... the quadraplegic would win. It's bad enough that AMD can't compete with Intel, but do they have to go and shoot themselves in the foot as well? And do they have to do it EVERY TIME that they launch a new product?
If AMD ever hope to compete with Intel again, they need to start working harder on their QA process instead of just shoveling sh!t out the door.
Personally I'd call ATI's chipsets pretty competitive with Intel's. The CPU is not the only factor on a PC these days, if anything now more than ever the present performance levels make the other features more important particularly on a HTPC.
I don't blame "AMD" so much as the ATI team, and this is also why I don't mind building a HTPC with an AMD CPU and nVidia chipset. Slower than some w/Intel CPUs, yes, but still plenty fast enough for my HTPC use and it cost less to build.
8ch LPCM over HDMI.
means i can hook up my htpc to my reciever with hdmi 1.3 and get both 1080p video and 7.1 channel (everything but TrueHD/DTS-MA) and be happy.
its the same way ps3's work. instead of sending the lossless encoded HD audio (trueHD/DTS-MA), it decodes it and sends 8ch of LPCM through.
so much for speding $500 for an AVR just to get those logos.
I am still amazed that people are willing to build HTPCs. Buy discrete components and have it JUST WORK. No screwing with drivers, incompatibilities, and companies that don't even know their own products. I am so sick and tired of PC crap that doesn't WORK. *#@%(&(*@#&^
The thing is, you have to start appreciating that a HTPC, even if every last feature doesn't work, still adds substantial features that discrete components don't.
The catch is you just need to avoid being anal about what doesn't work the same as always when dealing with such rapidly evolving hardware (regardless of 785 being only an incremental update).
If everything were delayed until it was perfect it would leave us without all the other gains we enjoy at a rapid pace... kind of amazing how far PC tech has come over the last couple decades.
also, according to amd, "Not all features will be supported on all machines. Check with your PC manufacturer for specific model capabilities and supported technologies."
that means that there is a possibility that it is enabled on some boards manufacturers and not on others. just like some boards only support am3/ddr3 vs others have am2+/am2 and the cheaper ddr2.
just called msi but their tech support doesnt even have white paper on their new 785/am3/ddr3/sideport mobo. the support tech didnt even know that it was on sale on newegg already. so when he said that the 8ch LPCM through HDMI is based on the software, i could barely hold out on laughing at his face.
and gary, epic fail. regardless, thanks for manning up.
You have a good point - life is much simpler with discrete components. I also built an HTPC (mostly out of spare parts) and I hardly use it. Its neat to run google earth, or emulated games on the big screen once-in-a-while, but it's far easier just to use dicrete components for watching movies, and using my computer for everything else.
I don't get the big deal with lack of 7.1 sound over HDMI...Most people I know who would spend the money on a 7.1 sound system would also spend the money on a Blueray player and other discrete hardware.
Like the others, I applaud you for owning up to your mistake in such a visible and public way. The correction was not silently made to the original article, nor was it buried in some obscure place. Yes, that is why AnandTech.com is trusted and is part of my morning routine.
I also hope you don't beat yourself up over it too much.
When you were working under time constraints, was it reasonable to skip a couple of items you had already tested? Yes. Can that come back and bite you? Obviously. But you could also make a similar judgment call, accept a reasonable risk, a hundred times and not encounter a problem.
After all, it is not like you never tested it, and fabricated test results. You did test it, and merely assumed it would continue to work. After I have my brakes fixed on my car, I assume the turn signals and passenger door will continue operating as they had before.
A bit of egg on your face. A lesson learned. Cheer up and keep up the good work. We look forward to your next article.
Had Gary repeated the ENTIRE test suite, he would have (probably) noticed the problem, unless there is a problem with his test setup as AMD has suggested. Personally, I trust Gary.
That's why I said a "bit" of egg on his face. Only a small bit. He got caught doing a 99% job instead of his usual 110%.
@In fact, we stuck to our guns that we would only utilize retail BIOS releases for testing. The three sample boards that AMD shipped for review all contain excellent BIOS’, but they were hand tuned at AMD.@
Huh? So, does it work with boards provided and "hand tuned" by AMD, or not?
hmm ... so I am back to 790GX/SB750 boards with DDR2 memory (Asrock A790GMH/128M).
Anyway the whole 785G thing is (as another reader said) similar to the G92 rebrands by Nvidia. There is not one feature that would warrant an upgrade from 780G.
I hope this mess will unravel itself. It does seem that AMD did NOT communicate clearly with the reviewers and media people.
Note that some of the 785G chip shots in the the various reviews showed date codes from late '08. Packaged silicon for this chip has existed for a LONG time...
If this:
" In fact, we stuck to our guns that we would only utilize retail BIOS releases for testing. The three sample boards that AMD shipped for review all contain excellent BIOS’, but they were hand tuned at AMD. "
means "LCPM 7.1. works with boards provided by AMD", then this whole story stinks quite a bit.
Actually, we had LPCM 7.1 working on a board not provided by AMD. I did not use the boards provided by AMD in the review since they did not contain public release BIOS'. ASUS committed to posting the BIOS releases that AMD used on the press samples so we will see the results next week.
That said, I will go to my grave believing that it was working correctly, even though AMD believes based upon my setup that it was actually multi-channel S/PDIF across HDMI and their drivers did not set the flags properly for the A/V receiver to interpret it correctly. I am an old fart but I know my ears heard gunshots across the rear surrounds in the final gun scene in Crank. ;)
The fact that you posted a second article instead of silently correcting the original only solidifies my belief in AT. Mistakes happen and some of the "other" review sites would not have acted as honestly. Great job Gary!!! AT is still one of the few "morning coffee" sites on my list.
... or did you just see what your receiver displayed? It's pretty easy to just run the Vista (or 7) channel test. You really can't trust the receiver. I've had 5.1 PCM show over S/PDIF-limited HDMI, but the center didn't work, and the fronts apparently were being duplicated into the backs, so it didn't violate the 2-channel PCM of S/PDIF.
LOL that's what I was going to say. AT, like every other real tech site out there, has been pulling for AMD while they're going through this rough patch, hoping to get competition back like in the Athalon days.
Disappointing that AMD just lost the low-cost HTPC market. Bad move on their part.
What do you mean "just lost"? nVidia 8*** and 9**** series chipset based motherboards already had that market preference and the retraction mentioned it as well as prior articles.
Even so, plenty of people are satisfied with 780G for HTPC (some even find 760G w/o HD acceleration acceptable since not everything can be fully accelerated anyway so you have to have the CPU muscle for more versatile use) despite the missing feature but I have to wonder if history is only repeating itself, just another ATI driver bug that may or may not be fixed before the chipsets' reasonable lifecycle is over.
This is ridiculous on multiple levels... (and I don't mean AT's reporting, which is typically very solid.)
"An audio block exists which limits LPCM output to two-channels"??? So, AMD is doing this crap ON PURPOSE?!?! They lost how much money last quarter? And now they're effectively telling people, "go buy Intel products"? Unbelievable.
First, I'd like to say classly retraction, owning up is always hard, but, proper. Kudos to that.
Second, I hope this serves as a final lesson learned in testing and reporting results to your AnandTech community. ONLY BIOS's and Drivers that are publically available on the websites of these products should be used - NOTHING MORE. I have said on past comments to you, that if these manufacturers see fit foist upon the general public their products with the BIOS and Drivers they give us, then they need to learn to live with the results you display to millions of AnandTech readers.
If that means they get 40% less performance, it crashes twice a day in testing: Oh well!!! That's what you should be reporting.
To keep going back to these manufacturers for BIOS's and Driver's to correct things you're finding - and respectfully, if you're finding them, they should have found them long ago when they did their QC work - is just plane insane. It's wasting your time in re-testing, in doing other reviews, and it sends a completely wrong message to them that they can ship sh1t but their reviews will be Golden.
Often AT gets pre-released versions of the hardware. This serves two purposes, it allows for objective, credible reviews by those outside the company, but also serves as its own test bed by the user community. Sort of like releasing beta version of Windows to the public.
There is no way manufacturers can test for every scenario; there are so many different hardware/software configurations. AT has its own software/testbed that it uses that are proprietary to AT. If AT let out its methods, the tests would not be as credible, since products could be tailored to better performance. Not to mention, it would allow the user community the ability to set up their own similar tests and report them, thus being competition to AT.
In any case, it seems AMD was either providing false information, or at some expo the exclusive enthusiast community leaked false information amongst themselves.
Regardless, AMD is most likely doing testing all throughout the time they send out these pre-released versions.
From reading Gary's post, it looks like the problem wasn't that he used drivers and BIOS revisions that weren't available to the public, it was that a feature (Multichannel LPCM) that was working (or he believed to have working) in an earlier release wasn't properly tested with the latest driver and BIOS revisions that were published in the review.
In fact, Gary states that he in fact DID ensure that the BIOS included in the review would be available to the public.
"This also occurred with the BIOS releases. In fact, we stuck to our guns that we would only utilize retail BIOS releases for testing. The three sample boards that AMD shipped for review all contain excellent BIOS’, but they were hand tuned at AMD. Only ASUS has committed that the BIOS release on their review sample boards will be available to the public. The BIOS releases we utilized on the two Gigabyte 785G boards are what shipped on the boards that we purchased from Newegg when they first went on sale. "
"Enjoy the latest audio technologies using HDMI with 7.1 digital surround sound support."
"Let your desktop PC be the center of your entertainment lifestyle with full 1080p HD playback and 7.1 surround sound."
There's more to this story than just "it doesn't work." I wouldn't be the slightest bit surprised if at some point down the road, AMD magically enables 7.1 over HDMI. This wreaks of a faulty driver issue or a bad spin of silicon. At some point this will get fixed, but until then, my plans to buy one of these motherboards for a HTPC are on hold.
If this is simply a case of "multi-channel LPCM is on some boards but not others" then no problem, that is clearly stated on AMD's website.
If this is a case of "it's a standard feature but we just haven't had time to perfect it yet in the drivers" then it's annoying but not unprecedented. It's a new chipset and these things do happen.
If this is a case of "it doesn't work in this spin of silicon so were keeping our mouths shut until the next spin is ready" then AMD is being disingenuous and deserves our scorn.
If this is a case of AMD intentionally being deceptive on it's website and letting false information and expectations spread, then AMD needs to go bankrupt.
''There's more to this story than just "it doesn't work." I wouldn't be the slightest bit surprised if at some point down the road, AMD magically enables 7.1 over HDMI. This wreaks of a faulty driver issue or a bad spin of silicon. At some point this will get fixed, but until then, my plans to buy one of these motherboards for a HTPC are on hold.''
It relates to number four in the foot note which states "Not all features will be supported on all machines. Check with your PC manufacturer for specific model capabilities and supported technologies."
Probably AMD is stating 8 channel LPCM as an option to let motherboard manufactures and PC makers decide to include it which will be in software. Eight channel LPCM might be supported in 785G, but they do not want to say for every board or else DMCA could bring AMD to court and more resources is taken out for new chipset models. The 780G chipset is an open chipset for digital audio because how it passes audio through HDMI.
I have used a 780G chipset with Linux and I experience that a SXRD projection TV does not always output audio that is coming from HDMI. I do not think the 785G is not any different, but it includes eight channel output only if it complies to standards which means encryption have to be done in software.
My HTPC is on hold but not because of this problem. It is because of money. I am also waiting for more motherboards that includes sideport memory and DDR3 in a ATX motherboard with 785G.
Hi,
Yes, I use TMT3 in testing, at times I prefer it to PDVD 9. We still have the same problem with a 2-channel lock on LPCM. At this point we are getting conflicting reports from AMD about what was/should be/is working on this chipset. The left hand has not met the right hand as of a couple of hours ago. ;)
In order for eight channel LPCM to work, encryption have to be used. It is OK to use two channel LPCM without encryption. If it is not OK to use unencrypted eight channel LPCM because it is not in compliance.
Go back to the Anandtech's article that explains about the requirements of eight channel LPCM.
It seems AMD is providing the responsibilities on the software developers to make sure the data being transfer over HDMI is in compliance.
With out encryption, DCMA is involved because a "black box" can be used in between a consumer player and monitor to extract the data to computer. This digital copy is illegal to do. It brings in DCMA and copyright agencies to cover this issue.
Your software that you suggested may not be in compliance to handle eight channel LPCM. It is most likely they are hoping for the sound card manufacture to do this for them.
Thank you for the prompt clarification. This is why I read AT everyday. I was going to grab one of these mobo's, and probably still will, but won't receive a nasty surprise upon installation.
It's process and essentially feature identical to 780G, and in an apples to apples DDR2 comparison performs essentially the same or worse.
Is it just a "GTS 250" type rebrand to get some headlines prior to LGA-1156? If so, AMD can suck it for wasting our time and ruining your weekend. Fake "New Release!" PR is the epitome of lame.
Do you even read the article? The 785G north bridge uses a newer UVD2 core, which does reduce CPU utilization over the 780G with Blu-Ray playback. However, the rest of the the north bridge and the SB710 south bridge still sound like the GTS 250 deal.
I don't understand why AMD wouldn't want to have this feature. It sounds like the chip is perfectly capable of doing it and they have to add a block so it can't be used.
Oh well, i guess that's why I'm an engineer and not in marketing.
Obviously, they had to admit they were wrong, it wasn't a matter of character, or anything else. Actually, the blog is disingenous. He apologizes for mistakes made by AMD, to a large extent. First it worked, then it didn't, and then AMD rushed the release. They are all excuses, and his apology is a bit disingenous. Not that I doubt him, it probably was AMD's fault. They're sloppy these days. I guess it's too be expected with a company that's in a death spiral and bleeding cash like crazy. Still, this kind of stuff doesn't help them.
There was no covering it up. Other sites had said the same thing. A retraction was necessary.
I used to hope AMD would stay around as competition, but with what they're doing, let's hope they go belly-up, and let ATI go again. AMD is run by monkeys these days. They need to die. Someone else with better management will buy the scraps of the company, and run it correctly. I still think that company should be IBM.
I still don't understand their product line at all. Does anyone? They have goofy as Hell names, that make no sense to anyone. The 785 shouldn't have come out alone. It's better in some things than the 790, but not in others. Confused customers don't buy. It's pretty remedial. They should have come out with a 785, and 795, and discontinued the two others. Their whole naming system is nuts. Not that Intel's is much better, but at least I know if I buy a G45, it's better than a G41. Their processor names are cryptic as Hell though, but, probably not as bad as AMD's. Then again, with the Lynnfield now spanning a few names, I think they'll pass AMD in the confusing naming system department.
I personally have little to no interest in the feature in question so it makes no difference to me either way, but thank you all the same for a quick and honest correction. This is why you guys remain so well trusted.
I'm with you on that feature. As a matter of fact I was not impressed by this chipset at all. After breifly reading through the article, I see no need to pdate my 780G board in my beadroom PC.
Gary, all my due respect to you and the entire crew for the timely update.
You are certainly not the only reviewer touting 7.1 channel stereo on this chipset. I wonder how this got generally misreported. You say it is not available on the current Bios' and drivers. Will this be available through driver or bios updates?
I am still working with AMD and the manufacturers on this whole fiasco. In fact, several of the board suppliers were just as surprised as we were about this feature disappearing. Early roadmaps, drivers, and BIOS' had 7.1 support or listed it as supported. Up until this morning, AMD's own PR page and white paper listed support.
Theories abound at various R&D departments if it is a software lock or something changed during mass production of the chipset. AMD's own internal departments said one story and then another. Bottom line is that I did not complete regression testing before the review went live and if I had, well, our story would have been correct, but confusing I imagine. :)
While you look as mistakes in a bad way, I actually think that it gives AT more credibility in the fact that you acknowledge when one is made and (if there is one) you notify your readers.
So long as there aren't too many, we're not too upset.
Agreed, you guys did the right thing and it's reasons like these that you guys are on my top 5 sites to look at everyday.
Still, what the hell is AMD thinking? Long before the 785G was released everybody was saying it would have 7.1 channel LCPM support. They didn't step up to anyone before the boards went out and said, "Hey this doesn't work."?
"Still, what the hell is AMD thinking? Long before the 785G was released everybody was saying it would have 7.1 channel LCPM support. They didn't step up to anyone before the boards went out and said, "Hey this doesn't work."? "
What? Because "everybody" was spreading false or at least only partially correct information doesnt make it AMDs fault. They apparently never claimed it would work with *every* product. And you can bet that it wont before long we'll see this feature in some shape or form...
I'm not saying it was their fault I am saying it was a very foolish move on AMD's part to not correct the misinformation because it hurts their credibility. Heh, Though I was probably stretching it a little bit with the "everybody" part.
AMD probably just has too much stuff on their PR plate right now to worry about to notice rumors being spread about future chipsets on forums in the community. Either way, you have an excellent point, AMD suffers from this the most.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
65 Comments
Back to Article
rocky5 - Thursday, August 13, 2009 - link
Does anyone know what brand/if any supports this feature now?rocky5 - Thursday, August 13, 2009 - link
Just brought the Gigabyte MA785G-UD3H MB.It has a little sticker on the box that reads
Blu-ray Disc Full Rate Lossless Audio
Is this not right?
juampavalverde - Wednesday, August 5, 2009 - link
Well, seems that there's no reason to update my 780g to 785g, phew...ipay - Wednesday, August 5, 2009 - link
... the quadraplegic would win. It's bad enough that AMD can't compete with Intel, but do they have to go and shoot themselves in the foot as well? And do they have to do it EVERY TIME that they launch a new product?If AMD ever hope to compete with Intel again, they need to start working harder on their QA process instead of just shoveling sh!t out the door.
medi01 - Thursday, August 6, 2009 - link
Well, if they cannot compete, is not because their products, but because of perception of the user base. 100$ AMD CPU vs 100$ Intel CPU, AMD wins.Heck, AMD didn't manage to get more than a third of the market when having VASTLY superior product (P4 vs Athlon 64s era)
mindless1 - Wednesday, August 5, 2009 - link
Personally I'd call ATI's chipsets pretty competitive with Intel's. The CPU is not the only factor on a PC these days, if anything now more than ever the present performance levels make the other features more important particularly on a HTPC.I don't blame "AMD" so much as the ATI team, and this is also why I don't mind building a HTPC with an AMD CPU and nVidia chipset. Slower than some w/Intel CPUs, yes, but still plenty fast enough for my HTPC use and it cost less to build.
jabber - Wednesday, August 5, 2009 - link
I buy the motherboard...plug in my X-Fi or similar like most folks would and....??What exactly is so upsetting here?
declan croix - Wednesday, August 5, 2009 - link
8ch LPCM over HDMI.means i can hook up my htpc to my reciever with hdmi 1.3 and get both 1080p video and 7.1 channel (everything but TrueHD/DTS-MA) and be happy.
its the same way ps3's work. instead of sending the lossless encoded HD audio (trueHD/DTS-MA), it decodes it and sends 8ch of LPCM through.
so much for speding $500 for an AVR just to get those logos.
jabber - Wednesday, August 5, 2009 - link
Oh, is that all?pervisanathema - Wednesday, August 5, 2009 - link
I am still amazed that people are willing to build HTPCs. Buy discrete components and have it JUST WORK. No screwing with drivers, incompatibilities, and companies that don't even know their own products. I am so sick and tired of PC crap that doesn't WORK. *#@%(&(*@#&^mindless1 - Wednesday, August 5, 2009 - link
The thing is, you have to start appreciating that a HTPC, even if every last feature doesn't work, still adds substantial features that discrete components don't.The catch is you just need to avoid being anal about what doesn't work the same as always when dealing with such rapidly evolving hardware (regardless of 785 being only an incremental update).
If everything were delayed until it was perfect it would leave us without all the other gains we enjoy at a rapid pace... kind of amazing how far PC tech has come over the last couple decades.
declan croix - Wednesday, August 5, 2009 - link
this is integrated. not discrete.also, according to amd, "Not all features will be supported on all machines. Check with your PC manufacturer for specific model capabilities and supported technologies."
that means that there is a possibility that it is enabled on some boards manufacturers and not on others. just like some boards only support am3/ddr3 vs others have am2+/am2 and the cheaper ddr2.
just called msi but their tech support doesnt even have white paper on their new 785/am3/ddr3/sideport mobo. the support tech didnt even know that it was on sale on newegg already. so when he said that the 8ch LPCM through HDMI is based on the software, i could barely hold out on laughing at his face.
and gary, epic fail. regardless, thanks for manning up.
kmmatney - Wednesday, August 5, 2009 - link
You have a good point - life is much simpler with discrete components. I also built an HTPC (mostly out of spare parts) and I hardly use it. Its neat to run google earth, or emulated games on the big screen once-in-a-while, but it's far easier just to use dicrete components for watching movies, and using my computer for everything else.I don't get the big deal with lack of 7.1 sound over HDMI...Most people I know who would spend the money on a 7.1 sound system would also spend the money on a Blueray player and other discrete hardware.
snakeoil - Wednesday, August 5, 2009 - link
so you made a ''sort of'' review and now you blame amd for your what's supposed to be your job?i hope this doesn't happen again.
justaviking - Wednesday, August 5, 2009 - link
Gary,Like the others, I applaud you for owning up to your mistake in such a visible and public way. The correction was not silently made to the original article, nor was it buried in some obscure place. Yes, that is why AnandTech.com is trusted and is part of my morning routine.
I also hope you don't beat yourself up over it too much.
When you were working under time constraints, was it reasonable to skip a couple of items you had already tested? Yes. Can that come back and bite you? Obviously. But you could also make a similar judgment call, accept a reasonable risk, a hundred times and not encounter a problem.
After all, it is not like you never tested it, and fabricated test results. You did test it, and merely assumed it would continue to work. After I have my brakes fixed on my car, I assume the turn signals and passenger door will continue operating as they had before.
A bit of egg on your face. A lesson learned. Cheer up and keep up the good work. We look forward to your next article.
dingetje - Wednesday, August 5, 2009 - link
it seems to me the egg is on amd's face and gary did a good jobjustaviking - Wednesday, August 5, 2009 - link
Had Gary repeated the ENTIRE test suite, he would have (probably) noticed the problem, unless there is a problem with his test setup as AMD has suggested. Personally, I trust Gary.That's why I said a "bit" of egg on his face. Only a small bit. He got caught doing a 99% job instead of his usual 110%.
AMD has much more left to explain.
yacoub - Wednesday, August 5, 2009 - link
It's okay, we still love you Gary.medi01 - Wednesday, August 5, 2009 - link
@In fact, we stuck to our guns that we would only utilize retail BIOS releases for testing. The three sample boards that AMD shipped for review all contain excellent BIOS’, but they were hand tuned at AMD.@Huh? So, does it work with boards provided and "hand tuned" by AMD, or not?
Gary Key - Thursday, August 6, 2009 - link
No, I checked.MrPoletski - Wednesday, August 5, 2009 - link
Is a wonderful thing./pats author on the back.
Proteusza - Wednesday, August 5, 2009 - link
Is AMD saying that the 785G is technically capable of multi channel LPCM output, but that its disabled at a software level? Why?haplo602 - Wednesday, August 5, 2009 - link
hmm ... so I am back to 790GX/SB750 boards with DDR2 memory (Asrock A790GMH/128M).Anyway the whole 785G thing is (as another reader said) similar to the G92 rebrands by Nvidia. There is not one feature that would warrant an upgrade from 780G.
Let's just hope the DX11 GPUs will be good.
solgoldberg - Wednesday, August 5, 2009 - link
Yes, I agree: back to the 790GX with the SB750.I hope this mess will unravel itself. It does seem that AMD did NOT communicate clearly with the reviewers and media people.
Note that some of the 785G chip shots in the the various reviews showed date codes from late '08. Packaged silicon for this chip has existed for a LONG time...
dingetje - Wednesday, August 5, 2009 - link
"Our recommendation for this audience continues to be the GF9300/9400 motherboards paired with an E6300 or E7200 processor."^ PWND !!
medi01 - Wednesday, August 5, 2009 - link
If this:" In fact, we stuck to our guns that we would only utilize retail BIOS releases for testing. The three sample boards that AMD shipped for review all contain excellent BIOS’, but they were hand tuned at AMD. "
means "LCPM 7.1. works with boards provided by AMD", then this whole story stinks quite a bit.
Gary Key - Thursday, August 6, 2009 - link
Actually, we had LPCM 7.1 working on a board not provided by AMD. I did not use the boards provided by AMD in the review since they did not contain public release BIOS'. ASUS committed to posting the BIOS releases that AMD used on the press samples so we will see the results next week.That said, I will go to my grave believing that it was working correctly, even though AMD believes based upon my setup that it was actually multi-channel S/PDIF across HDMI and their drivers did not set the flags properly for the A/V receiver to interpret it correctly. I am an old fart but I know my ears heard gunshots across the rear surrounds in the final gun scene in Crank. ;)
atbeliever - Tuesday, August 4, 2009 - link
The fact that you posted a second article instead of silently correcting the original only solidifies my belief in AT. Mistakes happen and some of the "other" review sites would not have acted as honestly. Great job Gary!!! AT is still one of the few "morning coffee" sites on my list.andy o - Tuesday, August 4, 2009 - link
... or did you just see what your receiver displayed? It's pretty easy to just run the Vista (or 7) channel test. You really can't trust the receiver. I've had 5.1 PCM show over S/PDIF-limited HDMI, but the center didn't work, and the fronts apparently were being duplicated into the backs, so it didn't violate the 2-channel PCM of S/PDIF.snakeoil - Tuesday, August 4, 2009 - link
its clear that this site is on the intel's payroll.anyway.
james jwb - Tuesday, August 4, 2009 - link
it's also clear you are trolling.Ryanman - Tuesday, August 4, 2009 - link
LOL that's what I was going to say. AT, like every other real tech site out there, has been pulling for AMD while they're going through this rough patch, hoping to get competition back like in the Athalon days.Disappointing that AMD just lost the low-cost HTPC market. Bad move on their part.
mindless1 - Wednesday, August 5, 2009 - link
What do you mean "just lost"? nVidia 8*** and 9**** series chipset based motherboards already had that market preference and the retraction mentioned it as well as prior articles.Even so, plenty of people are satisfied with 780G for HTPC (some even find 760G w/o HD acceleration acceptable since not everything can be fully accelerated anyway so you have to have the CPU muscle for more versatile use) despite the missing feature but I have to wonder if history is only repeating itself, just another ATI driver bug that may or may not be fixed before the chipsets' reasonable lifecycle is over.
medi01 - Wednesday, August 5, 2009 - link
Yep, low cost HTPC guys are exactly those guys, who do care about passing "7.1 true HD sound" over HDMI, right.Donny Bahama - Tuesday, August 4, 2009 - link
This is ridiculous on multiple levels... (and I don't mean AT's reporting, which is typically very solid.)"An audio block exists which limits LPCM output to two-channels"??? So, AMD is doing this crap ON PURPOSE?!?! They lost how much money last quarter? And now they're effectively telling people, "go buy Intel products"? Unbelievable.
Meanwhile there are manufacturers out there who continue to tout multi-channel audio over HDMI - http://usa.asus.com/999/images/products/2129/1.jpg">http://usa.asus.com/999/images/products/2129/1.jpg -- Does it seem odd to anyone else that not just AnandTech, but also ASUS (and who knows how many other mfrs) had misconceptions about this chipset?
Griswold - Wednesday, August 5, 2009 - link
Think again, toolboy.chucky2 - Tuesday, August 4, 2009 - link
Gary,First, I'd like to say classly retraction, owning up is always hard, but, proper. Kudos to that.
Second, I hope this serves as a final lesson learned in testing and reporting results to your AnandTech community. ONLY BIOS's and Drivers that are publically available on the websites of these products should be used - NOTHING MORE. I have said on past comments to you, that if these manufacturers see fit foist upon the general public their products with the BIOS and Drivers they give us, then they need to learn to live with the results you display to millions of AnandTech readers.
If that means they get 40% less performance, it crashes twice a day in testing: Oh well!!! That's what you should be reporting.
To keep going back to these manufacturers for BIOS's and Driver's to correct things you're finding - and respectfully, if you're finding them, they should have found them long ago when they did their QC work - is just plane insane. It's wasting your time in re-testing, in doing other reviews, and it sends a completely wrong message to them that they can ship sh1t but their reviews will be Golden.
PLEASE STOP THIS PRACTICE!!!
Chuck
vol7ron - Tuesday, August 4, 2009 - link
Chuck,Often AT gets pre-released versions of the hardware. This serves two purposes, it allows for objective, credible reviews by those outside the company, but also serves as its own test bed by the user community. Sort of like releasing beta version of Windows to the public.
There is no way manufacturers can test for every scenario; there are so many different hardware/software configurations. AT has its own software/testbed that it uses that are proprietary to AT. If AT let out its methods, the tests would not be as credible, since products could be tailored to better performance. Not to mention, it would allow the user community the ability to set up their own similar tests and report them, thus being competition to AT.
In any case, it seems AMD was either providing false information, or at some expo the exclusive enthusiast community leaked false information amongst themselves.
Regardless, AMD is most likely doing testing all throughout the time they send out these pre-released versions.
-vol7ron
cghebert - Tuesday, August 4, 2009 - link
From reading Gary's post, it looks like the problem wasn't that he used drivers and BIOS revisions that weren't available to the public, it was that a feature (Multichannel LPCM) that was working (or he believed to have working) in an earlier release wasn't properly tested with the latest driver and BIOS revisions that were published in the review.In fact, Gary states that he in fact DID ensure that the BIOS included in the review would be available to the public.
"This also occurred with the BIOS releases. In fact, we stuck to our guns that we would only utilize retail BIOS releases for testing. The three sample boards that AMD shipped for review all contain excellent BIOS’, but they were hand tuned at AMD. Only ASUS has committed that the BIOS release on their review sample boards will be available to the public. The BIOS releases we utilized on the two Gigabyte 785G boards are what shipped on the boards that we purchased from Newegg when they first went on sale. "
acejj26 - Tuesday, August 4, 2009 - link
http://www.amd.com/us/products/desktop/chipsets/7-...">http://www.amd.com/us/products/desktop/...integrat..."Enjoy the latest audio technologies using HDMI with 7.1 digital surround sound support."
"Let your desktop PC be the center of your entertainment lifestyle with full 1080p HD playback and 7.1 surround sound."
There's more to this story than just "it doesn't work." I wouldn't be the slightest bit surprised if at some point down the road, AMD magically enables 7.1 over HDMI. This wreaks of a faulty driver issue or a bad spin of silicon. At some point this will get fixed, but until then, my plans to buy one of these motherboards for a HTPC are on hold.
lifeblood - Wednesday, August 5, 2009 - link
If this is simply a case of "multi-channel LPCM is on some boards but not others" then no problem, that is clearly stated on AMD's website.If this is a case of "it's a standard feature but we just haven't had time to perfect it yet in the drivers" then it's annoying but not unprecedented. It's a new chipset and these things do happen.
If this is a case of "it doesn't work in this spin of silicon so were keeping our mouths shut until the next spin is ready" then AMD is being disingenuous and deserves our scorn.
If this is a case of AMD intentionally being deceptive on it's website and letting false information and expectations spread, then AMD needs to go bankrupt.
jmurbank - Tuesday, August 4, 2009 - link
''There's more to this story than just "it doesn't work." I wouldn't be the slightest bit surprised if at some point down the road, AMD magically enables 7.1 over HDMI. This wreaks of a faulty driver issue or a bad spin of silicon. At some point this will get fixed, but until then, my plans to buy one of these motherboards for a HTPC are on hold.''It relates to number four in the foot note which states "Not all features will be supported on all machines. Check with your PC manufacturer for specific model capabilities and supported technologies."
Probably AMD is stating 8 channel LPCM as an option to let motherboard manufactures and PC makers decide to include it which will be in software. Eight channel LPCM might be supported in 785G, but they do not want to say for every board or else DMCA could bring AMD to court and more resources is taken out for new chipset models. The 780G chipset is an open chipset for digital audio because how it passes audio through HDMI.
I have used a 780G chipset with Linux and I experience that a SXRD projection TV does not always output audio that is coming from HDMI. I do not think the 785G is not any different, but it includes eight channel output only if it complies to standards which means encryption have to be done in software.
My HTPC is on hold but not because of this problem. It is because of money. I am also waiting for more motherboards that includes sideport memory and DDR3 in a ATX motherboard with 785G.
wiak - Tuesday, August 4, 2009 - link
DCMA has nothing to do with multi-channel lpcmi think its a software problem aka PowerDVD, Driver or BIOS problem
@anandtech have you guys tried TotalMedia Theatre 3?
http://www.arcsoft.com/public/software_title.asp?P...">http://www.arcsoft.com/public/software_...ductID=3...
so if 785G was gonna be released late auguest, you can say its to early for it to support LPCM acording to bios, drivers,software updates ;)
Gary Key - Thursday, August 6, 2009 - link
Hi,Yes, I use TMT3 in testing, at times I prefer it to PDVD 9. We still have the same problem with a 2-channel lock on LPCM. At this point we are getting conflicting reports from AMD about what was/should be/is working on this chipset. The left hand has not met the right hand as of a couple of hours ago. ;)
jmurbank - Wednesday, August 5, 2009 - link
In order for eight channel LPCM to work, encryption have to be used. It is OK to use two channel LPCM without encryption. If it is not OK to use unencrypted eight channel LPCM because it is not in compliance.Go back to the Anandtech's article that explains about the requirements of eight channel LPCM.
It seems AMD is providing the responsibilities on the software developers to make sure the data being transfer over HDMI is in compliance.
With out encryption, DCMA is involved because a "black box" can be used in between a consumer player and monitor to extract the data to computer. This digital copy is illegal to do. It brings in DCMA and copyright agencies to cover this issue.
Your software that you suggested may not be in compliance to handle eight channel LPCM. It is most likely they are hoping for the sound card manufacture to do this for them.
xfile - Tuesday, August 4, 2009 - link
Thank you for the prompt clarification. This is why I read AT everyday. I was going to grab one of these mobo's, and probably still will, but won't receive a nasty surprise upon installation.GeorgeH - Tuesday, August 4, 2009 - link
Given this news, what's the point of 785G?It's process and essentially feature identical to 780G, and in an apples to apples DDR2 comparison performs essentially the same or worse.
Is it just a "GTS 250" type rebrand to get some headlines prior to LGA-1156? If so, AMD can suck it for wasting our time and ruining your weekend. Fake "New Release!" PR is the epitome of lame.
quanta - Thursday, August 6, 2009 - link
Do you even read the article? The 785G north bridge uses a newer UVD2 core, which does reduce CPU utilization over the 780G with Blu-Ray playback. However, the rest of the the north bridge and the SB710 south bridge still sound like the GTS 250 deal.QChronoD - Tuesday, August 4, 2009 - link
I don't understand why AMD wouldn't want to have this feature. It sounds like the chip is perfectly capable of doing it and they have to add a block so it can't be used.Oh well, i guess that's why I'm an engineer and not in marketing.
Marquis - Tuesday, August 4, 2009 - link
The measure of a person's character is not how they act when they are right, but how they act when they are wrong.You, sir, have proven to have character in spades.
TA152H - Wednesday, August 5, 2009 - link
Obviously, they had to admit they were wrong, it wasn't a matter of character, or anything else. Actually, the blog is disingenous. He apologizes for mistakes made by AMD, to a large extent. First it worked, then it didn't, and then AMD rushed the release. They are all excuses, and his apology is a bit disingenous. Not that I doubt him, it probably was AMD's fault. They're sloppy these days. I guess it's too be expected with a company that's in a death spiral and bleeding cash like crazy. Still, this kind of stuff doesn't help them.There was no covering it up. Other sites had said the same thing. A retraction was necessary.
I used to hope AMD would stay around as competition, but with what they're doing, let's hope they go belly-up, and let ATI go again. AMD is run by monkeys these days. They need to die. Someone else with better management will buy the scraps of the company, and run it correctly. I still think that company should be IBM.
I still don't understand their product line at all. Does anyone? They have goofy as Hell names, that make no sense to anyone. The 785 shouldn't have come out alone. It's better in some things than the 790, but not in others. Confused customers don't buy. It's pretty remedial. They should have come out with a 785, and 795, and discontinued the two others. Their whole naming system is nuts. Not that Intel's is much better, but at least I know if I buy a G45, it's better than a G41. Their processor names are cryptic as Hell though, but, probably not as bad as AMD's. Then again, with the Lynnfield now spanning a few names, I think they'll pass AMD in the confusing naming system department.
Rolphus - Tuesday, August 4, 2009 - link
I personally have little to no interest in the feature in question so it makes no difference to me either way, but thank you all the same for a quick and honest correction. This is why you guys remain so well trusted.jonup - Wednesday, August 5, 2009 - link
I'm with you on that feature. As a matter of fact I was not impressed by this chipset at all. After breifly reading through the article, I see no need to pdate my 780G board in my beadroom PC.Gary, all my due respect to you and the entire crew for the timely update.
ratbert1 - Tuesday, August 4, 2009 - link
You are certainly not the only reviewer touting 7.1 channel stereo on this chipset. I wonder how this got generally misreported. You say it is not available on the current Bios' and drivers. Will this be available through driver or bios updates?Gary Key - Thursday, August 6, 2009 - link
I am still working with AMD and the manufacturers on this whole fiasco. In fact, several of the board suppliers were just as surprised as we were about this feature disappearing. Early roadmaps, drivers, and BIOS' had 7.1 support or listed it as supported. Up until this morning, AMD's own PR page and white paper listed support.Theories abound at various R&D departments if it is a software lock or something changed during mass production of the chipset. AMD's own internal departments said one story and then another. Bottom line is that I did not complete regression testing before the review went live and if I had, well, our story would have been correct, but confusing I imagine. :)
Eeqmcsq - Tuesday, August 4, 2009 - link
I was just about to ask that, because EVERY article I've read so far claimed that the 785G was capable of 8-channel over HDMI.nafhan - Wednesday, August 5, 2009 - link
Maybe they saw Anandtech listed 7.1 sound as a feature, and didn't check it for themselves?Just kidding! That would never happen...
p1agu3 - Wednesday, August 5, 2009 - link
I wonder how often this happens in other news...hmmm... ;Dvol7ron - Tuesday, August 4, 2009 - link
While you look as mistakes in a bad way, I actually think that it gives AT more credibility in the fact that you acknowledge when one is made and (if there is one) you notify your readers.So long as there aren't too many, we're not too upset.
Seramics - Wednesday, August 5, 2009 - link
Yup, Anandtech is my no. 1 hardware site as far as credibility is concerned. Anand, u hv my support all the way man. Keep up the good work!oldabelincoln - Tuesday, August 4, 2009 - link
Thanks for prompt action. That's all we can ask for when mistakes are made - and the always will me made.Excellent job.
Ryun - Tuesday, August 4, 2009 - link
Agreed, you guys did the right thing and it's reasons like these that you guys are on my top 5 sites to look at everyday.Still, what the hell is AMD thinking? Long before the 785G was released everybody was saying it would have 7.1 channel LCPM support. They didn't step up to anyone before the boards went out and said, "Hey this doesn't work."?
Seriously bad move.
Griswold - Wednesday, August 5, 2009 - link
"Still, what the hell is AMD thinking? Long before the 785G was released everybody was saying it would have 7.1 channel LCPM support. They didn't step up to anyone before the boards went out and said, "Hey this doesn't work."? "What? Because "everybody" was spreading false or at least only partially correct information doesnt make it AMDs fault. They apparently never claimed it would work with *every* product. And you can bet that it wont before long we'll see this feature in some shape or form...
Ryun - Wednesday, August 5, 2009 - link
I'm not saying it was their fault I am saying it was a very foolish move on AMD's part to not correct the misinformation because it hurts their credibility. Heh, Though I was probably stretching it a little bit with the "everybody" part.p1agu3 - Wednesday, August 5, 2009 - link
AMD probably just has too much stuff on their PR plate right now to worry about to notice rumors being spread about future chipsets on forums in the community. Either way, you have an excellent point, AMD suffers from this the most.