A Crosshair in the Soup
Since I was building the X-Ray for the new project for the Computer Ed Show I thought lets use the best we can get and the Asus Crosshair series is widely regarded as the best of the AMD motherboards.
Once the system was up and running I launched into benchmarking, pushing the X6 to it’s limits. This was to be a none typical Computer Ed Build and I was determined to go all Enthusiast on this system. This was my first mistake.
You see over the years I have developed a system of testing that relies on looking at the system in use in the real world and I forgo traditional benchmarks. For this build however I was so bent on going enthusiast that I left my normal testing procedures behind and went at this with the typical enthusiast testing methods. I find it very telling BTW that traditional enthusiast sites have missed what I almost missed.
Once the base testing was done I set into building the X-Ray and started down the road again but this time using my usually methods. It was during this round of testing that I noticed that in some lower thread count apps the performance of the system was awful. I commented on this to my friend Paul Marini from Hi Tech Legion and he told me he had noticed the same issues, even brining them to AMD but the issues he saw where so minor he was persuaded that they would not matter in real world use.
As I continued my real world testing however more deeply the issue came into more clear focus. The real eye opener came when I fired up my current MMO, Star Trek Online for a gaming comparison. The game speed was AWFUL, I was lucky to see more than 30 FPS. This was on a system with 8 Gigs of RAM, a 5870 for video along with the 1090T playing at 1680×1050.
It took about and hour of testing to find the specific culprit. For whatever reason the 1090T was not coming out of Cool n Quiet mode. The cores where staying at 800 MHz and not stepping up as they came under load. I tested a few other games and did not see the issue and so submitted my findings to Cryptic and AMD.
As I waited for the companies responses I noticed that when using the Windows 7 built in ZIP for compression and decompression the cores where not kicking in as they should. Again the cores seemed to hover at 800 MHz. The only two ways I found around this issue was to either turn off CnQ in the BIOS or to artificially set the affinity for the game or folder window to two or three cores.
With this new information I knew we where not looking at an issue specific to STO and so started playing with various hardware settings to try and get the problem to stop. Today I got in a 1055T for testing and so stepped up my efforts. I grabbed an old Gigabyte 770 motherboard and dropped in the 1055T once I updated the BIOS. The chip worked perfect and CnQ performed as it should.
Next I moved the 1090T chip to the 770 board, again the CnQ worked perfect. Finally I put the 1055T in the Asus board, BINGO the problem immediately hit again. With this information in hand i was able to get a test copy of a newer BIOS from Asus and the problem continues to persist. I have sent all the material I have accumulated over the last few days of testing to AMD for them to forward to Asus.
Now this would normally be the end of the story, I would tell people to pass on the Crosshair right now unless they are aware of this issue and know they might have to work around it until Asus gets this resolved. However this incident got me thinking and so I went back over my older notes.
This board marks the third time that I have seen an Asus motherboard at the time of a new AMD processor release. This is the third time that I have found a BIOS glitch in the Asus board. This is a disturbing trend. Now I am not saying that anyone should not buy Asus boards but in each of these instances I was looking at one of their top boards. These where not inexpensive parts but the cream of the Asus crop.
I can tell you that I personally will recommend people hold off on Asus high end boards until they have had a few BIOS updates. This particular incident is not a show stopper. The issue does not show up on everything and so far I have found it is actually only happening in a few instances. However the fact is that when it does hit the processor is running at only a fraction of it’s potential. A less experienced DIYer could find himself in extreme levels of frustration with AMD and ASUS over an incident like this. It would be easy for a less experience tech to blame the processor in this case even.
Again let me say I love the features and look of the Crossfire IV. This is a serious enthusiast motherboard and has a feature set worth the cost. However until the BIOS issue is resolved I would suggest either waiting for your purchase or looking elsewhere.