Matrox RT.X100Xtreme with socket 939 mobo?

[Slushprodukt]

I just bought a new RT.X100 and put it into the system last night. After installing all the drivers and plug-ins, the system switched off, of its own accord. Every time I switch the computer on, Windows loads up and stays on for a minute or two and then switches itself off.

So I've taken out the RT.X100, but left all the drivers and plug-ins installed, and the computer hasn't switched off.

I have now realised how stupid I was not to check in advance whether the RT.X100 is compatible with my motherboard. It is the ECS KV2 Extreme.

I was intending to replace the motherboard with a better model anyway, as I want to have more RAM and the PCI Parhelia and so on. To this end, I have just ordered the AMD FX-57 processor.

The problem is, the FX-57 is socket 939, and I've just looked at the list of motherboards that have been validated for use with the RT.X100, and none of them are socket 939!

So my question is, what should I do? Any advice here would be appreciated. I really want to build the best, fastest and most stable system I can, and thanks to a recent windfall, money isn't much of an issue. So how should I proceed?

Please help!

[Dr Mordrid]

VIA chipsets are NOT compatable with the RT.X100, not even close. In fact they are the worst possible option. Neither are the NVIDIA or ATI chipsets. This is because of insufficient PCI to host memory bandwidth.

Unfortuneately it's also the one major aspect of mainboard performance that goes un-tested in reviiews because it requires $$ hardware to test

The RT.X100 lives and breathes by this performance indicator, so boards without sufficient bandwidth are not to be considered at all.

This info is on Matrox's mainboard and chipset compatability pages, save for the info about ATI which is my own personal experience. Unfortunately they haven't kept them updated to reflect the new chipsets and boards.

Chipsets that have a long history of working with the RT.X100 are those by Intel (no socket 939 there ) and SiS. ASUS makes a SiS 756 chipped socket 939 based board in the A8S-X;

http://www.asus.com/products4.aspx?modelmenu=2&model=755&l1=3&l2=15&l3=228

I have no personal experience with it and the RT.X100 yet, but plan to soon as I upgrade all my editing systems to Athlon 64 X2's.

One insight I've gained in my (many) years of beta'ing and using the RT.X100 is that cost matters not when it comes to mainboard compatability. I've seen $300 boards fail miserably and $60 boards run it like a jackrabbit (MSI 745 Ultra, ECS K7S5A, ECS K7S6A, all SiS chipped).

Basically: PCI to host memory bandwidth is where you find it, and it has nothing at all to do with the price tag.

Dr. Mordrid

[Fluff]

Sorry to jump topic a bit. But does the same PCI to host memory bandwidth issue apply with PCI soundcards for DAW use?

[Dr Mordrid]

Depends on the sampling rate. For high bandwidth audio it could well apply.

For 48/16 probably not much of an effect. For 96/24 I'd definitely go with as low latency a card as possible (M-Audio, Echo etc. for affordable DAW's) and a high bandwidth mainboard.

I'm currently using an Echo GINA24 in my DAW, but it's out of production. Their new products, however, are great. Same goes for M-Audio, which is now an AVID subsidiary....the company that makes studio video editing gear and software.

I'd never use a Creative card for serious work. Unfortunately their 24 bit support isn't really. That's why they had to settle a class action lawsuit earlier this year;

SINGAPOREAN soundcard maker, Creative Labs, has agreed to settle a class action lawsuit related to misleading marketing of its Audigy and Extigy range. Actually, it appears this settlement was agreed in principle at the end of last year, but few potential claimants yet know about it.

Creative claimed that the products in question could handle 24-bit audio at 96Khz – indeed this was stated on the product boxes in bold letters, and in all advertising. But complaints filed in 2003 pointed out that this was only true in a very limited set of circumstances, and pretty much all of the audio passing through the cards would actually be processed at lower quality.

The difference probably wouldn't concern the average gamer or casual MP3 enthusiast, but many of those planning to use the Creative cards for professional-quality audio were outraged by the labelling.

Nice, 'eh?

Dr. Mordrid

[Slushprodukt]

Chipsets that have a long history of working with the RT.X100 are those by Intel (no socket 939 there ) and SiS. ASUS makes a SiS 756 chipped socket 939 based board in the A8S-X;

I have no personal experience with it and the RT.X100 yet, but plan to soon as I upgrade all my editing systems to Athlon 64 X2's.

One insight I've gained in my (many) years of beta'ing and using the RT.X100 is that cost matters not when it comes to mainboard compatability. I've seen $300 boards fail miserably and $60 boards run it like a jackrabbit (MSI 745 Ultra, ECS K7S5A, ECS K7S6A, all SiS chipped).

Thanks very much for your in-depth reply, Dr Mordrid. I posted this question yesterday on five different forums and you are the only person who has replied!

Do you think my RTX is likely to have been damaged by being used with an incompatible motherboard? I have ordered the Asus mobo you suggested as it has all the specs I needed (4gb unbuffered RAM and socket 939 for FX-57 basically) and I'd like to know if you think the RTX is likely to be malfunctional since it has been in a system with an incompatible motherboard and, indeed, an ATI graphics card.

Thanks for replying,
Slush.

[Slushprodukt]

Dr. Mordrid,

I thought you would like to know something that I've been told by one of the Matrox advisors on the official forum regarding the Asus A8S-X...

This motherboard uses the SIS 965L South bridge. we tested the SIS 963L South bridge and found it to be incompatible with the RT.X100. Unfortunatly, this motherboard has not been tested and may or may not function properly with the RT.X100.

In any case, I have ordered the A8S-X and when it arrives, which I expect to be tomorrow, I will try using the RT.X100 with it and let you know if it works.

[Nowhere]

Hmm...that's interesting, because AFAI remember the board in which Doc used his RT.X succesfully had 963L southbridge

[Dr Mordrid]

I used it with good results in an ASROCK K7S8XE+ that was chipped with the SiS 748 northbridge and the 964 southbridge.

IMO Matrox has been pretty lazy the last couple of years in testing new boards with the RT.X100. Several of their tests seem to have been done on early or prototype boards and not production models. Not good because of the state of the BIOS etc.

I say this because I've run it sucessfully on production boards & chipsets that they have said it shouldn't run on.

Dr. Mordrid