs pretty good.
Quote:
Originally Posted by justapost
Dont worry about temps that much. Whatever it gets with stock cooler at stock volts must be acceptable or AMD wouldnt sell it like that. They would lose too much money on RMAs (do they sell those retail with a warranty? AFAIK oem has 0 warranty now)Quote:
Originally Posted by Methious
The 6400+ Black edition doesn't come with a cooler, I got the retail version and ran it with the nMedia PC Icetank at first, that's a 92mm 4 heat pipe and under load wasn't sufficient, ThermalTake Volcano wouldn't touch it, ThermalRight 120 Ultra did a good job if you want a cooler the size of a child's shoe box in there that weighs upwards of 2 LBS. Zalman 9700 does a good job.
With coolers of that magnitude the only ones suitable for full load it becomes an issue. It's a 95nm design, more nm means more electrical resistance, more resistance means more heat. Take a look at the AMD site and recommended coolers for the 6400, then do a search for them, after an hour I gave up and got the ThermalRight.
Small correction Methious,6400+ is made on 90nm(not 95) :).
Methious, are those temps stock or oc results? Zalman 9700 was what i had in mind. Don't know if I oc that machine, if there is an more energy efficient point above stock i'll oc it to that point but not further. Looking at 6400+'s stock voltage requirements 1,35V-14V there is not much room left.
Currently I use the thermaltake big typhoon on the m3a. Does a good job in component cooling and the temps are about 3° above the reserator 1p with the reservoir outside in ~1-5° ambient.
Had an Zalman 7000 mounted before, with whom I had ~15-20° higher temps that with the reserator setup in the above 1,5V cpuv area.
You shoul put a picture for your nb waterblock on your legit thread, modding sounds adventurous.
Roommate can back with camera I will take pictures of the L6740L area now.
http://www.abload.de/img/front2i4l.jpg
http://www.abload.de/img/frontud9.jpg
http://www.abload.de/img/back3eoq.jpg
Hmm, not the best quality but better than my eyes.
@informal: BTW, tried the be2400 with the zalman 7000, max I could reach with load temps below 70° was 3,1GHz at 1,4V, close to 0,1V more cpuv required than with watercooling. So the 3600+ I have should do 2.8-2.9GHz max on air.
hi all id like to know is , is the asus m3a a good overcloker with the phenom 9500 ??? i dont realy understand all the high spec terms u guys are using , so a simple yes or no wud be great:)
There is alway a third alternative. Short answer, we don't know cuz the bios limits the voltage you can apply to the cpu to 1,25V. With that voltage you can oc an 9500 to ~2350MHz.
aaaaa but ive heard that the asus and the gigabyte has the best stability and overclockableity than any of the other boards. i was hoping to get 2500mhz stable , is there any chance???
You need more cpu voltage for that speed, so it does not work with the current bios. Once this bios issue is fixed, I assume it will be a good overclocker for phenom, got very good results for the ref HT/fsb speed ~270MHz.Quote:
Originally Posted by sargentbash
My bad on the 95nm got a lap dog interfering with me :rofl: , it is 90nm. Those temps are @ 3.52 full load Orthos X 1 hour. I just got the water block on, using the above described method. Neoprene hand cut gasket about 1/8" and it's rock solid (well rubber rock).
I'll have to get some time to run low multiplier high FSB tests to see if it's doing me any good. I ran it full load about a half hour and the NB was at 38C @ 220.
It was up around 53C if I'm reading the temps right. Been using AOD and Core Temp, and Speed Fan. What are you using to read it, and which one are you reading from so were on the same page.
That's probably not indicative of what you'll get temp wise, my Radiator isn't real great. Aluminum 1 1/2 thick, was a throw in to a trade so I'm using it until I can get a triple 120mm rad.
I took some installed pics, they should be up in about a half hour.
Congrats for the succesfull mounting. You'll need more that 1,45V for that clock I guess.Quote:
Originally Posted by Methious
Hmm, my be2400 reached 392MHz ref HT but only with an 2x HT multi, nb temps had never been an issue but I have the mobo on the table and had one 12" fan for mobo cooling.Quote:
Originally Posted by Methious
With watercooling and without the 12" fan the nb temps reached 50° but I worked a full day with that temps without a problem.
EDIT: I use speedfan for temp reading, Temp1 is cpu, Temp2 is the nb and Temp3 is sb temp.
Great, thanks.Quote:
Originally Posted by Methious
Right now that MSI is the best Phenom oc board unfortunately. Justapost is 100% correct on the need for voltage. I have a 9600 Phenom and can hit 2.46 100% stable.Quote:
Originally Posted by sargentbash
Those pics are up over on Legit, here's the one you were interested in.
http://i243.photobucket.com/albums/f...lSizeMount.jpg
It originally had 2 legs that have to come off at the bottom of the block, then notch maybe 1/32" right where the legs were back into the block, cut one side of the retainer clip right behind the hole for mounting to the wire loop. Keeps it from riding on top of the capacitor. I don't think heat is going to be an issue for most on the NB, I got mine the day they released them, early on there was speculation it was NB heat holding Phenom back, your right it's not it's voltage. Figured since it was a free WB I might as well use it.
Temp on Speed Fan temp2 is 38c idle, 48c full load.
See ya in a while I'm a week behind on Company of Hero's :shocked:
Leaving the first (closer to the cpu) DIMMA1 empty and filling the other 3 with 3x1024Gb "turns on" unganged mode, if cpuz and everest are correct.
The bad news is that AOD locks on the start up,SANDRA locks during Memory Bandwith test and Everest locks during Cache/Memory test.
During POST
M3A reports "single channel".:(
Using Phenom BE with the M3a.
finally my acct got activated.
Anyways, if you hit F4 in bios it shows lots more settings, including unganged mode, more memory timings, and a few other things. Voltage is still stuck at 1.25 though. :(
Max OC I can get out of my Phenom is 2.7 single core, 2.6 dual, and 2.3 quad. btw I'm using the 601 bios.
http://crfx.homestead.com/files/unganged2.JPG
Thanks for the F4 tip, I'll have to run my fingers through it :p:
Great find, thanks for sharing. I saw it has bank and channel interleaving fields now. Was always curious what channel interleaving will bring compared to bank interleaving.Quote:
Originally Posted by CRFX
You rock:up:
Looks more like a dfi bios now!
Thanks a lot!
got to go to work now:(
Wow, how did you find that? Like the others said, thanks for sharing!
:welcome:
Random button mashing, lol
Started wondering if there were any other functions available, so I hit everything.
3 cheers for random key mashing :up:
Any one try vga coolers heat taped on the mosfets for increased OC stability? I know a world class over clocker that swears by it.
I was thinking you might have done something like that. Well that was the response I was hoping for :) that way I don't feel stupid.
My mosfets are still running cool to the touch. If they get hot I will go buy some thermal adhesive and start attaching some of the ramsinks I have laying around.
Are yours running hot now Methious? I have good air flow there, that might make a difference.
My mosfets are running burn my finger off hot. Have some small ram sinks on them with good airflow and still way too hot. I was wondering if this board can even handle the power needed for a good over clock with a Phenom. Could that be why the vcore is still limited to 1.25?
Ran everest bandwidth test with ganged/unganged bankinterleave on/off and all types of channel interleave (including reserved ones).
As for everest and cinebench i get the best results ganged bankinterleave on(Auto) and using channel interleavemode 5 (default).
Channel interleaving has near no impact in ganged mode. In unganged mode
channel interleave works pretty equal in all modes beside mode 3 and 0 whom perform worse.
Tested it with 2x1G-1066 and 4x1G-800.
Channel Interleave Mode
0 Disabled
1 Address bits 6
2 Reserved
3 Address bits 12
4 Reserved
5 XOR of Address bits [20:16,6]
6 Reserved
7 XOR of Address bits [20:16,12]
Is someone aware of a benchmark, whom profits from unganged mode?
Have you guys with the hot mosfets tried to detect which settings affect the temps most?
Try the rightmark :
http://cpu.rightmark.org/download/rmma372bin.exe
Use the Multi threaded test. I also noticed that the TLB ("fix") can be disabled in BIOS now (with the f4) before I had to go work.
Can't wait to get back home.
Other staff (including source code)
http://cpu.rightmark.org/download.shtml
Thank you SocketMan, will give it a try tomorrow. Are you sure about the tlb fix thought i tried F4 on all pages but have not noticed it, can be myfault.Quote:
Originally Posted by SocketMan
I don't have it infront of me , but I think it's the part about the "Microcode" enable/disable, possibly under CPU configuration.
Ah, ok I saw that field and I'm also away from the computer atm. thxQuote:
Originally Posted by SocketMan
BTW: Searched M3A manual for "F4", it's definately nowhere mentioned there.
Just got home turned off that " Microcode update" as well as the secure virtual machine,but unfortunately the AOD still came up with the green/yellow circle and everest gave memory latency of 156, so I was wrong about the Microcode=
TLB fix. My bad ,:hitself: for some reason I thought it was related. Anyway, switching the AOD circle to red gave the Latency of 65 which is much better.
Also the "unganged" mode let's you configure the 2 IMC separately -2CL's,2trp's
so on. Guess we can use mismatched ram sticks in "unganged" mode now.
Edit*
It seems like turning the "Microcode updating" setting to off -disables the AOD's and/or other software ability to turn the "TLB fix off" judging by the high latencies in Everest,even thou the circle goes yellow/red in AOD.
Sandra's (x11sp1 x64 Edition) mem bandwidth test got a nice bump in "unganged mode" slightly over 10k.
It's interesting how there is noway of checking the NB temperature other then thermometer/fingers, mine got up to 80c once (measured with thermometer @the bottom of the actual heatsink) had to put a fan on top
of the video card to blow some air into it.
My mosfet's are hot to the touch @ 220 FSB, 2 120mm front and back in push pull, one 80mm blow hole blowing directly on them. It should be more of an issue at higher FSB speeds, but with phenom pushing 4 cores if they get to hot their going to cause stability problems. For that matter any cpu will tend to destabilize with the mosfet's running to hot.
Not to say that's the only problem, I'm positive lack of vcore is a big part. But it's some thing to look at.
I've been polling the high end over clockers I know and they all agree that on a value based board (especially under water) one of the top suspects in instability is excess mosfet heat.
So the microcode update is required to apply the tlb-fix. :up:Quote:
Originally Posted by SocketMan
Tried Rightmark and I can seen an improvement using unganged now. The improvement is much higher with 1066 that with 800.
Played with memory timings to push the ref HT limit beyond 276MHz (seems to be similar if not equal in unganged mode, tried 272MHz with no problem). So far no luck to get higher with more relaxed timings.
Running with such high ref HT speeds results in no warm restart sometimes, the system works flawless after a cold start.
Your friends are right. If you want to appliy more than ~1,45V 24/7 to the cpu you should buy an high end mobo. But below that voltage temps don't become an issue.Quote:
Originally Posted by Methious
As I have to pay my elictricity bill, I'd buy an better cpu if i need higher clocks.
noob questions here :confused:
this will be my first time doin raid, and im trying to slipstream the driver into my windows install disc. im not sure if im supposed to use the amdchipset driver or the raidcontroller. :shrug: im using this old maximum pc tutorial and it says raid controller, but in the m3a manual when it tells you how to create the floppy for raid it says to put the amdchipset onto the floppy.
well when following that tutorial it says to place the .sys file from the driver in the windows i386 folder or w/e it is. the raid controller driver has a .sys file, but the amd chipset file does not.
im gonna keep reading about it today, but any help would be appreciated. i at least need to know what driver needs to be slipstreamed into the install so i dont need to use a floppy. if someone could help me with that i sould be able to figure out the rest. thanks
Looks like a typo in the manual, the screeshot says "AMD Raid Driver Disk", so use the raid controller driver.Quote:
Originally Posted by crazydiamond
cool, i am pickin up the last of the parts today and doin the build tonight. ill post back and let you know how it goes.
Asus has never responded to me!
I am going to ditch this board. I will put it and the Phenom on Ebay this weekend. I need the money from one of my boards/cpu, and:
It is good with the x2, but I have to run 3.25@1.525v to equal the same benchmark scores I can get with 3.15@1.45 on the old board. My old board won't run a Phenom and most Phenom boards are not worth the money...at the price I could nearly buy a Blood Iron and q6600.
So I don't have much choice, keeping the M3A seems to be a gamble. My x2 will live longer with the old board and that is that.
For me (and I think it's a personal choice) running my 5000+ BE
on stock volts @ 3.00G was/is better then running it @ 3.2/3.3 with 1.55+ volts.But hey it is xtream systems after all:D
I hear ya I tend to stay close to stock vcore on Amd, higher volts on the X2 series doesn't yield that dramatic an increase in speed for the price you have to pay in heat. I might push a little more volts now that I have the new Swiftech triple 120 radiator installed an leak testing. Only for testing though 24/7 stock volts and what I can get from there.
So you contacted their developers directly and never got a response? May I ask what you old mobo is?Quote:
Originally Posted by jpierce555
Yep it really is a gamble, first I had to wait around five weeks till the board got delivered and now I have to wait till the bios matures.
I wonder if this voltage thing is a general limitation of the 770 chipset. The phenom 9500 itself has a upward locked CPU VID at 1,25V. VCore settings rely on the CPU-VID settings (atleast on the MSI Premium mobo) so an upward lock on cpu-vid should cause an upward lock for cpu-vid also.
But it seems the 1,25V upward cpu vid lock is no issue on the 790FX chipsets, they can apply higher vcores there.
Can understand you decision to sell the phenom I have no use beside exploring specs for him too. Damn I don't like to say but an wolfdale at 3GHz for an similar price would do a better job here, 4GHz at ~1,35V are not bad.
Yep, glad to see this thread did not get closed due to bad oc results. :)Quote:
Originally Posted by SocketMan
Must say I liked the feature on the m2a-vm which allowed two nios profiles to be saved. Id like to have two profiles, one for normal work (may be even underclocked for lower power consumption) and one for heavy workloads.Quote:
Originally Posted by Methious
Here is a comparison of bandwidth in ganged vs. unganged mode.
http://www.abload.de/img/everest-gangedtp4.jpg
http://www.abload.de/img/everest-unganged7is.jpg
Memory timings are identical with KTE's, the only difference is the cpu multi 11 (9500) vs 11,5 (9600BE).
http://www.gamewiki.amd.com/overclocking-howto?f=print
cpuvddc=more cpu vid? it looks like:(
My other motherboard is an ECS (believe it or not) KA3 MVP...and both my cpu and ram run at lower voltage, with higher benchmarks.
Yes I contacted the developers twice with no response. I made a final comment yesterday evening to Asus customer support that this is no way to be and not to expect future business from me. I don't like getting blown off, ANY answer at all would have settled me.
Justapost I agree that it may be a limit of the chipset itself, or the mosfets powering it. I don't see good reviews for the other 770 chipsets either.
Due to the chipset or do you expect no suitable bios here also. :)Quote:
Originally Posted by jpierce555
My phenom failed prime95 at 210x11 last night, so switched back to my old m2a-vm and voila the system requires less power in idle (no CnQ) now.
System:
2x1GB Crucial Ballistix 8500 (800MHz,5-5-5-18,2V),
HD 2400pro 256MB,
80GB IDE HD
M2A-VM, GP-9500(2,2GHz,1,8GHz,1,25V)
Idle: 67,8W DC
Load: 138,2W DC
M3A, GP-9500(2,2GHz,1,8GHz,1,2V/1,025V)
Idle: 93,2W DC
Load: 129,9W DC
M3A, GP-9500(2,2GHz,1,8GHz,1,185V/1,2V)
Idle: 102,5W DC
Load: 136,6W DC
M2A-VM however needs higher vcore for higher frequencies than on the M3A.
M2A-VM, GP-9500(210x11 2,31GHz,1,89GHz, 1,3V/1,3V)
Idle: 72,3W DC
Load: 145,5W DC
M3A, GP-9500(210x11 2,31GHz,1,89GHz, 1,2375V/1,0625V)
Idle: 96,7W DC
Load: 142,5W DC
hey socketman, i see your runnin raid on the m3a too. do you know if there is a way to not have the SATA dvdrw being seen as a raid dvd drive? in my boot priority it shows regular dvd drive and a raid dvd drive, just curious if this is how its gonna be or if i can change that. also, on the 'press ctrl f' screen (after asus splash screen) it shows my 2 hd's in raid by #1 and right under that it shows #4 raid dvd drive and that one is blinking. is there a way to bypass this screen or something?
and finally, my last build was pretty much the same except for the last one was running on 1 hard drive. i was getting boot times about 25 seconds and now on a fresh install with this raid0 its taking like 40-45 seconds. im going to run some benchies on my HD's tonight. im getting same superpi 1m scores as my last rig so performance wise its about the same, its just bootup takes much longer. any ideas?
heres the hardware:
asus m3a
x2 5000+ be
4x1gb ballistix
evga 8800gt
2x320gb 7200.10 16mb raid 0
lite-on dvdrw
xclio 500w greatpower psu
When Asus gave me the email address to the engineering department they also stated they don't support overclocking, so in a word yes I don't expect a suitable bios.Quote:
Originally Posted by justapost
Under "storage" in bios it gives options of:
1.Enable/disable The SATA chip
2.When enabled you can toggle the "function" IDE/RAID/AHCI
3.I tried AHCI first and it gave me some problems,then it turns out it's made by Intel,so I switched to "RAID" right away :D
Then made the floppy of the Asus Support CD and used it during install.
4.My SATA DVD only works when the SATA "function" set to IDE, but then you loose the RAID:(
So as far as I know you can't have raid and SATA DVD drive at the same time.
I ended up using a sata/ide to usb adapter- came in handy.
Then just swapped DVD drives later with my other machine (it had IDE DVD).
Fixed! :up:Quote:
Originally Posted by justapost
Exactly! ;)Quote:
Originally Posted by SocketMan
If you place 000 or 1.550VID on the CPU with very good cooling and it only gets X MHz, you won't get more because it's a chip limit and that's 1.536VCore you're feeding it.
I was afraid you would say that.:(
I'll call of the hit on Asus Bios team;) after all 1.7v for cpu and cpuNB is plenty.
Guess we were hoping that there was some "mysteries" core logic (as AMD calls it) setting that was missing from the m3a bios. I haven't given it that much yet,(1.7v)
trying to "break it in" sort of speak -like a new car or a virgin- lick it before you stick approach.:p:
We could use a voltage control for theCPU HT setting, as Justapost can get it in the 270 range.
Plus 1.55V-ish on cool good air will damage/degrade/kill the chip unless with cold. ;)
I went to buy some thermal adhesive today and found the Gigabyte 790fx on a good sale and thought what the heck. Now I have to sell an old sempron/mobo, the KA3, the M3A, and one of my processors or else!
My raid and sata run perfectly off the "raid" option in the bios no problem on both the m3a and m3a32 deluxe;)Quote:
Originally Posted by SocketMan
Sorry to derail the M3A side of this:
I just ran science mark on the Gigabyte 790fx with the Phenom stock and got the same score I got with the M3A turned up to its max, I consider that a really good start. It has voltage up to 1.38v for the cpu. The bios looks more familiar. It has its own oc software, good thing AOD crashes on it. Interesting is Windows does not realize the difference, I did not have to re-activate. It was a very easy install. The layout fits my case much better, still a little clearance issue with the nb and the Infinity hs. I also see that the SATA cable attachments are in a bad place. A major plus is all of the fan sockets.
I will be back tomorrow to let you know if I can oc this board or not.
Any one try setting the HT width to 16x and the HTmax IN/OUT to 16x...WHILE using a Phenom
If the Bios defaults any of these 4settings to 8x your losing performance/CPU,comunication with the Pci-E link with..These new bioses seem to cap the linkwidths for everything when set to 8x..
If you have the Option to cahnge it in Bios @ stock settings it would be under HT link width and HT width...CHANGE all Link Widths TO 16x in BIOS..
Youll see what im talking about.
Ive noticed the 3Dmarks scores posted ABOVE are like only 5000 if that,, ....This is a sign the HT link to PCi-E is being capped with the Bioses. Especially if you have a Bios that the( HTlink and width Autodefaults) even if you set it!
It would be interesting to see if it can be set on the M3A as i can do this with the TA770 and Stock Bios @ Stock clocks.
Polygon from RebelsHaven was so kind to mod the 601 Bios in a way that the Lan-Boot field can be used to disable the fix (enabled the tlb).
I did not yet test it, cuz i'm running few stability test's with phenom on the m2a-vm atm.Quote:
I modded the 601 BIOS for the M3A board to disable the TBL-Fix when the LAN Boot is enabled in the BIOS...
The BIOS is untested and you must be able to recover from a Bad Flash before trying this BIOS. I don't know if the BIOS is corrupt or not. I have done some other AMI BIOS this way, but mainly I Mod Award BIOS and I'm just not sure of this one.
In case you want to try it I uploaded the bios to rapidshare here.
Any body been hearing about the new offerings from AMD due out 2nd quarter 2008? 45nm process, jump from 2 meg cache to 6 meg L2 cache, higher frequencies, dual and quad core designed for the am2+ platform. They'll be having an AM3 platform later initially supporting ddr2 and ddr3. I talked to a friend from college who works for AMD, if what their telling him is true it's a move back to embracing the enthusiast users and will over clock similar to the Intel chips. can't wait to see some engineering samples handed out and tested.
Thanks for the tip, you are right, I can see that on the Gigabyte board. Without changing to 16x it won't oc more than a couple of bumps in the fsb. I did not see that on the M3A, even on this you must type CNTRL+F1 to find it. Maybe somebody should email Asus and ask about it again.Quote:
Originally Posted by gOtVoltage
Woa... :eek: !!! lucky you...Quote:
Originally Posted by justapost
new bios from the master :worship:
You are too kind, my friend! Without your work on the A64 ROM Patcher, we would be nowhere today :)
Just a warning:
The BIOS is UNTESTED and the BIOS chip is soldered in. If the BIOS is corrupt, RMA may be the only answer!
Edit: I will have the board in a few days and will test the mod BIOS :)
Polygon, tried your modded bios. Flashed it from USB from within the bios without problems.Quote:
Originally Posted by RebelsHaven
System boots without problems. But the mod does not work. I get the same everest results with Lan-Boot en/dis-abled.
Also Crystalcpu still has the tlb-disabled flag set (checked only the firt core).
Hmmm... may be a little confusion here... When you say it's "set", you mean it's a "1" correct? I'm already writing a 1b to the 4th bit of the MSRC001_0015 register. If the BIOS already has written a 1b, of course, there will be no change. Do you want a 0b written to the 4th bit or a 1b? I'm writing a 1b. Very easy to change. Here's a BIOS with it the other way:
http://www.lejabeach.com/ASUS/M3A/601RHCF3.zip
This BIOS is UNTESTED and the BIOS chip is soldered in. If the BIOS is corrupt, RMA may be the only answer! Be careful!
Hi Polygon,Quote:
Originally Posted by RebelsHaven
Yepp the bit must be unset, that reenables the TLB.
Tried the new bios and it works. :) But you remove that bit only on core0. It should be done on all four cores, but that may cause problems if cores are disabled via the bios.
Anyway, thank you for your effort, having an option for the tlb-fix is a comfortable thing. :toast:
Hello!
Interesting. That same fix sets or unsets all 4 cores in the Biostar TA770 board. There are several MSR's that there is only 1 per CPU. Isn't the HWCR one of the MSR's that are 1 per CPU and not in every core?
I ran RightMark multi-threaded mem test and it gave me the same bandwidth on all cores.Quote:
Originally Posted by RebelsHaven
Guess you are right about the HWCR. But it's odd that CrystalCPU still reads 8h for cpu's 1-3 but 0h for cpu0. If it's the same register for all cores it should read the same value, or is only the core0 register used in a per cpu case?
I honestly don't know if it's 1 per CPU or 1 per core. I assumed 1 per CPU because setting it on core0 shows it changed on the other 3 in the Biostar board. I'm not sure how to address the other cores from the BIOS anyhow.
The BKDG for the Phenom is very tough to read and many things are not explained, where in the BKDG for AM2 and A64, it was very clear...
But, I'm glad it seems to be working correctly... Thanks for testing it!
Only if you change it from within the bios or does it also behave like this if you change that register with crystalcpu?Quote:
Originally Posted by RebelsHaven
Dunno the guide for A64/AM2 but the BKDG really is tough, hope a new revision will add abit extra background information.Quote:
Originally Posted by RebelsHaven
Glad I could help.Quote:
Originally Posted by RebelsHaven
BTW: A you aware of an 770 chipset mobo, whom is able to apply more than 1,25V to a k10 cpu?
Writing the Biostar from the BIOS, changes all 4 cores for the TBL-Fix only. Writing to the other patch you discovered(MSRC001_1023), does not write to all 4 cores. Also writing to P-State0(MSRC001_0064) only writes to P-State0 for core0... I can't figure out how to write to all 4 cores!
Writing to MSRC0010015 with CrystalCPUID only changes 1 at a time...
I've had the Biostar TA770-A2+ running with 1.5Vcore for my Phenom....
I'm getting the ASUS M3A and if it needs a Vcore mod, I can do that, no problem. Is that Vcore limited?
I took a look in the linux kernel sources. The function whom is affected by the linux-tlb patch gets called for core0 only. I'll inspect the msr registers in linux now.Quote:
Originally Posted by RebelsHaven
My theory for the M3A is that those fields labled prozessor and nb voltages are the cpu/nb vid's. I can select more than 1,25V but the internal decimal places get mapped to the 1,2V-1,25V range.Quote:
Originally Posted by RebelsHaven
AMD Power Monitor reports exact the values i set in the bios, cpu-z's vcore is pretty close 1,264V max if I select 1,25V in the bios.
You have an BE or an 9500/9600 as those have upward locked vid's?
I have a 9500. I'll have to measure the Vcore at 1.50V to see what it is...
The registers all showed the modified value. The patch adds a note to the kernel log if it changes the msr registers.Quote:
Originally Posted by justapost
Here is the excerpt of the log for cores 1-3, core 0 get's called on an earlier stage.
You can see the patch get's applied for each core in opposite to my first adoption.Code:SMP alternatives: switching to SMP code
Booting processor 1/4 APIC 0x1
Initializing CPU#1
Calibrating delay using timer specific routine.. 4400.40 BogoMIPS (lpj=8800810)
CPU: L1 I Cache: 64K (64 bytes/line), D cache 64K (64 bytes/line)
CPU: L2 Cache: 512K (64 bytes/line)
CPU 1/1 -> Node 0
AMD erratum 298 bios fix disabled
CPU: Physical Processor ID: 0
CPU: Processor Core ID: 1
AMD Phenom(tm) 9500 Quad-Core Processor stepping 02
SMP alternatives: switching to SMP code
Booting processor 2/4 APIC 0x2
Initializing CPU#2
Calibrating delay using timer specific routine.. 4400.39 BogoMIPS (lpj=8800788)
CPU: L1 I Cache: 64K (64 bytes/line), D cache 64K (64 bytes/line)
CPU: L2 Cache: 512K (64 bytes/line)
CPU 2/2 -> Node 0
AMD erratum 298 bios fix disabled
CPU: Physical Processor ID: 0
CPU: Processor Core ID: 2
AMD Phenom(tm) 9500 Quad-Core Processor stepping 02
SMP alternatives: switching to SMP code
Booting processor 3/4 APIC 0x3
Initializing CPU#3
Calibrating delay using timer specific routine.. 4400.45 BogoMIPS (lpj=8800903)
CPU: L1 I Cache: 64K (64 bytes/line), D cache 64K (64 bytes/line)
CPU: L2 Cache: 512K (64 bytes/line)
CPU 3/3 -> Node 0
AMD erratum 298 bios fix disabled
CPU: Physical Processor ID: 0
CPU: Processor Core ID: 3
AMD Phenom(tm) 9500 Quad-Core Processor stepping 02
Brought up 4 CPUs
Aint there no simple assembler commando to select the cpu? Don't know much about bios coding, I assumed you change the jump target in the gui part of the bios to a place where you have stored your modified machine code.
VID decides VCore on AM2+.
Achim: Check MSR C001_0064 and post this back please:
Attachment 71811
Tell me what you set in BIOS for CPU voltages and what CPUZ reads too.
In x86 assembly, there is no directive that I know of to select a different core. I'm afraid it might have to be performed with a call to a seperate module. I really don't know... But I think the other 3 cores may get addressed by APIC directives...Quote:
Originally Posted by justapost
To do the mod I did, I wrote a substitute PCI LAN Boot ROM and replaced the original. The new ROM writes to the MSR when it's executed and that is when the LAN Boot is enabled during POST....
Aight, here we go.Quote:
Originally Posted by KTE
PState-0 Status
http://www.abload.de/img/msr64m38.jpg
COFVID Status
http://www.abload.de/img/msr711j3.jpg
Results in 11010b for the vid equals 1,225V That's what I set in the bios as prozessor voltage.
CPU-Z 1,216V/idle 1,232V/load.
How much higher can vcore be choosen than the vid on your mobo?
Last time I wrote assemble code was more than twenty years ago on an atari st, so i'm not up to date here. I'll swap disc again and try to find out how it is done in the linux kernel.Quote:
Originally Posted by RebelsHaven
That sounds like a pretty save way to mod the bios. :) Is this A64 Rom Patcher public available? Googled for it but the only related result was your post here.Quote:
Originally Posted by RebelsHaven
Those are stock VID/VCore on 9500, yep. VID is locked on 9500/9600. Max voltage on 1.25VID is around 1.488V on AM2+ CPU.Quote:
Originally Posted by justapost
What happens if you change the processor voltage in BIOS, do those bit values change?
See above. Differs for each VID for unlocked version though.Quote:
How much higher can vcore be choosen than the vid on your mobo?
Same here. Took a year to get back into it, but it hasn't changed at all!Quote:
Originally Posted by justapost
Visit:Quote:
Originally Posted by justapost
The Rebels Haven BIOS Workshop
That correlates with my max stable vcore on the m2a-vm. On that mobo the cpu-vid (AMD PM reporting) stays at 1,25V and i can change the vcore. Not stable beyond 1,475V.Quote:
Originally Posted by KTE
Tried it with 1,25V in the bios. 34xxh is 30xxh now. -> 11000b -> 1,25V.Quote:
Originally Posted by KTE
Thanks, did not expect that much info. :rolleyes:Quote:
Originally Posted by RebelsHaven
Tried to find how linux modifies core related msr info. Found a few wrapper functions for reading and writing to msr registers. There are smp versions of this functions available whom have a core number variable. But those versions are just wrappers to the single core versions and the core number variable is not used.
The core selection must be hidden somewhere else.
Yes, there is a lot to it...Quote:
Thanks, did not expect that much info.
Vcore check:
Biostar TA770 A2+
Set Vcore: +.225V above default.
BIOS Reads: 1.47V
DMM Reads: 1.493V
Hey guys, I finally got a response from Asus, they said that I should try the 601 bios. I am assuming that means they think the voltage issue is fixed, one of you should drop them an email again to let them no it did not.
Thank you that sorts out my fear that it's an common chipset issue.Quote:
Originally Posted by RebelsHaven
Made my way thru the linux source and found this function. She's used to call the cpu initalisation function whom applies the tlb-fix related stuff.
arch/x86/kernel/smp_64.c line 329
Seems it's done true IPI (Inter Prozessor Interrupts).Code:/*
* this function sends a 'generic call function' IPI to all other CPU
* of the system defined in the mask.
*/
static int
__smp_call_function_mask(cpumask_t mask,
void (*func)(void *), void *info,
int wait)
{
struct call_data_struct data;
cpumask_t allbutself;
int cpus;
allbutself = cpu_online_map;
cpu_clear(smp_processor_id(), allbutself);
cpus_and(mask, mask, allbutself);
cpus = cpus_weight(mask);
if (!cpus)
return 0;
data.func = func;
data.info = info;
atomic_set(&data.started, 0);
data.wait = wait;
if (wait)
atomic_set(&data.finished, 0);
call_data = &data;
wmb();
/* Send a message to other CPUs */
if (cpus_equal(mask, allbutself))
send_IPI_allbutself(CALL_FUNCTION_VECTOR);
else
send_IPI_mask(mask, CALL_FUNCTION_VECTOR);
/* Wait for response */
while (atomic_read(&data.started) != cpus)
cpu_relax();
if (!wait)
return 0;
while (atomic_read(&data.finished) != cpus)
cpu_relax();
return 0;
}
The involved functions seem to solve that true APIC register modification.
APIC inline functions are defined in include/asm/apic.h.
Hmm I already wrote an email to that asus engineer whom sent me the 501 bios, about that issue. Guess he understood my english as good as I understood his chinese. :DQuote:
Originally Posted by jpierce555
How far did you get with your GB mobo jpierce555? I ordered the sapphire clone of the dfi mobo today btw. ;)
OK, thanks! Inter-Processor Interrupts was one of my suspicions...