Anandtech - Problems with 125W AMD CPUs on budget boards.

[smilingcrow]

Anandtech has an article showing the problems that they had running 125W TDP Athlon X2s and X4s on boards with 3 or 4 phase power; they died very quickly.
Not a problem for over-clockers as they generally go for higher spec boards but people looking for a mATX board with integrated graphics may hit this problem: Anandtech.

[andyb]

I only skimmed the first page, but that article looks like a load of toss.

Point 1:Who goes out and buys a £160 CPU, and uses it with is supreme integrated graphics on a £35 motherboard.???

Point 2: I didnt see anywhere in their "comments" (it sure as hell wasnt a review) what other components they were using, it is easily possible that they have a dodgy CPU, or more likely a dodgy PSU.

Point 3: Take their "comments" with a piece of salt, and have a look around the net for other users with 125W CPU's and those motherboards (if any end users with that combination exist) and see whether their boards have died.


Andy

[jaganath]

Point 1:
Who goes out and buys a £160 CPU, and uses it with is supreme integrated graphics on a £35 motherboard.???

such people do exist:
viewtopic.php?t=46929&start=30

[andyb]

Yes, they will exist, but are not that common.

Anyway, my point is that "Nighthog" has also been suckered in with Anandtech's "comments", they have provided no proof at all.

For All I know they were running the system off of a Pico 120 with a 90W power brick and thats why they fried boards, or maybe they were testing them on a metal desk covered in iron filings. Their "comments" are as worthless as reviews I have seen of things that have never been used.

Such as this shyte review - WARNING, link to a 4.5Mb PDF.

http://www.ultimatepc-reviews.co.uk/PDF ... eviews.pdf

I will be happy to admit that I am wrong if anyone had done any serious testing and even listed the hardware that they used, and proven that these boards that claim to be compatable with these 125W CPU's are not compatable at all.


Andy

[andyb]

I forgot to add point 4.

Point 4: They "claimed" to have blown mobos whilst trying to overclock, this is a revelation, I suppose that this is the first and only time Anandtech have managed that


Andy

[leo_fischer]

Anandtech has been around for over a decade, and they are one of the most thorough and respected sites around. They do this stuff for a living, so if they say there's a problem, I'd belive there's a problem. They set out to benchmark these boards and some didn't last more than 2 seconds with a 125W cpu installed (after being tested exhaustively with several other cpus.)
You didn't even bother to read the article before commenting!

[smilingcrow]

I posted this here because quite a lot of SPCR people seem to use IGPs which potentially might lead them to encounter this problem.
It’s not so unusual to use a cheap mATX board and with the Phenom Quads being inexpensive this combination will appeal to people looking for good performance on a tight budget. IMO the high TDP of Phenom X4s makes them relatively unattractive for the SPCR crowd but there are plenty of AMD enthusiasts that might fancy a crack at one.

As Anandtech stated this is not an AMD problem but an issue caused by the motherboard manufacturers. If Intel’s mainstream chips were consuming as much power as AMD’s then we would likely see the same problem on the Intel platform.

I think the article is worth reading even if just to remind people to check carefully on ‘exactly which CPUs a motherboard supports’ and not rely on the headline features list which might state something like ‘supports quad core’ or ‘the latest 45nm CPUs’.
I was considering ordering an Abit or Asus P35 board and after reading this article I double checked and neither board lists support for the Q9300. The Asus lists the Q9450 but not the Q9300 which surprised me.

[lucas82]

Maybe these motherboards aren't suposed to run the CPUs at full load all the time, they are targeted to offices, HTPCs, etc.

[nitram_tpr]

Goodness, do people here believe that a respected site such as Anandtech would be scaremongering?
What they are pointing to (if you can be bothered to actually read the article) is (accidental) misinformation by the motherboard manufacturers.
Here is what the overview for the Gigabyte GA-MA78GM-S2H says (amongst others):

Supports AMD AM2+ Phenom/ AM2 Athlon64 series processors

So, for a casual browser thinking about getting the best bang for your buck they are advertising Support for the new Phenom processors but in actual fact they do not support the 125W processors.

Anandtech are not just whacking in the fastest processor and imeediately overclocking, they are testing a wide range of processors.

Here are some quotes for the lazy people who can't be arsed to read the whole article.

Our normal course of testing has us installing a wide variety of processors in each board, regardless of the target market. We do this to ensure compatibility, and at times (like now) we wish this was not the case. This week, we tested the 780G boards with the LE1600, 4400+ X2, 4850e X2, 6400+ X2, Phenom 9600BE, Phenom 9900, and now the 9850BE.

and

We discovered quickly that running the 9900/9850BE or 6400+ X2 on these products resulted in the loss of the board, in a matter of a few seconds to a few minutes. Granted, it will probably be rare that a user will purchase a 9850BE to run on this platform, but in case you were considering that course of action, we highly suggest you do not. Let’s get this out of the way quickly; it is not a 780G chipset problem. In fact, it is not strictly a board problem either, but rather a design issue.

The vast majority of the 780G boards have a three-phase or four-phase PWM circuitry design. These designs are completely acceptable for the 45W, 65W, 89W, and 95W TDP rated processors; however, drop in a 125W TDP processor such as the Phenom 9850e or 6400+ X2 and you are asking for trouble. Trouble is exactly we found, as each board we tested eventually succumbed to the greater power requirements of these processors.

What did we learn? Do not trust the product information and specification pages in the vast majority of cases. The CPU support pages tell the real story - some better than others, but in all cases the 125W TDP processors are not supported by the current 780G motherboards. It just takes some effort to find that information (except for J&W) and this is something we do not think the user should have to do. In our opinion, the product pages lead the user down one path while the CPU support pages (which are sometimes difficult to find) tell another story. At this point, it pays to read the fine print or hope your favorite review site is able to provide this information.

This last point is the most important. If you are looking to buy a new motherboard, do you always go to the manufacturers website and look up the full tech specs and supported processors list? No, I don't suppose you do. Especially on a board that advertises support for the latest processors.

I agree with some of the sentiment regarding who would put a 125W TDP processor in a 'budget' motherboard. Well, there will always be someone who will do this. After all why spend double and sometimes triple the price on a motherboard when (apparently) the cheaper one will support the more powerful processor.

[Nighthog]

I wouldn't call me getting suckered....

I trust Anandtech and what they say, and I will wait and see. I don't want to have all the parts home and then see it was as they said and I had wasted my money. I had planned a full new computer and everything together would have costed quite a bunch and then maybe beeing usless parts that won't work together and no more money...

I went whit the best choice of not getting anything and stoping all my orders.

I've read over at Fudzilla that DFI might have a m-ATX 780g motherboard comming. Thinking off DFI it should be something to look forward to.

As it's now there really isn'y anything but budget styled m-ATX boards for AMD which is kinda disapointing.

[andyb]

I dont dislike Anandtech, I am as wary of their reviews as every discerning buyer should be.

But at this point in time I would say that yes they are "scaremongering".

The reason why I say that they are scaremongering is that they have only given the bad news, with no other information whatsoever.

When they decide to reward their avid readers and follower with their testing platform and methodology you can believe them, until then be sceptical.


Andy

[smilingcrow]

I dont dislike Anandtech, I am as wary of their reviews as every discerning buyer should be.

Having read Anantech for many years my discernment says that they are a highly reputable site.

But at this point in time I would say that yes they are "scaremongering". The reason why I say that they are scaremongering is that they have only given the bad news, with no other information whatsoever.
When they decide to reward their avid readers and follower with their testing platform and methodology you can believe them, until then be sceptical.

The article is a precursor to a bigger article but it seems as if they wanted to share their experience in advance to warn people of potential problems. If you wait for the main article I’m sure they’ll give full details of their test setup as is their usual policy. In fact if you can’t wait then email them as I’ve found their writers helpful in the past when I’ve contacted them about certain reviews.

For All I know they were running the system off of a Pico 120 with a 90W power brick and thats why they fried boards, or maybe they were testing them on a metal desk covered in iron filings.

There’s a big difference between scepticism and keeping an open mind. Your position seems to be closed to the possibility of there being a problem. I posted this because it’s the sort of information that I’d want to know if I was in the market for a similar setup. It’s premature to come to a conclusion at this point and I certainly haven’t. Why don’t we wait until Anandtech publish their full article before jumping to firm conclusions?

[Avalanche]

Thanks for posting this, SC. I have also read Anandtech for ~ 10 years and I believe this is an important problem if they posted a separate editorial to address it. I don't know why Andy is disrespecting them. Clearly from a SPCR point of view they don't do good acoustics tests, but I'm sure they were not "running the system off of a Pico 120" or "testing them on a metal desk covered in iron filings."

One thing that I think people might be glossing over... These CPU / MB combinations are not just incompatible. These combinations fry the motherboard!

[smilingcrow]

One thing that I think people might be glossing over... These CPU / MB combinations are not just incompatible. These combinations fry the motherboard!

I think you’ve hit the nail on the head there and that’s probably the main reason Anandtech published their preliminary findings. It’s better for everybody to reduce the chance of this happening.

[andyb]

The article is a precursor to a bigger article but it seems as if they wanted to share their experience in advance to warn people of potential problems. If you wait for the main article I’m sure they’ll give full details of their test setup as is their usual policy.

It may be the precursor to a bigger article, but anyone who says that they have fried xxx No of motherboards with xxx CPU and have provided no information about their test system, or testing methodology should be disregarded simply for the fact that they have nothing more than circumstantial evidence.

As I have said before, I am willing to admit that I spoke out of turn and I am wrong if Anandtech "prove" that these motherboards are being fried by a CPU that the motherboard manufacturers claim will work. If the xxxx 125W CPU is on that motherboards supported list I will make a public apology in this very post.

Until that time, I say that it is worthless to believe their words in this context with no evidence at all.

I think you’ve hit the nail on the head there and that’s probably the main reason Anandtech published their preliminary findings.

Shyte newspapers print "preliminary findings" all of the time, look at how often they are wrong. Yes I totally agree that Anandtech should have posted their problems earlier rather than later, but they should have atleast done so with a degree of transparancy, knowing only 2 of the items that they are testing is far from transparent.

Why did they not publish half a review, simply stating that several boards have failed during their testing, at least then Page 1 would have included the other components they are/were testing with.

Or they could have fried a couple of board with 125W CPU's and then finished the review with the remaining boards and sub-125W CPU's - or did they decide to "fry" each of the boards after the first couple failed, just to see how long they would last

I'm sure they were not "running the system off of a Pico 120" or "testing them on a metal desk covered in iron filings

Why are you so sure that all of those boards failed because they cant work with 125W CPU's - because Anandtech told you.? So because Anandtech have NOT told you, you are SURE that they were testing using a 100% healthy PSU.

That is a kind of faulty logic

Until Anandtech actually publish the review we wont know what or how they were testing these motherboards with, and until then this ridiculous bicker will continue until I get bored - or there is some kind of proof.


Andy

[juamez]

The article is a precursor to a bigger article but it seems as if they wanted to share their experience in advance to warn people of potential problems. If you wait for the main article I’m sure they’ll give full details of their test setup as is their usual policy.

It may be the precursor to a bigger article, but anyone who says that they have fried xxx No of motherboards with xxx CPU and have provided no information about their test system, or testing methodology should be disregarded simply for the fact that they have nothing more than circumstantial evidence.

As I have said before, I am willing to admit that I spoke out of turn and I am wrong if Anandtech "prove" that these motherboards are being fried by a CPU that the motherboard manufacturers claim will work. If the xxxx 125W CPU is on that motherboards supported list I will make a public apology in this very post.

Until that time, I say that it is worthless to believe their words in this context with no evidence at all.

Do you even know what "their words" are? I found this:

What did we learn? Do not trust the product information and specification pages in the vast majority of cases. The CPU support pages tell the real story - some better than others, but in all cases the 125W TDP processors are not supported by the current 780G motherboards. It just takes some effort to find that information (except for J&W) and this is something we do not think the user should have to do. In our opinion, the product pages lead the user down one path while the CPU support pages (which are sometimes difficult to find) tell another story. At this point, it pays to read the fine print or hope your favorite review site is able to provide this information.

Personally, I was very disappointed in the type of information available on the websites and in the product manuals. Sure, a manufacturer can hide behind the CPU support lists, but the information provided on the product description and specification pages would lead the majority of users to think using any Phenom or X2 processor is perfectly acceptable when it is not. One would think the manufacturers would be especially sensitive to this problem unless they just enjoy the RMA process and pissed off customers.

The only thing they are stating is that the motherboard manufacturers claim they support the new phenoms, period. This is actually not the case. CPU Support pages for most mobo's state that 125W TDP cpu's are not supported, contradicting the claim that the motherboard supports "all new phenoms". Anandtech calls to their readers to be weary of this kind of marketing. This is what's Anandtech pointing out, nothing more.

I think you’ve hit the nail on the head there and that’s probably the main reason Anandtech published their preliminary findings.

Shyte newspapers print "preliminary findings" all of the time, look at how often they are wrong. Yes I totally agree that Anandtech should have posted their problems earlier rather than later, but they should have atleast done so with a degree of transparancy, knowing only 2 of the items that they are testing is far from transparent.

Why did they not publish half a review, simply stating that several boards have failed during their testing, at least then Page 1 would have included the other components they are/were testing with.

Or they could have fried a couple of board with 125W CPU's and then finished the review with the remaining boards and sub-125W CPU's - or did they decide to "fry" each of the boards after the first couple failed, just to see how long they would last

Better safe than sorry? I know I'd rather be.

Also it's better to post nothing than to post a half complete review - with possible errors included - mind the angry motherboard companies sueing Anandtech for the false accusations, HA!

I'm sure they were not "running the system off of a Pico 120" or "testing them on a metal desk covered in iron filings

Why are you so sure that all of those boards failed because they cant work with 125W CPU's - because Anandtech told you.? So because Anandtech have NOT told you, you are SURE that they were testing using a 100% healthy PSU.

That is a kind of faulty logic

It is true that they have not included the used setup to test their claim, BUT they grant Anandtech credibility because of their 10yr flawless (?) reputation.

Until Anandtech actually publish the review we wont know what or how they were testing these motherboards with, and until then this ridiculous bicker will continue until I get bored - or there is some kind of proof.


Andy

Anandtech merely tries to warn their readers, not spook them. They post some premature findings and expect readers to look out for themselves, nothing more. They do certainly not claim to posess total and godly wisdom - which I think you actually believe.


Bottomline: Anandtech - a reputable site - posts some important findings and decides to share it with the world before any more damage will be done in the time the article is being published in its completeness. Many users take a note of it, but you - as a knight against Anandtech fanaticism and blind, radical belief of everything Anandtech posts - goes for the Quad Scepticism power-up and tries to mow down all the infidel Anandtech followers with the BFG-of-truth-and-justice.

No seriously, people actually can think for themselves. That's in my truly humble opinion also what Anandtech want's you to do. They shout something that will change your way of looking at numerous 780G motherboards, only for your own good. Even if they are bribed by some mobo manufacturer, their flawed results will be uncovered in their flawedness sooner or later by individuals that decide to redo the testing for themselves and compare their findings with those of Anandtech's.

So, steady on, old boy.

[andyb]

Well said "juamez".


Andy

[smilingcrow]

'Four MSI µATX boards do support 125W Phenoms' – Fudzilla.

[Avalanche]

We spoke to J&W about this issue, and they explained the Mosfets may handle a 9850 when properly cooled, but most customers will stuff such boards in small cases and even when not overclocked the VRM will fail and blow up eventually because of bad airflow in such cases.

With all the safety features in hardware these days, it surprises me that this could happen when combining two physically compatible parts.

[andyb]

The concept of support is ridiculous, either the motherboard supports xxx CPU or it does not support it, but it seems that not all of the problems are down to the motherboard manufacturer.

Did AMD state to the mobo manufacturers that they were going to release a 125W CPU.??? I doubt it, I strongly suspect that it was only released so they could claim they had a 2.5GHz Quad-Core CPU - even though they had to bump the voltage and the CPU draw to the max.

Anyway, I only build with Gigabyte motherborads at the moment, and I trust what they say (or dont say), I am currently using the GA-MA69VM-S2 for most builds. They state that the 125W x2 CPU's will work, and that the 95W x4's will work, but have not yet declared the 125W x4's as being valid.

To my mind this is sensible practice to have a huge list of supported and unsupported, but they should be clearer on their marketing than they currently are - such as having a link directly to they supported CPU's page.


Andy

[smilingcrow]

With the 9850BE having an unlocked multiplier it is clearly aimed at over-clockers. So if you release a board that supports this CPU and has over-clocking facilities which seemingly most do these days even mATX boards, the question is how much juice will the board be required to deliver to enable it to over-clock and stay stable?
That must surely be on the mind of board makers with regard to mATX and budget systems.
Once you start pushing over 3GHz with these CPUs the power consumption is going to put it in water cooling territory only I imagine. Has anyone seen any power consumption data for these chips when running at 3GHz+?

[dragmor]

There a few people getting the new B3's to 3ghz+ with stock volts and cooling. Lots of 3.2ghz with high end air.

But I'm more interested in how far these undervolt.

[smilingcrow]

But I'm more interested in how far these undervolt.

Anyone seen data on that yet?

Anandtech have an update on the original article posted on their Blog.

[juamez]

+3GHz is not bad at all! Can you post the voltage increase that was needed to attain that high of a clockspeed?

[Firetech]

Just posting to add my vote of confidence in Gary Key. Given what that guy has been through in the last year, I would trust his testing and work ethic 100%.

Sorry AndyB, I think your viewpoint on this one is wrong.
It may be worth reading some of his previous reports and forum posts to get a bit of background on the guy, then decide if he's the type to scaremonger.

[Nighthog]

Concerning the Gigabyte GA-MA78GM-S2H board they have now updated their cpu support list.

It supports all AMD cpu's except Phenom X4 9850.

Take a look yourselfs. It suports all other 125W like X4 9750, FX-62, X2 6400+

http://www.gigabyte.com.tw/Support/Moth ... uctID=2758

I might now try to build a 9750 whit this board instead of my orginal 9850 plan. As 9750 is supported I should get no problems, though OC in any matter is probably not so wise.

Will 2.4Ghz be enough to play any 1080p video whitout hardware accelration(no help from elsewhere but cpu making the work?)

[Firetech]

Now they're saying the 9950 will be a 140w unit. Better get a Delta High Speed over those mosfets....

[juamez]

Now they're saying the 9950 will be a 140w unit. Better get a Delta High Speed over those mosfets....

Or maybe even better: wait for the 45nm quadcores, who are supposedly much better.