Bookmark and Share

Articles: Mainboards

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 ]

Performance in Nominal Mode

We are going to test the new ASRock X58 SuperComputer mainboard on the following platform:

  • Mainboard: ASRock X58 SuperComputer, rev. 1.02, BIOS P1.30, P1.40, L1.44;
  • CPU: Intel Core i7-920 (2.66GHz, 133MHz base frequency, 8MB L3 cache, Bloomfield, rev. C0, 1.225V Vcore);
  • Memory: 3 x 1024MB DDR3 Kingston HyperX DDR3-1866, KHX14900D3T1K3/3GX, (1866MHz, 9-9-9-27, 1.65V voltage);
  • Graphics card: ATI Radeon HD 4870 512MB (RV770, 750/750/3600MHz, 800 SP, 40 TMU, 16 ROP, 256-bit 512MB GDDR5);
  • HDD: Samsung SP2504C (250GB, SATA II, 7200RPM, 8MB, rev. A);
  • CPU cooler: Cooler Master GeminII (120mm Protechnic Electric MGA12012HB-O25 fan, 1500-2500RPM);
  • Thermal interface: Noctua;
  • PSU: Antec NeoPower Neo HE 550 rev. A4 (550W);
  • Case: Antec Skeleton.

As I have already said, ASRock X58 SuperComputer mainboard is one of the flagship solutions. It costs about the same as the mainboards of the same kind from AsusTek or Gigabyte. Therefore, it is quite natural that we will be comparing ASRock X58 SuperComputer against one of these board, namely against Gigabyte GA-EX58-Extreme. We have recently reviewed Gigabyte mainboards: GA-EX58-Extreme and GA-EX58-UD5. Their functionality turned out almost the same and they made a very good impression.

Unfortunately, my excitement about the functionality of the ASRock X58 SuperComputer BIOS disappeared when I got to deal with it face to face. Let’s start with the “Smart” section that I liked so much. It turned out that the board works just fine in safe mode with the default settings, but failed even to boot with optimized parameters. In the meanwhile, we were promised an advantage of ASRock X58 SuperComputer over other mainboards in this particular optimized mode. We managed to partially solve the mystery of this optimized operational mode by replacing the Kingston HyperX DDR3-1866 memory with other modules. It turned out that the board sets the memory frequency at 1333MHz in the optimized mode, and not to 1066MHz like other boards. True, in this case we will see that ASRock X58 SuperComputer does better in some tests than its competitors, especially in synthetic memory tests. But why doesn’t the optimized mode work for Kingston HyperX DDR3-1866 memory remained a mystery. I did check that the board works just fine with this particular memory at 1333MHz when this frequency was set manually without any voltage increase or timings adjustment.

Speaking of the memory, we have to say that the board didn’t support the XMP (Extreme Memory Profile) technology. However, we could have guessed it from looking at the BIOS, since the board assumes that the memory frequency is 1066MHz for the first profile and 1600MHz for the second. However, we know from our Kingston HyperX DDR3-1866 Review that the modules SPD contain respective 1866 and 1800MHz frequencies. As a result, no matter what profile we picked, ASRock X58 SuperComputer just wouldn’t boot.

However, the incompatibility we discovered between the ASRock mainboard and high-speed memory is not fatal. As we have just learned from ASRock directly, the company engineers are aware of the problem and are working to resolve it. Next BIOS versions will allow using XMP profiles and Performance mode with optimized settings in any configuration without limitations. At this time, however, users have to do a little work and adjust the settings manually in order to achieve higher performance.

By the way, the “Boot Failure Guard” technology that monitors the POST process and reboots the board in case of failure worked impeccably. Right after a failed boot-up or at least after several failed attempts the board would boot in safe mode and display a warning message offering to access the BIOS and correct the parameter settings.

It was especially pleasing, because the Clear CMOS jumper on ASRock X58 SuperComputer is not much fun to work with: it would reset all the parameters including date and time. If you have previously saved the profile in the BIOS, you will be able to restore the settings quickly. However, you will still have to change the date and times manually.

But let’s get back to the memory. It is not really a big deal that the XMP technology doesn’t work on ASRock X58 SuperComputer mainboard. It did work on Gigabyte GA-EX58-Extreme, but I still configured most of the parameters manually, because the board increased the voltages way to much in my opinion. However, it turned out that the maximum available memory frequency on ASRock mainboard was only 1333MHz. Even at 1600MHz memory frequency setting the board refused to boot, not to mention 1866MHz, which is the nominal frequency for the Kingston HyperX DDR3-1866 modules. It was not enough to set the desired frequency and increase the memory voltage, Gigabyte GA-EX58-Extreme required an increased QPI bus voltage, too. The problem is that ASRock mainboard doesn’t know to adjust this voltage, there is no parameter like that in the BIOS. I suspected that “IOH CSI Voltage” parameter was the one, because the serial interface with point-to-point topology that is currently called QPI (QuickPath Interconnect) used to be known as CSI (Common System Interface). However, increasing this voltage setting didn’t help: the board still wouldn’t boot with the memory frequency set above 1333MHz.

Luckily, these were the only problems we discovered when we tested ASRock X58 SuperComputer in the nominal mode. If you do not choose the optimized mode but stick to the default one and then configure all parameters and technologies manually, the board works perfectly fine. The processor clock frequency multiplier and Vcore will be lowered in idle mode.

Under heavy load the multiplier will increase to 21, if you enable Turbo Boost.

We tested the mainboard in these conditions and compared the results against the performance of Gigabyte GA-EX58-Extreme also tested with default settings:

Overall, the performance of both boards in nominal mode is very close, which we have actually expected to see, taking into account similar settings. However, there are a few situations when we really can’t explain why one mainboard is faster than another.

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 ]

Discussion

Comments currently: 0

You must log in to add comments.

Forgot password? Registration

remember me