Gigabyte P35 motherboards and DPC

WARNING TO OVERCLOCKERS!!!

I found a real problem with this beta BIOS, at least on my system. I had been testing yesterday at stock clock settings, and everything seemed great. Well, this morning I decided to use my standard settings for a 3.0GHz (9x333) overclock with my Q6600. The vcore stays at 1.15v, no matter what I set it to in the BIOS. That is enough for my chip at stock speeds, but not for o/c. I tried with auto voltage and the same thing. I bumped up the voltage to 1.3375, but it stayed the same at 1.15. My VID is 1.3v. I tried with C1E both enabled and disabled, no change. I tried with Loadline Calibration disabled and enabled, no change.

Is anyone else having this problem?

ORIGINAL: Desperate Dan

I have PCI soundcards, and I've never used a firewire interface. Have you evr tried your interface on another system and achieved better results?

No, but these Units are supposedley the best Money can buy in their price range. I doubt the Fireface is faulty. I used to have an RMI 96/8 which was PCI Based and it definately performed better in certain ways, like saving while playing back and rendering while playing back which you cant really do with Firewire.

I would post a new topic asking users of the Fireface what results they are getting on their systems. More people would see it because I think only Gigabyte users are looking at this thread. It does seem like you should be able to get much better performance. It would be interesting to see if it is the motherboard or something else in the chain.

ORIGINAL: vanceen



Yes, but that's still a big difference: 37 compared with 21. You said you got 22 with the 13B BIOS. Have you had a chance to confirm that?

Here are my results at stock speed, using the Sonar6/7 benchmark:

WDM
64 samples = 41 multibands/ reopen 42
128 samples = 70 multibands/ reopen 70
256 samples = 87 multibands/ reopen 86
512 samples = 96 multibands/ reopen 96

ASIO
64 samples = 20 multibands/ reopen 20
128 samples = 47 multibands/ reopen 45
256 samples = 61 multibands/ reopen 61
512 samples = 74 multibands/ reopen 74

UPDATE FOR OVERCLOCKERS:

I flashed back to the previous BIOS just to make certain that my board's vcore settings were still functional, and everything is as it should be, albeit with the DPC latency spikes. I contacted Gigabyte to thank them for working on this, and tell them of the problem with the vCore settings. I feel certain that if they were willing to tackle something as obscure as DPC latency spiking, they will fix the vCore issue in the new BIOS, too. But it may help if anyone else seeing this problem contacts them about it, too.

ORIGINAL: HighestOlive

So my questions becomes this, I know there are several of us who are using the GA-P35-DS3L, has any one with this board contacted gigabyte yet to complain about the DPC spiking issue? From my understanding of these latest posts in this thread gigabyte is now sending out board specific beta version fixes as they are contacted. So unless it is brought to their attention that the GA-P35-DS3L is also having these same issues with there be a fix made for this board.

If someone has to contact that how specific about the problem would we have to be. Would it be as simple as saying the GA-P35-DS3L is expiriecing DPC spiking issues the same as other boards?

Does any one have a copy of their email that the sent to gigabyte we could use as a template to send follow up emails for other boards?

I reported on the same board, rev. 2.0 is mine GA-P35-DS3L

They gave me F8n bios and asked me to test with that.

I updated to F8n and tested, still having spike around 500 and just aboce about every 5 bars, so same thing. I told them about the results and also sent them a link to this thread and I am awaiting a response again.

Nice of Gigabyte to pursue our issues! They are making me feel that they may resolve all of our issues and I am definitely beginning to feel they are possibly a go to company, as I feel about Cakewalk.

ORIGINAL: Lanceindastudio


They gave me F8n bios and asked me to test with that.

Lance,

Did you notice if your new beta BIOS held vCore settings when you changed them?

The new bios on my DS3P changes Vcore settings just fine. However, I did notice that the Auto voltage sets the voltage about 0.1V higher than the old bios did (overclocking my Q6600 to 3.2GHz.)

-Clayton

In the light of all the sporadic issues folks are encountering due to the beta BIOS patches, I'm just going to hold off for a bit, install my new MB (non-gigabyte), and keep my gigabyte around just for testing only.

-mr moon

UPDATE:

I had gone back to the last official BIOS release to make sure my vCore settings worked, and was overclocked to 3.0 with a vcore of 1.26v. I reflashed to the beta BIOS, and it held the 1.26v setting, so I am now overclocked at 3.0 with the new beta BIOS. It still won't respond to vcore settings, and just keeps it at whatever it was when I flashed it.

However, I was able to run the Sonar 6/7 benchmarks at 3.0:

ASIO:
64 samples = 31 multibands/ reopen 31
128 samples = 61 multibands/ reopen 62
256 samples = 80 multibands/ reopen 80
256 samples = 105 multibands/ reopen 103

WDM:
64 samples = 55 multibands/ reopen 55
128 samples = 90 multibands/ reopen 91
256 samples = 112 multibands/ reopen 112
512 samples - 123 multibands/ reopen 123

On my system, the ASIO performance is dramatically improved. But the WDM performance is essentially the same at low latencies, and slightly less at higher latencies than it was with the DPC spiking. It may just be variations in performance on my system on the different test dates, or it could be the changes in the beta BIOS affected the performance of the WDM drivers at the higher latencies. This is with Sonar 6.2.1 and my Delta soundcards and the second to latest drivers. I'd be curious to hear the results others are getting.

ORIGINAL: mr. moon

In the light of all the sporadic issues folks are encountering due to the beta BIOS patches, I'm just going to hold off for a bit, install my new MB (non-gigabyte), and keep my gigabyte around just for testing only.

-mr moon

Mr. Moon,

Have you tried the beta BIOS in your system? It would probably be helpful to Gigabyte if you reported your findings to them, so they can work out the bugs for an official release.

ORIGINAL: Dan DAmico


Mr. Moon,

Have you tried the beta BIOS in your system? It would probably be helpful to Gigabyte if you reported your findings to them, so they can work out the bugs for an official release.

Hi Dan,

I've decided that I'm not going to run *any* beta BIOS on my "production" DAW, as I have too many projects that I cannot loose or have anything "bad" happen to. I will probably install the Gigabyte board in one of my other non-primary-DAW quad rigs to mess around with, but that will not be until I get my new non-Gigabyte MB in my primary DAW, which will probably be later tonight. Sorry to be a party-pooper, but the thought of loosing data on my primary DAW is not something I'm OK with.

I will keep an eye on this thread to see how it all pans out. I appreciate the sacrifice the "risk-takers" around here, you included, are making!

EDIT: I'd like to think that if Gigabyte sees that at least one customer (me) has jumped ship from their MBs due to this issue, that others may do the same, so it is in their best interest to fix this issue ASAP.

-mr moon

I understand, Mr. Moon. I am putting my faith in Gigabyte, since I gave up my opportunity to return the board for a refund. I have an Asus P5K Pro arriving tomorrow, which I may hold onto for a few days to see if Gigabyte fixes the beta BIOS. Luckily, the retailer I'm using, newegg, is understanding of my situation, and has agreed to allow me to return the Asus without a restocking fee as long as it's unopened.

ORIGINAL: Dan DAmico

I understand, Mr. Moon. I am putting my faith in Gigabyte, since I gave up my opportunity to return the board for a refund. I have an Asus P5K Pro arriving tomorrow, which I may hold onto for a few days to see if Gigabyte fixes the beta BIOS. Luckily, the retailer I'm using, newegg, is understanding of my situation, has has agreed to allow me to return the Asus without a restocking fee as long as it's unopened.

Yup, Newegg rocks!!

I have an ASUS P5K and P5K-SE, and they are both solid! If the Gigabyte doesn't work out, rest assured you'll have a great MB anyways!

-mr moon

that is interesting that WDM gets worse overall performance at higher audio latencies, with lower DPC latencies in the beta BIOS. How much worse? I think at higher DPC latencies, there will be less thrashing in the system, so that would give the system more breathing room to handle the larger buffer settings. makes some sense. Still, if is only slight, then I think I would prefer the lower DPC latencies and ability to handle really low audio latency when I want to.

It's only slightly worse or essentially the same if you account for any performance differences in my machine on the two different test days. Or it could be the beta BIOS has had some effect on the WDM drivers, or the way they interact with the hardware. The change for the better in ASIO performance means a lot to me because I am syncing two Delta cards together via their drivers, which sync their clocks when used with ASIO drivers. I can use WDM drivers if I sync the clocks by linking the cards with a SPDIF cable.

Your results on your system may vary quite a bit from mine, so I'll be curious to see the results of others.
Remember that these are only beta BIOS, report your results to Gigabyte so they can come out with a solid official release.

ORIGINAL: mr. moon

EDIT: I'd like to think that if Gigabyte sees that at least one customer (me) has jumped ship from their MBs due to this issue, that others may do the same, so it is in their best interest to fix this issue ASAP.

-mr moon

If I end up having to give up on the Gigabyte board and go with the Asus, that will be the second time in 5 years that I will have been in that circumstance. I had nothing but problems with an earlier Gigabyte board, and I ended up replacing it with an Asus A8V Deluxe that is still working perfectly.

It will also be the last time I ever buy anything from Gigabyte.

Guys I've been lurking on this thread for about a week and finally had a chance to test things for myself. I own the GA-P35-DS3L board and was also getting spikes, but none higher than about 700. After updating to the latest BIOS the spikes only got worse (this is not the beta BIOS mentioned upthread, I just want to be clear on that). So I rolled back the F3 and the spikes are totally gone. So I'm just going to do as others have said and hold off until there's a permanent fix.

Really, I have the same board. Do you know what Bios version you had in the first place?

Thanks.

ORIGINAL: emwhy

So I rolled back the F3 and the spikes are totally gone. So I'm just going to do as others have said and hold off until there's a permanent fix.

emwhy,

If you have an official BIOS that is working for you, then you can just stick with that. But if are experiencing the DPC spiking and you'd like Gigabyte to develop a new BIOS to fix it, you shouldn't shy away from testing the beta BIOS. They are going to need some feedback as to what works and doesn't work in order to get enough information for an official release.

I haven't had any instability, had any crashes, nor lost any data with the beta BIOS.

I have a GA-P35-DS3L that had F6 and 500us spikes.

I just got brave enough to zap it back to F2 and all my spikes are gone!

Thank you all ever so much!

Thanx Dan, I'll try the beta one when I have more time later this week. I have two heavy duty projects on my machine and can't afford to mess things up too much until their finalized on Wednesday.

Bitman, I'm running F3 and the spikes are gone as well.

ORIGINAL: emwhy

Guys I've been lurking on this thread for about a week and finally had a chance to test things for myself. I own the GA-P35-DS3L board and was also getting spikes, but none higher than about 700. After updating to the latest BIOS the spikes only got worse (this is not the beta BIOS mentioned upthread, I just want to be clear on that). So I rolled back the F3 and the spikes are totally gone. So I'm just going to do as others have said and hold off until there's a permanent fix.

+1

Using GA-P35-DS3L Rev2.0 with BIOS f3 here with no issues:

I am on the beta driver for my P35-DS4 Rev 2.1 (13b).
I do not overclock.
I adjust RAM multiplier and RAM voltage.
I've only loaded 16 track audio prjects with 5-6 instances of effects and no soft synths.
Although I consider this to have have not pushed the system yet, all is fine so far.

Ed,

Are the DPC latency spikes gone on your system with the beta BIOS?

Dan,

Yes they are.

Not sure if you guys are aware of this or not, but there is a gigabyte forum where a lot of gigabyte specific issues are discussed and perhaps even some Gigabyte support folks hang out on there from time to time. perhaps some of you would post your experiences with the old and beta Bios and see if this helps to pick up some momentum on the issue for Gigabyte to make the fix part of the next official release.

http://forums.tweaktown.com/f69/

Hey Dan, sorry I didn't state this earlier, but it's been a looooong week. Anyway, you and I aren't using the same board, so the Beta BIOS you're using wont do me any good, however I am going to email their tech support people and see if they can come out with a fix for the board I have. As it stands now with the F3 BIOS every thing is working great, but I would be limited in the future if I wanted to upgrade to 1,000 GB HD or a newer Intel CPU.


PS Guys thanx to all for their contributions to this topic, and the other threads as well. My system was very stable, but now thanx to everyone here I feel even better about things.

ORIGINAL: dewdman42

Not sure if you guys are aware of this or not, but there is a gigabyte forum where a lot of gigabyte specific issues are discussed and perhaps even some Gigabyte support folks hang out on there from time to time. perhaps some of you would post your experiences with the old and beta Bios and see if this helps to pick up some momentum on the issue for Gigabyte to make the fix part of the next official release.

http://forums.tweaktown.com/f69/

Thanks dewdman. I posted a brief description of the problem and a link to this thread.

ORIGINAL: emwhy

Hey Dan, sorry I didn't state this earlier, but it's been a looooong week. Anyway, you and I aren't using the same board, so the Beta BIOS you're using wont do me any good, however I am going to email their tech support people and see if they can come out with a fix for the board I have. As it stands now with the F3 BIOS every thing is working great, but I would be limited in the future if I wanted to upgrade to 1,000 GB HD or a newer Intel CPU.


PS Guys thanx to all for their contributions to this topic, and the other threads as well. My system was very stable, but now thanx to everyone here I feel even better about things.

Thanks emwhy!

I guess maybe I will give F3 a try tonight and see if it helps any.
I will report back.

ORIGINAL: emwhy

Guys I've been lurking on this thread for about a week and finally had a chance to test things for myself. I own the GA-P35-DS3L board and was also getting spikes, but none higher than about 700. After updating to the latest BIOS the spikes only got worse (this is not the beta BIOS mentioned upthread, I just want to be clear on that). So I rolled back the F3 and the spikes are totally gone. So I'm just going to do as others have said and hold off until there's a permanent fix.

And there is this gigabyte forum site too: http://groups.google.com/group/alt.comp.periphs.mainboard.gigabyte/topics

I got the same DPC problems with my GA-P35-S3G.
I've contacted the GIGABYTE support and after some days i got that scandalous reply :

Dear Sir,

After confirmed by our RD, DPC high latency was caused by ICH6390 Label Data update. ICH6390 Label data is used for initail seting while system boot up . This is INTEL specification Data and also certified by INTEL. We will not sure if caused any unpredicable issue if changed it. Thus, we are sorry we can't rework this BIOS. We are really sorry for the inconvenience you have with our product.

If you still have any further question to this issue or suggestion about our products/service, please do not hesitate to contact with us. If you have any issue different from original, make sure you issue a new form in order to get proper assistance and it may be response faster as well. We will try our best to help you resolve the problem ASAP.

Best Regards,
GIGABYTE TECHNOLOGY


That people are really crazy,they sold crap boards...i call that a fraud!!!!!!
It's time to start an ANTI-GIGABYTE campaign all over the net,personally i've already blacklisted Gigabyte when i build DAW's for my clients and when i do counsels.

That is a pretty shocking reply from Gigabyte, but it goes against what the rest of us are getting from them. The beta BIOS they have asked us to test has reduced the DPC latency spikes for most testers, so they must have been able to change something to reduce it. Also, other manufacturers using the same chipset don't seem to have the problem.

Before you give up completely, try contacting them here: Gigabyte Tech Support

That's where I was able to get the beta BIOS for testing.

If Gigabyte is saying the problem is caused by Intel spec of the southbridge, then maybe we should be contactng Intel to see if there is anything they can do to correct the problem.

ORIGINAL: Dan DAmico

If Gigabyte is saying the problem is caused by Intel spec of the southbridge, then maybe we should be contactng Intel to see if there is anything they can do to correct the problem.

However, IMHO, if Gigabyte boards are the only ones that are experiencing this behavior, that would probably be pointless as Intel would just say that Gigabyte's implementation of the spec is what's causing the spikes, rather than the spec itself.

If other boards start having his issue, *then* Intel might be who to contact about this issue.

Just my 2 centavos worth...

As an aside, I rolled out my ASUS P5K-SE MB into my primary DAW last night, and it's working flawlessly with the latest BIOS update (version 1008?). I may even just stick with this board depending on how my "new" MB works out (which is being RMA'd back to Newegg with a bad temp sensor).

-mr moon

Agree with that Mr. Moon. That is a simple use of variables there.

But, on another note, my board is awesome for overclocking, but since I changed my bios, I CANNOT hold an overclock. It doesnt register on bootup or in cpu-z. I turned off HEIST and I have triend the previous bios as well as others, the one they sent and othr betas, I think I tried 4 or 5 bios versions.

This sux! Why would my board not OC now? I mean, it doesnt hold the settings I make in the bios.

Thanx, Lance

Does it default to stock settings?

I've noticed the beta BIOS stays at whatever vcore I set it to when I flashed it. It doesn't respond to changes in vcore settings, but everything else seems OK. When I flashed it back to the last official BIOS, it was fine and I could change the vcore.

I am corresponding with tech support, but having a hard time communicating with them. They keep implying that overclocking isn't guaranteed to work, but changing settings in the BIOS is such a basic thing I can't imagine they intend to not deal with it.

Yeah they will be difficult about overclocking stuff.

Any thing I set, it goes psots at stock setting on the cpu. The ram shows 534mhz in cpu-z but does keep 4-4-4-12 setting and the cpu ghz goes back to stock 1.86.

The v-core setting also shows correct, upped to 2.2 as per the ram recommendation.

2.2v should be your DIMM voltage, the vcore is your CPU voltage. Are you able to adjust that?

I can adjust my cpu voltage, but 1.35 always worked I shouldnt need to, but I did try.

Also, even though I put my ram timing at 2.0, it still shows 534mhz and soesnt go to 858 post like before when booting up.

AlsomI tried using gigabyte tuning software to OC and it would not do it whenI pushed go. It always stays at stock settings for the cpu.

So even if you adjust the clock speed in the BIOS it just stays at the stock speed, even if you flash to an official release BIOS?

ORIGINAL: sessionthree

GA-EP35-DS3P here... just tried the F4c bios and the DPC latency values certainly came down (now running ~10us or lower on average.) I wasn't having performance issues at all before, so I don't expect to see any improvement in performance.

Just a note... with the F4c bios, my system will not boot if the onboard LAN is disabled in the bios. You have to make sure it is enabled. I'm not sure if this will be true on other people's systems with DS3P and the F4c bios, but just thought I'd share that in case someone could not get the system to boot. (If you disable the LAN and it won't boot, you have to clear the bios by shorting the clear cmos jumper on the mobo.)

-Clayton

OK. I just got done with about an hour or two of testing the F4c BIOS rev. and had the exact same results you reported. Fot the time being, I'll use this as a secondary internot/DAW rig until they release a full version that works 100%. I may then roll it out as my main rig, but we'll just have to see.

-mr moon

Well, I did some research and found some info on the net!

I guess my particular model has some interesting situations happen.

After flashing bios, I couldnt get the overclock to hold. Even though the settings said OC to 3.0 it would not show up in cpu-z and EasyTunePro5. Also, I could not OC in easytunepro5.

Well, people say if you take the battery out of the board for about 30 minutes, it will completely drain and reset the board. I did it and it worked. My OC settings hold perfect again!

Not only that, but I decided to push it a little further since I have been at 3.0 ghz for 2 or 3 months.

I got it to 3.2ghz and the cpu core temps are 39-42, which is completely safe! This is on a stock sink/fan.

I will run prime tonight and see how it goes. Looking good though! Still god the dpc latency spikes up to around 5000 every 4-7 bars though, but mine is not nearly as bad as other boards. Really it is always in the green, so I guess I cant complain.

Cheers, Lance

I had the system running prime 95 across all 4 cores since last night, and all is well.

I also wanted to report that my system still overclocks without any problems, however, I'm only making a FSB adjustment; no voltage changes or anything else.

I ran the DPC checker while I had an audio project running, and it ran well in the green the entire time.

-mr moon

ORIGINAL: mr. moon
I also wanted to report that my system still overclocks without any problems, however, I'm only making a FSB adjustment; no voltage changes or anything else.

Does that work for you any differently and/or better than changing the CPU clock multiplier? (Just curious)

ORIGINAL: losguy

ORIGINAL: mr. moon
I also wanted to report that my system still overclocks without any problems, however, I'm only making a FSB adjustment; no voltage changes or anything else.

Does that work for you any differently and/or better than changing the CPU clock multiplier? (Just curious)

I just jack the FSB up to 333 MHz, which pushes the proc up to 3 GHz and sets the memory at ~800MHz, which is what it's rated for. A number of folks at various forums and other sites report that this is a very easy, effective, and stable way to OC the q6600. I would have to agree that it is *very* easy, *very* stable, and *very* effective on all my q6600 rigs (I have three of them).

I'm sure that there are better ways to OC, but I'm not trying to eek out every last drop of performance out of these rigs, just trying to get them up to 3GHz for a bit of a bump.



-mr moon

Is there a different way to overclock if your multiplier is locked? If the max multiplier on a Q6600 is 9 then you would have to raise the clock speed to overclock. Is there another way I don't know about? If so, please enlighten me, I'd like to try it.

Dan,

Did you try what Lance did?

Well, people say if you take the battery out of the board for about 30 minutes, it will completely drain and reset the board. I did it and it worked. My OC settings hold perfect again!

I have not done this as I still do not overclock and therefore do not have the need for overclocking BIOS changes to "stick".

ORIGINAL: Dan DAmico

Is there a different way to overclock if your multiplier is locked? If the max multiplier on a Q6600 is 9 then you would have to raise the clock speed to overclock. Is there another way I don't know about? If so, please enlighten me, I'd like to try it.

Just set your FSB to 333MHz?

I don't touch any multipliers at all, I just raise the FSB to 333MHz and that sets the CPU to 3.01 GHz and the memory to ~800MHz.

-mr moon

ORIGINAL: mr. moon
I just jack the FSB up to 333 MHz, which pushes the proc up to 3 GHz and sets the memory at ~800MHz, which is what it's rated for. A number of folks at various forums and other sites report that this is a very easy, effective, and stable way to OC the q6600. I would have to agree that it is *very* easy, *very* stable, and *very* effective on all my q6600 rigs (I have three of them).

I'm sure that there are better ways to OC, but I'm not trying to eek out every last drop of performance out of these rigs, just trying to get them up to 3GHz for a bit of a bump.

OK, thanks, like I said, I was just curious. Some BIOSes let you adjust the CPU and memory clocks/multipliers separately, but yes, you can certainly also bump the FSB if it gets you the right clocks to make all the components happy in the end.

just jack the FSB up to 333 MHz, which pushes the proc up to 3 GHz and sets the memory at ~800MHz, which is what it's rated for. A number of folks at various forums and other sites report that this is a very easy, effective, and stable way to OC the q6600. I would have to agree that it is *very* easy, *very* stable, and *very* effective on all my q6600 rigs (I have three of them).

I'm sure that there are better ways to OC, but I'm not trying to eek out every last drop of performance out of these rigs, just trying to get them up to 3GHz for a bit of a bump.

Set your Multiplier to 8 and the frequency to 400mhz and a RAM Multiplier of 2 will give you 8x400=3.2Ghz and RAM 2 x 400= 800.
Increase CPU Voltage to 1.375. Got mine to 3.2 safely on almost stock cooling.

Dear Sir,

After confirmed by our RD, DPC high latency was caused by ICH6390 Label Data update. ICH6390 Label data is used for initail seting while system boot up . This is INTEL specification Data and also certified by INTEL. We will not sure if caused any unpredicable issue if changed it. Thus, we are sorry we can't rework this BIOS. We are really sorry for the inconvenience you have with our product.

If you still have any further question to this issue or suggestion about our products/service, please do not hesitate to contact with us. If you have any issue different from original, make sure you issue a new form in order to get proper assistance and it may be response faster as well. We will try our best to help you resolve the problem ASAP.

Best Regards,
GIGABYTE TECHNOLOGY


That people are really crazy,they sold crap boards...i call that a fraud!!!!!!

I think you got the wrong tech guy on a bad hair day. My response although not tried yet is at least Promising.

I reported my DPC Latency on the Gigabyte Link you guys suggested and I went to check there now and got the following reply from Gigabyte

Dear Sir,

Thank you for your kindly mail and inquiry. About the issue you mentioned, attached is the latest version BIOS for the motherboard. You should be able to update it and try again later. After bios updating, please do not forget to load BIOS fail-safe default setting and please also do not try to overclock your system at this moment.

For more information about BIOS upgrading, please kindly refer to motherboard user manual or you may visit our website: http://www.gigabyte.com.tw/Support/Motherboard/HowToReflashBIOS.aspx

Unfortunately, if the problem still remains, we suggest you contact your retailer/supplier, or, contact with local distributor for checking and repairing. For their information, please kindly visit our website by following link: http://www.gigabyte.com.tw/WhereToBuy/Default.aspx

At last, if you still have any further question or suggestion about our products/service, please do not hesitate to contact with us directly. We will try our best to help you resolve the problem ASAP.

******** Thank you for supporting Gigabyte products ***********
Attachment : ep35-ds3r_f3e.exe ( 787 KB ) :

ORIGINAL: Desperate Dan

Set your Multiplier to 8 and the frequency to 400mhz and a RAM Multiplier of 2 will give you 8x400=3.2Ghz and RAM 2 x 400= 800.
Increase CPU Voltage to 1.375. Got mine to 3.2 safely on almost stock cooling.

Hmmmm.... I may have to try that at some point. Thanks!



-mr moon

ORIGINAL: Desperate Dan


Set your Multiplier to 8 and the frequency to 400mhz and a RAM Multiplier of 2 will give you 8x400=3.2Ghz and RAM 2 x 400= 800.
Increase CPU Voltage to 1.375. Got mine to 3.2 safely on almost stock cooling.

That's what I've got mine set at right now, with a vcore of 1.3v and it's stable. I just installed a Xigmatek 1283 cooler, which is working fantastically. My temps are about 20c cooler than they were with the stock cooler at 100% load in Prime 95. I used Tuniq TX-2 thermal paste and a Thermalright bolt through kit to mount it. It's a great deal right now through newegg.

In reference to these responses we're getting from Gigabyte, what do you guys think they will do? I'm hoping they will eventually release an official BIOS that fixes the DPC spiking and the minor problems related to voltage settings. I just talked with newegg and they said I could still return my Gigabyte board and keep the Asus, but I have to do it today. Should I trust that Gigabyte will eventually work all this out, or dump it and take my chances with Asus? I have to say that I like the Gigabyte board better than the Asus P5K Pro. It just seems to be made better, and has better heatsinks, and a TI firewire chip.

For those Guys that have put in incorrect BIOS values and cant access their PCs, simply take the Battery out, put it in the wrong way round for 5 Seconds to discharge the capacitor and put it back in the right way up. Trick showed to me by the Local Gigabyte reps works a treat.

My temps are about 20c cooler than they were with the stock cooler at 100% load in Prime 95.

I Have found that TAT (thermal Analysis tool) From Intel stresses the CPU more than Prime 95 does and 100% CPU on prime doesn't give the Same Temperature that TAT generates. Also it's made by Intel so it should be accurate, also you can have the stress levels set to 50% for 20 seconds then 60 etc etc to stop borderline overheating before it get's to CPU shutoff temperature. I mistakenly set my 2.4 CPU to 3.6 and it just ran a little bit hot but I could see it approaching the 63 Degree limit and went back and checked my settings and found incorrect Multiplier values.

Right I am going to switch off now and see if Gigabyte have performed any magic on my DS3R board. Back in a flash Ha Ha

ORIGINAL: Dan DAmico
Should I trust that Gigabyte will eventually work all this out, or dump it and take my chances with Asus? I have to say that I like the Gigabyte board better than the Asus P5K Pro. It just seems to be made better, and has better heatsinks, and a TI firewire chip.

IMHO, even if you like the Gigabyte board better, what good is it to you with a P.O.S. (non-functional) BIOS? If you have the money to spend or your out of time to return it (as I am, missed out by 2 weeks), you could hold onto it. Otherwise, get something that will work for you well enough to make music NOW!!



-mr moon

Regarding overclocking, I experimented with all kinds of different combinations of FSB speed, CPU multiplier, Ram divider, etc.. and benchmarked each setup. I was skeptical at first but it appears that definitely you want to try to keep the ram divider set to 1:1 if you can. This mobo has a lot of flexibility. I can't remember right now which setup I finally ended up with. I have some ram that is capable of overclocking, so I think I pushed the ram speed up to 900mhz by increasing the FSB. Then use the cpu multiplier to get the CPU speed where you want. But try to keep the ram divider at 1:1 if you can. I was skeptical that it would make such a difference, but it really does. All the combinations with ram divider other than 1:1 underperformed other close configurations with the 1:1.

Dan, if I were you in your shoes, I don't know what to tell you. I am holding out for the Gigabyte because of the LPT. If you really think you need the T1 firewire, then it might be worth waiting, but otherwise I would probably switch to the Asus. As far as the build quality, etc... eh.....(shrug)...the asus is fine too.... I am still not 100% confident that Gigabyte will release production BIOS to fix this DPC latency. Hopefully so, but in your shoes... Or.... you can take a chance, send back the asus....and then if it never gets fixed you will have to sell the Gigabyte to get something else. You'll lose $50 that way, but might be worth it to take the risk on the Gigabyte board if you are so inclined. Its too late for me to return my Gigabyte mobo, so basically the only option I have is to sell it. So I am going to try out the Beta Bios and wait to see what happens. Also, I need the LPT and the Asus doesn't have that option.

ORIGINAL: Dan DAmico

In reference to these responses we're getting from Gigabyte, what do you guys think they will do? I'm hoping they will eventually release an official BIOS that fixes the DPC spiking and the minor problems related to voltage settings. I just talked with newegg and they said I could still return my Gigabyte board and keep the Asus, but I have to do it today. Should I trust that Gigabyte will eventually work all this out, or dump it and take my chances with Asus? I have to say that I like the Gigabyte board better than the Asus P5K Pro. It just seems to be made better, and has better heatsinks, and a TI firewire chip.

Since they were pretty responsive getting Beta verisons out to various models I'm going to believe they will have a final version sometime soon. That alone is reason enough for me to not switch to another mobo. However, you overclockers should, IMHO, keep impressing on them the things that don't seem to work (they are beta drivers after all, even they admit it).

My $0.02 (adjusted for inflation and recession: probably not worth much then)

This is the response I got from Gigabyte this morning, after some exchange with them about my results testing the beta BIOS.

"We will pass this over to our team, when the bios is finalize it will be posted on our website"

I have to think they are taking this seriously if they even bothered to put the beta BIOS together. I do think that in the beginning, they were inclined to ingore it, because if you do a web search you find examples from earlier in the year where people were getting a standard response which basically said they were following Intel spec and weren't planning to change anything. But it appears that with enough of us asking for a fix they decided to work on it, so I do believe there is strength in numbers on an issue such as this.