LeonardoWake up and smell the glaciersEagle River, AlaskaIcrontian
edited January 2007
AM2 socket / DDR2 combination which is still very new
Right you are. Sorry I mis-interpreted your previous post.
0
LeonardoWake up and smell the glaciersEagle River, AlaskaIcrontian
edited January 2007
Also, were we off the track completely with reference to VIA chipsets ( I still won't touch them without rubber gloves). Which motherboard, exactly are we talking about? The M2V?
We've all got our biases. One of mine is no chipsets on my motherboards except Intel or nVidia. I do recognize though, that SiS, ALI, and ATI have made great strides. I think nVidia lifted up the whole industry by the competition they brought.
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
edited January 2007
Well, just got the beast back from the shop. Turns out the tech that worked on my system was someone whose done work for me before. He's a very good tech, so I trust his findings. We had a lengthy discussion this am when I picked up the sytem. Here's what his input was.
He, too was puzzled re: why the problem would suddenly appear after more than 6 months.
Judging by the symptoms, he didn't think of the BIOS either until he took a look at Asus' BIOS release history.
Turns out there have been a flood of BIOS releases since this mobo was set loose on the market.
Several of them addressed issues w/, yep!, memory. :rolleyes2 In fact, the last one released was on the 19th of this month (another one addressing memory issues, if I remember correctly)
I'd last updated the BIOS back in September or so (I thought I'd updated it more recently but that must've been my Gigabyte mobo... ) I'm from the old school when it comes to BIOS updates. That is, if it works don't fix it. So, I don't update the BIOS (or until now didn't) unless there's a pressing need.
So, of course, he flashed the BIOS.
I had told them about heat problems when I took it in yesterday, too (esp. in summer) so he swapped out the HSA for a really nice one that has about a gazillion fins and is copper (as opposed to aluminum) where it contacts the processor (better heat conductivity). BTW, A nice anectdote re: excellent service: I told him I wanted the fan off of the HSA that he removed because after a while obsolescence becomes an issue w/ parts like HSAs. He went one better and gave me another branny new HSA assembly. (the same new fangled one) What a guy! :bigggrin:
Moving along, so he ran MEMTEST86 and ORTHOS for hours and didn't get any errors after these changes.
I have it cooking right now w/ ORTHOS BLENDED mode. It's been running (w/ case closed for a change) for nearly 2 hours now w/ no errors. Yippee!!
Now, re: VIA (Leonardo & Goat was it? Ya'll might be interested in this)
He's not a fan of VIA either.
Interestingly, he said it was more because they haven't kept up w/ current technology than the general quality of their chipsets.
He told me of a case where he has to build out hundreds of systems for a client that uses mobos w/ one of VIAs chipsets. In an earlier version of the board, the southbridge doesn't properly support SATAII so he has to jumper SATAII drives to SATA.
I asked him what boards he usually sees the VIA chipsets on and he said boards running AMD processors and cheaper mobos.
Again I say, what happened to Asus' quality??
So that's it in a (big) nutshell. I'm going to burn it in overnight and if all goes well restore my HDD image.
I still want to look into customizing the BIOS but I'll start a new thread for that.
PS://
You all were great. Thankyou for contributing to this thread!!
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
Also, were we off the track completely with reference to VIA chipsets ( I still won't touch them without rubber gloves). Which motherboard, exactly are we talking about? The M2V?
We've all got our biases. One of mine is no chipsets on my motherboards except Intel or nVidia. I do recognize though, that SiS, ALI, and ATI have made great strides. I think nVidia lifted up the whole industry by the competition they brought.
Over the years I've stayed away from SiS chipsets. I hadn't realized there were issues w/ VIA chipsets until ya'll pointed it out.
As for the nVidia chipset, I'm iffy because of the experience I've had w/ my Gigabyte mobo. I'm not fond of their nForce update method either. I'd rather pick and choose my updates. I suppose I need to just get used to it.
0
Straight_ManGeeky, in my own wayNaples, FLIcrontian
Thankyou, RyderOCZ. I'm looking forward to learning about this. I've never futzed w/ the BIOS settings that way b4. Thought that was just for those into overclocking. Looks like regular folks can benefit by tweaking the BIOS, too.
Why do you think it started rebooting out of the blue like that, though? If the tolerances were too tight, why didn't if fail from the beginning do you reckon?
Simple-- for the first 5-6 BIOS flash tries on a new motherboard from mfr, figure they are fixing what goes wrong in the field. It's been tested under conditions that match the BIOS and the engineers know what RAM works best with the board given the BIOS it has and has been tested with it.
Real World:
Each mfr's RAM will have slightly different fine timings. So, when Joe gets his new machine it may or may not have RAM of same MFR as tested. The BIOS may be set too tight due to a quirk in the non-tested RAM. So, mobo mfr has to adapt if RAM brand of a popular RAM mfr is used in machines out there, and what they (Mobo mfr)do is isolate problem by getting some of that RAM and adjust BIOS timing range for the test RAM until it handles the RAM right for both the RAM originally tested when mobo is mfr'd and the RAM that did not work. The loosening of timings let BIOS code work for a wider set of brands and models.
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
edited January 2007
Thankyou Straight_Man. That makes sense. I'm very familiar w/ lab vs real world. However, it still doesn't explain why it took more than 6 months for the problems to surface.... There were NO hardware changes (including RAM) from the time the system was built 'till the time it began failing. If there was a RAM compatibility issue, I would have expected it to show up in the beginning. Therefore, it only stands to reason that something (in the original (which is also the current)) hardware and/or firmware had to have changed or changed states. THAT seems to be the eternal question here - WHAT changed?
Not sure what you mean by that. Do you mean it wasn't burned in? or Do you mean that a burn in caused the problem? I think more than 6 months of 24/7 use might constitute an ample burn in, yes??....
0
Mt_GoatHead Cheezy KnobPflugerville (north of Austin)Icrontian
edited January 2007
Great to hear all is well!
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
Thankyou Mt_Goat. Nearly 5 hrs into Orthos blended w/o a hitch. I'm a happy camper.
Thanx again for you participation. I really appreciate it.
0
nonstop30151° 27' 24.87" N // 0° 11' 38.91" WMember
edited January 2007
Just a final thought Pterocarpous with respect to the compatibility of your DDR2 memory with the M2V motherboard.
If you haven't done so already, you can refer to the Recommended and Qualified Vendors list for DDR2 memory in the motherboard manual to make sure the ones you are using are what ASUS and VIA recommend for your board.
I had a look at your related thread where you were attempting to troubleshoot the problems you encountered and I noticed the specifications of your SpecTek DDR2 533. At first glance it seems that it doesn't match the recommendations in terms of vendor model and component.
The best way to check is to look at your memory sticks and see if the recommended vendor model number or component number is designated on them. You can also refer to the SpecTek website to see whether the model you have qualifies for the M2V
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
edited January 2007
Thankyou nonstop301. And thankyou, too for going to the trouble to review the other thread.
You make a very good point. I thought about that, too when I saw the SpecTek RAM in the system. I'd thought I specified Kingston when I had the system built but can't remember now. I'll look into that. (Whether my current RAM is on the list.) My guess is the SpecTek won't be in the list. In which case, if it starts acting up on me again, I'll pull the RAM and replace it w/ RAM off Asus' approved list.
0
LeonardoWake up and smell the glaciersEagle River, AlaskaIcrontian
edited January 2007
Glad you got it sorted. Sorry, but I can't let this one go:
Again I say, what happened to Asus' quality??
I concede that Asus should have straightened out the BIOS shortcomings within the first couple of months, but with that said, you don't have to buy an Asus motherboard with VIA chipsets. There are so many choices with Asus. When I make my motherboard choices, I choose a chipset first; the board brand and model selections will conform first to the chipset choice. If a manufacturer does make not make a board with the desired chipset, that maker is automatically disqualified. I learned this lesson after the pain (did I say PAIN?) of the KT133/A chipsets that supported (sort of) the K7 Athlons. Once nVidia jumped in the game, I latched on to theirs and Intel's chipsets and haven't looked back. (but I repeat myself)
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
edited January 2007
I expect BIOS updates and revisions. Especially early on. What I was referring to was Asus' choice in chipsets. If VIA is known to produce inferior chipsets, why is Asus putting VIA on their boards? That's what gives me pause re: Asus' overall quality anymore.
I'll definetly follow your example from now on and choose a chipset then a motherboard in that order. Lesson learned! (now I think I'll go lick my wounds some more...)
Thankyou again Leonardo. I've really appreciated your and everyone's feedback. :smiles:
0
LeonardoWake up and smell the glaciersEagle River, AlaskaIcrontian
edited January 2007
If VIA is known to produce inferior chipsets, why is Asus putting VIA on their boards?
I don't know for a fact, but my assumptions why Asus will use them:
1. Asustek is one of the biggest if not the biggest motherboard maker. They can't stay that way if they don't cater to all or nearly all market segments. VIA helps them sell to the entry level market with low cost chipsets. You could have chosen an nVidia nForce 590 chipset (socket AM2) motherboard, but you didn't. They were more expensive, weren't they. Market forces in action. Asus didn't want you make your purchase from Gigabyte, ECS, or Biostar.
2. Not all series of VIA chipsets have been problematic. The latter chipsets for pre-64 Athlons, such as K266 and K333 were decent chipsets. At that time, VIA was the only silicon company turning out motherboard chipsets in sufficient volume for growing AMD demand. ALI and SiS also produced 'AMD' chipsets, but in smaller numbers and with ho-hum performance. AMD itself did not have the fabrication capacity nor the necessary capitol to produce chipsets in volume. They did make reference chipsets for their CPUs, which were mostly very good, but VIA, I think never lived up to AMD's hopes. When nVidia entered the fray, it only took about a year for VIA to be relegated to the entry level, low cost arena. (nVidia, in my opinion, was just as instrumental for AMD's market share increase in the past few years as AMD's technical advances.) Since about 2003, VIA has not been able to compete with nVidia or Intel in the mid-range and high performance markets. Maybe Asus thought VIA's reliability problems were behind them. Maybe the VIA chipset in the motherboard in question is just an anomaly as far as modern VIA reliability is concerned.
....but...maybe I'm just all washed up. After all, the Asus M2N32-SLI Deluxe seems to be notoriously buggy. It's an nForce 590-based board. Honestly, I'm not going to read that huge thread to see from where the problems sprang - BIOS? Memory incompatibilities? Chipset inadequacy? Hmm, maybe Asus does bear watching.
Just for reference, those problems with the M2N32-SLI Deluxe are almost all BIOS and memory compatibility issues. After several BIOS revisions, it seems to be quite a stable and high performance board.
This may speak more to ASUS' testing procedures than to quality of hardware. The hardware is great - it was the software that was the problem, and it clearly wasn't tested with a lot of different RAM. Consider us, the customers, as having been beta testers.
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
...I don't know for a fact, but my assumptions why Asus will use them:...You could have chosen an nVidia nForce 590 chipset (socket AM2) motherboard, but you didn't. They were more expensive, weren't they...
I don't skimp when it comes to building out new systems. I understand the importance of quality and reliability. That's why I chose Asus in the 1st place. (And the M2V wasn't cheap either, btw.) I simply wasn't aware of the chipset issue at the time. I ended up w/ the VIA chipset out of pure ignorance, not by being "frugal". :rolleyes2
...Market forces in action. Asus didn't want you make your purchase from Gigabyte, ECS, or Biostar...
And as I said earlier, I'm not too terribly taken w/ Gigabyte. That's the board I have in my other system and I've had problems w/ it from the beginning. I had the Asus system built after my Gigabyte system. After the experiences I was having on the Gigabyte system (nVidia chipset, btw), I insisted on Asus in my next system.
...2...Maybe Asus thought VIA's reliability problems were behind them. Maybe the VIA chipset in the motherboard in question is just an anomaly as far as modern VIA reliability is concerned...
We'll never know I suppose. Until some decent reviews hit the streets, I know I'm done w/ VIA.
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
Just for reference, those problems with the M2N32-SLI Deluxe are almost all BIOS and memory compatibility issues. After several BIOS revisions, it seems to be quite a stable and high performance board.
This may speak more to ASUS' testing procedures than to quality of hardware. The hardware is great - it was the software that was the problem, and it clearly wasn't tested with a lot of different RAM. Consider us, the customers, as having been beta testers.
Thankyou primesuspect. That's reassuring to hear. I'm hoping it was just firmware issues and that, w/ the revisions, the problem is really solved. (That flashing the BIOS wasn't just a bandaid for a deeper problem - one that caused the firmware to become corrupt, for instance.)
Well, Orthos has been running BLENDED mode for 17' 45" now w/out a hitch. I think I can safely say she's a happy little computer now (at least w/ respect to CPU & RAM). :bigggrin:
0
Straight_ManGeeky, in my own wayNaples, FLIcrontian
Thankyou Straight_Man. That makes sense. I'm very familiar w/ lab vs real world. However, it still doesn't explain why it took more than 6 months for the problems to surface.... There were NO hardware changes (including RAM) from the time the system was built 'till the time it began failing. If there was a RAM compatibility issue, I would have expected it to show up in the beginning. Therefore, it only stands to reason that something (in the original (which is also the current)) hardware and/or firmware had to have changed or changed states. THAT seems to be the eternal question here - WHAT changed?
Ok. take RAM that has been run a tib or more too fast for what it is designed for(as far as timing specs). BIOS gave nearest try it had in it's possibilities table, and that required RAM to run faster than it was designed for given BIOS possibilities. Under such circumstances the RAM will gradually get damaged due to heat buildup over time. Now, let's say that a BIOS release now treats the particular RAM you have in the box correctly as to timings. If RAM is not badly overheated any more(due to more recent BIOS), then there is a chance that the application of a RAM cooler might let the RAM, along with latest BIOS, perform right.
I would run MemTest and if it passes on an overnight run (to get the RAM as hot as it will work for a long test run) then install a cooler for the RAM. Otherwise, if it fails the RAM test run it is defective and should be returned to the system builder as defective.
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
Ok. take RAM that has been run a tib or more too fast for what it is designed for(as far as timing specs). BIOS gave nearest try it had in it's possibilities table, and that required RAM to run faster than it was designed for given BIOS possibilities. Under such circumstances the RAM will gradually get damaged due to heat buildup over time. Now, let's say that a BIOS release now treats the particular RAM you have in the box correctly as to timings. If RAM is not badly overheated any more(due to more recent BIOS), then there is a chance that the application of a RAM cooler might let the RAM, along with latest BIOS, perform right.
I would run MemTest and if it passes on an overnight run (to get the RAM as hot as it will work for a long test run) then install a cooler for the RAM. Otherwise, if it fails the RAM test run it is defective and should be returned to the system builder as defective.
Allright! Now THAT's a plausible theory! I like it. It's logical to my old linear-minded bean. :bigggrin:
It's been running Orthos Blended for 20' 25" now w/ no errors. I'll stop that test and start MEMTEST86 and let it cook for a l-o-n-g while. I'll let you know how it shakes out.
Thanx Straight_Man!!
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
edited January 2007
Ok, after 20+ hours running Orthos here are the voltage and temperature levels as shown in the BIOS:
CPU Temp: 39C/102F
MB Temp: 40C/104F
CPU FAN: 2678 RPM
NONE of these levels are steady but the variation is minimal:
VCORE: Ranges from 1.360 to 1.376 (~ .016V variation)
3.3V: Ranges from 3.344 to 3.360 (~ .016V variation)
5V: Ranges from 4.999 to 5.026 (~ .027V variation)
12V: Ranges from 12.281 to 12.342 (~ .061V variation)
The tech at the shop told me voltage levels are ok if they're w/in .5V
What d'ya'll think?
BTW://
MEMTEST86 is running now - right on the heels of the 20+ hours Orthos Blended Mode test
0
LeonardoWake up and smell the glaciersEagle River, AlaskaIcrontian
edited January 2007
OK, I admit, perhaps I was a bit harsh on VIA. Prime's point about software - BIOS coding, is cogent. Ptero, perhaps your motherboard selection was spot on and Asus did let you down. (I still won't get near anything VIA, from past experience.) Considering that Asus has within the last year released two boards, a VIA-based board and an nForce-based board that have had significant teething problems, Asus has now raised a yellow flag with me. I sure hope this doesn't turn out to be a trend. Abit went in the toilet a couple years ago. Let's hope Asus isn't also going down the path to mediocrity by sacrificing quality for quantity.
Concerning DRAM coolers. As Straight_Man mentioned, overheated RAM definitely can cause problems. Ram coolers though? I think it was Tom's Hardware or another big site a couple years ago did some thorough testing with RAM sinks. It was the conclusion of the testers that RAM sinks are essentially eye candy, making at most around a 1*C temperature lowering effect. (I like the eye candy though, as all my cases are windowed.) It is my opinion that if RAM is running too hot, it's an overvoltage problem or a poorly ventilated case. Ptero - I'm not implying that your computer case is poorly ventilated; I just wanted to relate what I read about RAM sinks.
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
...Asus has now raised a yellow flag with me. I sure hope this doesn't turn out to be a trend. Abit went in the toilet a couple years ago. Let's hope Asus isn't also going down the path to mediocrity by sacrificing quality for quantity...
They've caught my attention, too. I will no longer pick an Asus (or any for that matter) board w/out a thorough investigation into its reliability. I've watched over the years manufacturers' products, where before you could swear by them, plumet w/ respect to their quality...only to crawl back up the ladder again. Or they get sold off from company to company and which direction their quality takes is anyone's guess. Then you have the new kids on the block who throw their lot into the fray. Bottom line, it's damn hard to keep up.
......It is my opinion that if RAM is running too hot, it's an overvoltage problem or a poorly ventilated case. Ptero - I'm not implying that your computer case is poorly ventilated; I just wanted to relate what I read about RAM sinks...
That's good to know. And I agree. It sounds like RAM sinks are addressing a symptom (apparently not very effectively) and not the underlying problem. The case for this system is a mid tower. And there's plenty of empty space (it's not tight in there). I suggested adding fans and my shop balked - saying it would only make the system noisy. Hmmmm, I thought. What about quiet fans? I decided that if the heat problem persists, I'd order some fans from newegg and install them. I do have a new HSA now - much better quality. I posted my temps and voltage levels in my previous post in this thread and no one has commented so I will have to explore elsewhere whether those levels are ok. If they aren't then, I'll get the fans.
BTW, 20+ hours on MEMTEST86 w/out a hitch. That's on the heels of 20+ hours of ORTHOS Blended Mode. I'd say that the system is pretty sound now.
0
nonstop30151° 27' 24.87" N // 0° 11' 38.91" WMember
edited January 2007
Your temperatures and voltages are clearly within the normal range Pterocarpous, that's why they don't require further commenting
I agree with primesuspect's view that you are essentially a beta tester when you go out and buy the first release of a motherboard designed for a brand new socket.
I don't think the choice of VIA or nVIDIA has anything to do with it in this case and you would encounter a few nags here and there with either one especially if you don't receive the latest BIOS fixes. The manufacturers do point out quite forcefully to check for and perform BIOS updates
For me personally, it isn't a matter of deciding whether to go for Asus or Gigabyte or MSI or any other manufacturer. It's more about the details of the chipset and at moment the nVIDIA chipsets match the criteria I'm looking for more closely than VIA or any other chipset producer when it comes the AMD processors. With Intel processors the Intel chipsets are the rule but the nVIDIA SLi technology is very popular as well.
0
PterocarpousRosie the Riveter Lives On in CA, USA!New
...Your temperatures and voltages are clearly within the normal range Pterocarpous, that's why they don't require further commenting ...
Thankyou, nonstop301. It's good to know they are in tolerance.
As for not commenting, if folks don't respond to a question or comment, that equates to "no feedback w/out any implications" in my book. Otherwise I could draw an endless list of conclusions based on there having been no response. The written word is our only means of communication here. We don't have body language or other auditory or visual cues to go by. Just what we put in writing. IOW, The implications of an absents of a response are infinite. Therefore, the best way to convey your thoughts or opinions in this (a forum) context is to set them in writing. :smiles:
...I agree with primesuspect's view that you are essentially a beta tester when you go out and buy the first release of a motherboard designed for a brand new socket...
Unfortunately, I didn't realize at the time the mobo/socket were new. I didn't do my homework and I have reaped what I sowed as a result. I'll know next time. You can count on that!
...I don't think the choice of VIA or nVIDIA has anything to do with it in this case and you would encounter a few nags here and there with either one especially if you don't receive the latest BIOS fixes. The manufacturers do point out quite forcefully to check for and perform BIOS updates ...
I suppose I just need to change my approach a bit. I come from the old school of "leave the BIOS alone unless a change is warranted". Because the symptoms didn't, in my mind, lead me to think of the BIOS, I didn't consider it as the problem. So, flashing the BIOS never entered my mind.
...For me personally, it isn't a matter of deciding whether to go for Asus or Gigabyte or MSI or any other manufacturer. It's more about the details of the chipset and at moment the nVIDIA chipsets match the criteria I'm looking for more closely than VIA or any other chipset producer when it comes the AMD processors. With Intel processors the Intel chipsets are the rule but the nVIDIA SLi technology is very popular as well.
That's actually been a very valuable lesson learned for me w/ this whole evolution. That the chipset should be taken into consideration in higher priority than the motherboard manufacturer. W/ all the sage advice you all have given me, I'll definetly do a better job next time.
Comments
We've all got our biases. One of mine is no chipsets on my motherboards except Intel or nVidia. I do recognize though, that SiS, ALI, and ATI have made great strides. I think nVidia lifted up the whole industry by the competition they brought.
- Now, re: VIA (Leonardo & Goat was it? Ya'll might be interested in this)
- He's not a fan of VIA either.
- Interestingly, he said it was more because they haven't kept up w/ current technology than the general quality of their chipsets.
- He told me of a case where he has to build out hundreds of systems for a client that uses mobos w/ one of VIAs chipsets. In an earlier version of the board, the southbridge doesn't properly support SATAII so he has to jumper SATAII drives to SATA.
- I asked him what boards he usually sees the VIA chipsets on and he said boards running AMD processors and cheaper mobos.
- Again I say, what happened to Asus' quality??
So that's it in a (big) nutshell. I'm going to burn it in overnight and if all goes well restore my HDD image.I still want to look into customizing the BIOS but I'll start a new thread for that.
PS://
You all were great. Thankyou for contributing to this thread!!
Over the years I've stayed away from SiS chipsets. I hadn't realized there were issues w/ VIA chipsets until ya'll pointed it out.
As for the nVidia chipset, I'm iffy because of the experience I've had w/ my Gigabyte mobo. I'm not fond of their nForce update method either. I'd rather pick and choose my updates. I suppose I need to just get used to it.
Simple-- for the first 5-6 BIOS flash tries on a new motherboard from mfr, figure they are fixing what goes wrong in the field. It's been tested under conditions that match the BIOS and the engineers know what RAM works best with the board given the BIOS it has and has been tested with it.
Real World:
Each mfr's RAM will have slightly different fine timings. So, when Joe gets his new machine it may or may not have RAM of same MFR as tested. The BIOS may be set too tight due to a quirk in the non-tested RAM. So, mobo mfr has to adapt if RAM brand of a popular RAM mfr is used in machines out there, and what they (Mobo mfr)do is isolate problem by getting some of that RAM and adjust BIOS timing range for the test RAM until it handles the RAM right for both the RAM originally tested when mobo is mfr'd and the RAM that did not work. The loosening of timings let BIOS code work for a wider set of brands and models.
Thanx again for you participation. I really appreciate it.
If you haven't done so already, you can refer to the Recommended and Qualified Vendors list for DDR2 memory in the motherboard manual to make sure the ones you are using are what ASUS and VIA recommend for your board.
I had a look at your related thread where you were attempting to troubleshoot the problems you encountered and I noticed the specifications of your SpecTek DDR2 533. At first glance it seems that it doesn't match the recommendations in terms of vendor model and component.
The best way to check is to look at your memory sticks and see if the recommended vendor model number or component number is designated on them. You can also refer to the SpecTek website to see whether the model you have qualifies for the M2V
You make a very good point. I thought about that, too when I saw the SpecTek RAM in the system. I'd thought I specified Kingston when I had the system built but can't remember now. I'll look into that. (Whether my current RAM is on the list.) My guess is the SpecTek won't be in the list. In which case, if it starts acting up on me again, I'll pull the RAM and replace it w/ RAM off Asus' approved list.
I'll definetly follow your example from now on and choose a chipset then a motherboard in that order. Lesson learned! (now I think I'll go lick my wounds some more...)
Thankyou again Leonardo. I've really appreciated your and everyone's feedback. :smiles:
1. Asustek is one of the biggest if not the biggest motherboard maker. They can't stay that way if they don't cater to all or nearly all market segments. VIA helps them sell to the entry level market with low cost chipsets. You could have chosen an nVidia nForce 590 chipset (socket AM2) motherboard, but you didn't. They were more expensive, weren't they.
2. Not all series of VIA chipsets have been problematic. The latter chipsets for pre-64 Athlons, such as K266 and K333 were decent chipsets. At that time, VIA was the only silicon company turning out motherboard chipsets in sufficient volume for growing AMD demand. ALI and SiS also produced 'AMD' chipsets, but in smaller numbers and with ho-hum performance. AMD itself did not have the fabrication capacity nor the necessary capitol to produce chipsets in volume. They did make reference chipsets for their CPUs, which were mostly very good, but VIA, I think never lived up to AMD's hopes. When nVidia entered the fray, it only took about a year for VIA to be relegated to the entry level, low cost arena. (nVidia, in my opinion, was just as instrumental for AMD's market share increase in the past few years as AMD's technical advances.) Since about 2003, VIA has not been able to compete with nVidia or Intel in the mid-range and high performance markets. Maybe Asus thought VIA's reliability problems were behind them. Maybe the VIA chipset in the motherboard in question is just an anomaly as far as modern VIA reliability is concerned.
....but...maybe I'm just all washed up. After all, the Asus M2N32-SLI Deluxe seems to be notoriously buggy. It's an nForce 590-based board. Honestly, I'm not going to read that huge thread to see from where the problems sprang - BIOS? Memory incompatibilities? Chipset inadequacy? Hmm, maybe Asus does bear watching.
This may speak more to ASUS' testing procedures than to quality of hardware. The hardware is great - it was the software that was the problem, and it clearly wasn't tested with a lot of different RAM. Consider us, the customers, as having been beta testers.
Well, Orthos has been running BLENDED mode for 17' 45" now w/out a hitch. I think I can safely say she's a happy little computer now (at least w/ respect to CPU & RAM). :bigggrin:
Ok. take RAM that has been run a tib or more too fast for what it is designed for(as far as timing specs). BIOS gave nearest try it had in it's possibilities table, and that required RAM to run faster than it was designed for given BIOS possibilities. Under such circumstances the RAM will gradually get damaged due to heat buildup over time. Now, let's say that a BIOS release now treats the particular RAM you have in the box correctly as to timings. If RAM is not badly overheated any more(due to more recent BIOS), then there is a chance that the application of a RAM cooler might let the RAM, along with latest BIOS, perform right.
I would run MemTest and if it passes on an overnight run (to get the RAM as hot as it will work for a long test run) then install a cooler for the RAM. Otherwise, if it fails the RAM test run it is defective and should be returned to the system builder as defective.
It's been running Orthos Blended for 20' 25" now w/ no errors. I'll stop that test and start MEMTEST86 and let it cook for a l-o-n-g while. I'll let you know how it shakes out.
Thanx Straight_Man!!
CPU Temp: 39C/102F
MB Temp: 40C/104F
CPU FAN: 2678 RPM
NONE of these levels are steady but the variation is minimal:
VCORE: Ranges from 1.360 to 1.376 (~ .016V variation)
3.3V: Ranges from 3.344 to 3.360 (~ .016V variation)
5V: Ranges from 4.999 to 5.026 (~ .027V variation)
12V: Ranges from 12.281 to 12.342 (~ .061V variation)
The tech at the shop told me voltage levels are ok if they're w/in .5V
What d'ya'll think?
BTW://
MEMTEST86 is running now - right on the heels of the 20+ hours Orthos Blended Mode test
Concerning DRAM coolers. As Straight_Man mentioned, overheated RAM definitely can cause problems. Ram coolers though? I think it was Tom's Hardware or another big site a couple years ago did some thorough testing with RAM sinks. It was the conclusion of the testers that RAM sinks are essentially eye candy, making at most around a 1*C temperature lowering effect. (I like the eye candy though, as all my cases are windowed.) It is my opinion that if RAM is running too hot, it's an overvoltage problem or a poorly ventilated case. Ptero - I'm not implying that your computer case is poorly ventilated; I just wanted to relate what I read about RAM sinks.
BTW, 20+ hours on MEMTEST86 w/out a hitch. That's on the heels of 20+ hours of ORTHOS Blended Mode. I'd say that the system is pretty sound now.
I agree with primesuspect's view that you are essentially a beta tester when you go out and buy the first release of a motherboard designed for a brand new socket.
I don't think the choice of VIA or nVIDIA has anything to do with it in this case and you would encounter a few nags here and there with either one especially if you don't receive the latest BIOS fixes. The manufacturers do point out quite forcefully to check for and perform BIOS updates
For me personally, it isn't a matter of deciding whether to go for Asus or Gigabyte or MSI or any other manufacturer. It's more about the details of the chipset and at moment the nVIDIA chipsets match the criteria I'm looking for more closely than VIA or any other chipset producer when it comes the AMD processors. With Intel processors the Intel chipsets are the rule but the nVIDIA SLi technology is very popular as well.
As for not commenting, if folks don't respond to a question or comment, that equates to "no feedback w/out any implications" in my book. Otherwise I could draw an endless list of conclusions based on there having been no response. The written word is our only means of communication here. We don't have body language or other auditory or visual cues to go by. Just what we put in writing. IOW, The implications of an absents of a response are infinite. Therefore, the best way to convey your thoughts or opinions in this (a forum) context is to set them in writing. :smiles: Unfortunately, I didn't realize at the time the mobo/socket were new. I didn't do my homework and I have reaped what I sowed as a result. I'll know next time. You can count on that! I suppose I just need to change my approach a bit. I come from the old school of "leave the BIOS alone unless a change is warranted". Because the symptoms didn't, in my mind, lead me to think of the BIOS, I didn't consider it as the problem. So, flashing the BIOS never entered my mind. That's actually been a very valuable lesson learned for me w/ this whole evolution. That the chipset should be taken into consideration in higher priority than the motherboard manufacturer. W/ all the sage advice you all have given me, I'll definetly do a better job next time.
Thanx Again!!