uATX Update

[Gary Key]

Good Day...

I was held up an extra day on my trip late last week to AMD. While I would love to discuss what we saw, NDA prevents it. All I can say is that November through February will be a "very interesting time" for new product releases from AMD, this includes chipsets, GPUs, and of course CPUs.

On to the schedule -

NV7150 boards - I feel like I wasted the last three days testing these boards to some degree. We had numerous problems but after a whirlwind cycle of BIOS updates yesterday and an overnight marathon session of testing, I would say the boards are ready to be fully reviewed and released. 11/17**********

AMD690G -
Just about finished up with it but AMD provided new BIOS code last week (overclocking is excellent now if that is important) and the 7.9 drivers took advantage of some of the changes with improved IGP performance, especially in HD playback. I am retesting a few items now and will update tonight on the results and if we should take a another look at the IGP performance before publishing. I will provide a final update by tomorrow morning if we release it this week or wait until early next week and retest the five boards. Going up week of 11/17*****

G31/G965-
We have four new boards in-house and will publish the week of 11/26 ****.

G35-
It looks like early November now, will update once we have final silicon to test *** no changes. *** Boards are held again, drivers still not up to par, late November, move to December 5th launch now**********

Compatibility Chart-
This list of components that we tried in each uATX board will be published next week along with a pass/fail report for each board. This will become a standard feature in each motherboard article starting in mid-October. Finished, posting shortly.

uATX System Guide-
An overview of buildups including a SOHO, HTPC, and Gaming machine is coming shortly along with platform versus platform testing of each configuration against a similarly equipped ATX setup.

uATX Linux Testing -
Our first results will be published this month and we are looking to make this a monthly feature for all boards now.

Input Devices/Audio/Storage/Chassis/GPU Image Quality Results-
These articles will start in a couple of weeks and finish up in early November. I think the audio article will be the most interesting one and probably the most controversial.

 

[chucky2]

Thanks for the update Gary!

One thing I'd say regarding the 690G BIOS upgrades is that unless the manufactuers such as Asus, Gigabyte, MSI, Abit, etc. can commit they will provide the new BIOS's to us - the public - by time of article release, I don't think the new BIOS's should be used.

It's high time these manufacturers start being held publically accountable for BIOS's the public is forced to use that are not mature, or do not overclock well, or have incompatibility issues, etc.

While reading how great the new performance BIOS for 690G is for HD playback under Vista will be especially nice for someone like me who already runs a 690G board, what will not be nice is knowing it'll be another month or three before my mainboard vender decides I get access to it.

JMHO's...

Chuck

 

[Zap]

Originally posted by: Gary Key
the 7.9 drivers took advantage of some of the changes with improved IGP performance, especially in HD playback. I am retesting a few items now and will update tonight on the results and if we should take a another look at the IGP performance before publishing.
...
I think the audio article will be the most interesting one and probably the most controversial.

What I would like to see (besides the articles being published of course ) is testing the IGP against low end video cards. We all already know that high end cards provide a much better gaming experience. However, how much benefit is there to a low end card? I'm talking about $40 video cards. Besides gaming differences (if any) is there any difference in performance of non-gaming apps? I think this would be interesting information because there are people who use these really low end cards in non-gaming systems with IGP because everyone thinks that "onboard = slower." Would be nice if someone can quantify that.

Looking forward to the audio stuff... is it onboard versus PCI?

 

[Skott]

Originally posted by: chucky2
Thanks for the update Gary!

One thing I'd say regarding the 690G BIOS upgrades is that unless the manufactuers such as Asus, Gigabyte, MSI, Abit, etc. can commit they will provide the new BIOS's to us - the public - by time of article release, I don't think the new BIOS's should be used.

It's high time these manufacturers start being held publically accountable for BIOS's the public is forced to use that are not mature, or do not overclock well, or have incompatibility issues, etc.

While reading how great the new performance BIOS for 690G is for HD playback under Vista will be especially nice for someone like me who already runs a 690G board, what will not be nice is knowing it'll be another month or three before my mainboard vender decides I get access to it.

JMHO's...

Chuck

Here! Here! /cheer :thumbsup:

 

[Gary Key]

Hi,

1. Any of the BIOS updates that we will be testing tomorrow and Thursday will be released with the article. I am doing the same on the next P35 review as the BIOS we have totally blows away the current BIOS available for download from abit as an example. We will not review a board without providing a download link to the BIOS if it is not up on the supplier's website. That said, we are going to take one final pass at the 690G boards and will add in a few results with a 6000 or 6400+ just to show the top end from AMD now if anyone cares. I will provide an update tomorrow.

2. We will follow-up in the near future with a single GPU article showing the performance of the IGP solutions against the AMD 2400 Pro ($48) , NV 8400GS ($49), NV 8500GT ($55), and maybe a 2400XT ($56). The results are not pretty, mainly in this class, but your reason for upgrading will be related to video decoding, DVI/HDMI output, or running Tetris about 15FPS higher although a few games will benefit a little.

3. Gigabyte is shipping us a G33M-S2H board tomorrow so we can show the performance of the G33 with DVI/HDMI capabilities and with the latest Intel driver set, we are told 1080P should be a reality now. Image quality will be discussed but not shown as we still do not have a "legal" way of capturing the HD images and displaying them.

 

[Gary Key]

1. The first Gigabyte NV7150 board we received from NVIDIA was an early production sample as it turns out so that explains a few problems even after the BIOS updates. Retail boards are on their way next week so we will do a three page performance preview with the Biostar board we just received and follow-up with a mini-roundup later this month when the ASUS, MSI, Foxconn, and abit boards ship. This preview will at least provide some information regarding the chipset performance and a graphics quality comparison to the other IGP solutions in the market. I am hoping to get the preview ready for late Friday or a Saturday publish date.

2. We are working around the clock to update and retest the AMD690G boards. I expect to be finished with the five boards sometime tomorrow and will provide an update at that time. Realtek was suppose to release new HDMI audio drivers, but those were pulled at the last minute yesterday. We will publish next week. We will also post the BIOS releases we are using.

3. Still no confirmation on G35 being released this month but everything I hear now points to a November release with solid DX9 drivers. DX10 compatibility will not arrive until Q1 next year although I cannot think of a single DX10 game this chipset is capable of playing at decent frame rates, even at 800x600.

 

[Gary Key]

Originally posted by: chucky2


It's high time these manufacturers start being held publically accountable for BIOS's the public is forced to use that are not mature, or do not overclock well, or have incompatibility issues, etc.


JMHO's...

Chuck

No disagreement here..... :beer:

 

[imported_Kreed]

What about the NV7050 boards? I hope you haven't dropped them from the reviews.

 

[AmberClad]

Originally posted by: Zap
Looking forward to the audio stuff... is it onboard versus PCI?

Me too actually.

I've been meaning to add my old tuner card in, but I'd rather not put it right in front of my video card, and my sound card is currently in the other PCI slot. Between the upcoming uATX articles and the Creative article Anand posted last night, I wonder if the final nail in the coffin is coming for Creative vs Realtek...

 

[Zap]

Yeah, I read that article too. However, I think Creative can ride a while longer on the perspective of the enthusiast. After all, who else buys physics cards, extra video cards, Killer NICs, kilowatt power supplies... just because they "think" they need it.

Think about it. There are people who will never buy an AMD CPU because they had one bad experience with an unstable Athlon 700 on a crappy PCChips board, or who will never buy a Radeon because they had one game BITD which glitched on pre-Catalyst drivers.

These are the same people who had an Abit IS7 years ago who's onboard audio crackled, popped and buzzed, and since then have vowed to NEVER use onboard audio because it's the devil's work.

Anyways, to my ears the latest HD audio sound just fine, and I could never tell an FPS difference between onboard and PCI solutions.

 

[AmberClad]

Originally posted by: Zap
Anyways, to my ears the latest HD audio sound just fine, and I could never tell an FPS difference between onboard and PCI solutions.

Same here, to be honest, which is why I haven't upgraded to the X-Fi. And the way things are looking, it's unlikely I'll buy another discrete sound card ever again unless there's some major innovation to make that investment worthwhile.

Especially these days, I can't imagine a CPU being taxed noticeably by taking on the burden of audio processing. And from what I understand, the EAX advantage may be moot as far as Vista.

Reading the comments in the Creative article, it seems like they may be reluctant to work on developing onboard audio. They may be able to hang on for a few years longer, but I'm not sure where they'll be a few years from now.

 

[Zap]

Originally posted by: AmberClad
Reading the comments in the Creative article, it seems like they may be reluctant to work on developing onboard audio.

Some boards have used their chipsets, but generally speaking they tried to go their own way. Perhaps they wouldn't be in such dire straights if they had really designed for AC'97 BITD, and HD audio now. The margins would be pretty low per chip, but I'd imagine that the volume would be high because what motherboard manufacturer would not want to integrate a real Creative Labs chip into their enthusiast boards if it is just an easy drop-in for AC'97 or HD Audio at around the same cost as other solutions.

 

[Gary Key]

Originally posted by: Kreed
What about the NV7050 boards? I hope you haven't dropped them from the reviews.

No, they will come right after the 690G boards.

 

[Kenshin911]

Also, please look into Linux compatibility for the motherboards. I want to build a low cost mATX linux system with (maybe?) an integrated graphics chipset, but I don't want compatibility issues. I don't know what platform to go for because I don't know if any of the recent chipsets have decent linux support.

Thanks!

 

[heyheybooboo]

As difficult as it may be, I volunteer to subject myself to days and days of tedious MCP72XE testing at no cost to you.

It's my patriotic duty.

 

[chucky2]

Originally posted by: Gary Key
Hi,

1. Any of the BIOS updates that we will be testing tomorrow and Thursday will be released with the article. I am doing the same on the next P35 review as the BIOS we have totally blows away the current BIOS available for download from abit as an example. We will not review a board without providing a download link to the BIOS if it is not up on the supplier's website. That said, we are going to take one final pass at the 690G boards and will add in a few results with a 6000 or 6400+ just to show the top end from AMD now if anyone cares. I will provide an update tomorrow.

snip

Gary,

I really appreciate AnandTech doing this for us end users of the boards you're reviewing. However, I don't believe a board should be represented in the review with the new performance BIOS unless that performance BIOS is up on the actual manufacturers website. If they can publically take credit for the performance, then they can publically host that BIOS file to the rest of us.

You hosting it serves them in getting a better review, without providing us the customer with the means to achieve the same results in that review.

To me that is unacceptable.

There is quite simply no reason where they feel comfortable releasing a BIOS to you for you to give us results with, and then not be able to host that file to us themselves. Either they stand behind the BIOS or not. If so, host it. If not, then they shouldn't have their board reviewed with that BIOS.

Chuck

 

[Basilisk]

Originally posted by: Gary Key
1. Any of the BIOS updates that we will be testing tomorrow and Thursday will be released with the article. ...

The humor of this, regarding the 7150 boards, is -- unless I've totally blown Google, etc. searches -- that your BIOS's will be available but the 7150 boards aren't!?

Oh well... some day those boards may arrive. Be nice to see them have fewer problems than the 6xxx boards.

 

[Gary Key]

1. I will start releasing any BIOSs used in the article and not on the manufacturer's website with the article contents. If the manufacturers want us to use them, then you should be able to also.

2. Nothing but hell with the NV7150 boards this week, the retail boards from MSI and Gigabyte will not arrive until Tuesday or so. The Biostar board is the only one working correctly and that is with a new BIOS today.

3. We had a heck of a time with the AMD690G/NV7050 boards the last couple of days with S3 resume and USB floppy drives and optical drives, turns out it was Vista and not the BIOSs we have retested the last three days non-stop, download this hotfix - 94160 HotFix - as it solves numerous problems with just about any board and Vista, seems especially targeted to NV and AMD chipsets.

More later....

 

[imported_Kreed]

Gary,

I know I am being selfish here, but can you please publish the AMD690/NV7050 reviews before the NV7150. As you have said, the NV7150 are quite problematic right now and it would make more sense to me to concentrate on boards that can actually be purchased when the review is published. Of course, if there are other forces at hand that are impacting your decision, then I totally understand.

 

[Gary Key]

Hi,

I will publish the 690G/NV7050 before the NV7150.

The 690G/NV7050 is complete from a test viewpoint, running some additional background tests on 4GB configs under Vista64, having some compatibility problems that are not addressed yet. The article finally goes live next week. After that one, the NV7150 and the catch up with the new DVI/HDMI G33 boards and a couple of new G965 boards. I am happy to report that those five boards have all passed the 100 hour mark without a real problem.

 

[8steve8]

any update on g35? when will we see this chipset for sale at newegg.

are we talking in a few days, or nov1 or dec1 or nov15 or? ... or xmas?

 

[FrodeNilsen]

Originally posted by: Gary Key
Hi,

I will publish the 690G/NV7050 before the NV7150.

The 690G/NV7050 is complete from a test viewpoint, running some additional background tests on 4GB configs under Vista64, having some compatibility problems that are not addressed yet. The article finally goes live next week. After that one, the NV7150 and the catch up with the new DVI/HDMI G33 boards and a couple of new G965 boards. I am happy to report that those five boards have all passed the 100 hour mark without a real problem.

Do you have any idea if AMD is going to fix the 690g for vista64?

It cannot run using 64bit memory addressing. Is the 7050 also affected?

 

[imported_Kreed]

Originally posted by: 8steve8
any update on g35? when will we see this chipset for sale at newegg.

are we talking in a few days, or nov1 or dec1 or nov15 or? ... or xmas?

Given that the Intel Eaglelake G45 is due in Q2 2008, you would think that the G35 would be out very soon.

 

[gman81]

Over on the avsforum, someone has posed the question about an 8400GS - is it good enough for "higher bitrate 720p or 1080i MPEG2 .TS content. The HTPC basically starts choking on these .TS files with stuttering and dropped frames. The CPU usage doesn't seem to be very high (around 30% if I recall)"? I understand that Gary is going to compare a couple IGP boards with one that uses a low end card like the 8400GS. Just thought Gary was listening and could throw .TS files at the test systems.

 

[imported_Kreed]

Originally posted by: Gary Key
The 690G/NV7050 is complete from a test viewpoint, running some additional background tests on 4GB configs under Vista64, having some compatibility problems that are not addressed yet. The article finally goes live next week. After that one, the NV7150 and the catch up with the new DVI/HDMI G33 boards and a couple of new G965 boards. I am happy to report that those five boards have all passed the 100 hour mark without a real problem.

Gary, please tell me you will publish the article this week. Surely, most of the issues would have been resolved by now.