In search of my software Crossfire, I asked ATI for help. What I got was rude and unconstructive support. I’ve had it with Richard C, giving me rude one-liners. This is how the exchange went. Please, do read on. 

Me:

Hello,

I have a GeCube (No GeCube in manufacturer list) HD2600XT 256MB and I want to use them under software Crossfire. However, my “Enable Crossfire” option in CCC is greyed out. Device manager detects two HD2600XTs.

Another thing, installing the HD2600XT and its drivers causes an unknown PCI device in device manager. When I removed one, only one unknown PCI device appeared. PCI databases indicated that the VEN_1002 points to ATI. How do I resolve this?

I hope there is a solution, especially to the Crossfire problem.

Thank you very much.
–John Perez

Related specifications:
1. MSI K9A Platinum (ATI Crossfire Certified)
2. HEC 550W Power Supply
3. AMD Athlon64 X2 5600+
4. Corsair 2GB XMS2

It took ATI 2 days to respond.

Richard C.:

For your combination of cards and motherboard you need to run the internal interconnect cables for the cards to achieve CrossFire.

Please see the below chart for reference.

http://ati.amd.com/technology/crossfire/CF_combo_chart_Aug07-1.jpg

Me:

I believe this chart is not updated. Inspecting the filename of the chart, it was created last August. The new drivers were released October. The Catalyst 7.10 drivers support software Crossfire. There are even websites that have successfully benchmarked HD2600XTs in software Crossfire. Similarly, there others that experience the same issue. Here is an excerpt from the 7.10 driver release notes:

“Software CrossFire™ support for the ATI Radeon™ HD 2600 and ATI Radeon™ HD 2400 Series

This release of Catalyst™ introduces software CrossFire™ support for the ATI Radeon™ HD 2600 Series and ATI Radeon™ HD 2400 Series. Software CrossFire™ is only available when two graphics accelerators from the same family (i.e. 2 ATI Radeon™ HD 2600 products) are plugged into a CrossFire™ compatible motherboard with no hardware interconnect between the two graphics accelerators.”

I have two videocards that are in the same family and I have a Crossfire certified motherboard. As for the unknown PCI device issue, I have solved it. Apparently it was the HD2600’s audio capabilities. Additional drivers solved the PCI device issue. I hope there is a solution for the Crossfire.

Richard C:

The chart is still the most current.  For your combination you would need the internal interconnects.  It is specifically the motherboard and its chipset.

Me:

Then why are the drivers claiming to enable software Crossfire? The release notes did not specify chipsets. Only Crossfire compatible chipsets.

How is it possible that others have enabled software Crossfire with other chipsets? OCWorkbench (http://www.ocworkbench.com/2007/amd/catalyst-7.10/b1.htm) is able to do so with the X38 chipset. And according to that chart, the X38 is not even compatible.

Topic 737-28799 seems to discuss the software Crossfire issue. But it is apparently missing.

Richard C.:

The chart lists supported configurations, I have had feedback of customers enabling CrossFire on SLI motherboards as well but it is not something we would support or troubleshoot. 

Different motherboards, different chipsets, different CrossFire configurations.  With your setup you need the internal connections.

I decided not to continue the conversation. Richard C. was simply too rude and narrow-minded. He definitely does not address the points I raised and refute my arguments. It is clear that he is avoiding the argument and is not attempting to even resolve or be constructive on the issue. I have evidence of software Crossfire with the HD2600 being done and evidence of the inability to enable Crossfire as well. He does not have any evidence to back-up his claims. I haven’t read any case of software Crossfire being difficult on 580X chipsets. I still believe that the chart he presented are dated. He simply waved it off.

As far as this debate goes, I can eat Richard C. alive. He is just not worth the time and energy. Richard won’t even take the time to consult ATI engineers to consider the possibility of a driver bug. As far as this support ticket is concerned, I can tell that I did my part as a customer. It is ATI and Richard C. that did not.

I’ve dealt with Richard C. before, and he wasn’t this pissy.

I’m sorry, ATI, but you just blew your customer service for me. Very poor. Very disappointing. Hardly got anything done. And it took your representatives a few days to think of something such as linking an outdated chart.

Two thumbs down.