ASRock 775XFire-eSATA2

Written by Michael Larabel in Motherboards on 26 January 2006 at 01:00 PM EST. Page 5 of 11. Add A Comment.

Overclocking:

As we have brought up repeatedly in various Phoronix articles, ASRock overclocking in general is quite limited (with the exception of the previously reviewed 939Dual-SATA2) due to the lack of voltage adjustments and time spent by ASRock on ensuring the motherboard is enthusiast oriented and can achieve a nice overclock. When it came to the 775XFire-eSATA2, we had spent a fair amount of time attempting to push the motherboard far but even with high DRAM and Chipset voltages, we were unable to pass 220MHz. In fact, the highest frequency we were able to successfully boot at with our setup was a mere 215MHz. With the CPU running at 3225MHz (215MHz x 15), the system memory ran at DDR2-572 speeds with 3-4-4-11 timings as opposed to DDR2-533. Of course, the Untied Overclocking Technology implemented on this board certainly did not appear to boost any overclocking abilities.

Performance:

In addition to running the ASRock 775XFire-eSATA2 (Intel 945PL + ICH-7R) at its stock frequencies, as well as overclocked values (3225MHz - 215MHz x 15), we also ran the Abit AW8 (Intel 955X + ICH-7R), ASRock 775Dual-915GL (Intel 915GL + ICH-6), and ASRock 775Dual-880Pro (VIA PT880 Pro + 8237R) for comparison values. Of course, it is important to keep in mind that with ASRock comes innovation and some of the comparative motherboards boast various unique features as well as the VIA-powered motherboard is limited to PCI Express x4 bandwidth. For more information on these various products, their individual reviews can be found in the motherboard section at Phoronix. Throughout all testing, the hardware, and software, components were kept the same except for the instance of the 775Dual-915GL where we had to use 2 x 512MB of OCZ PC-3200 memory due to the lack of DDR2 support. Below is the configuration used during testing.

Hardware Components
Processor: Intel Pentium 4 530 (3.00GHz)
Memory: 2 x 512MB Kingmax Mars DDR2-667
Graphics Card: Gigabyte GeForce 6600GT 128MB
Hard Drives: Western Digital 160GB SATA
Optical Drives: Lite-On 16x DVD-ROM
Cooling: CoolJag JAC16EC
Power Supply: SilverStone Strider ST405 400W
Software Components
Operating System: Fedora Core 4
Linux Kernel: 2.6.14-1.656_FC4smp
GCC (GNU Compiler): 4.0.0
Graphics Driver: NVIDIA 1.0-8178
Xorg: 6.8.2

As far as the Linux compatibility goes with the 775XFire-eSATA2, after the fresh installation of Fedora Core 4 and applying the various official updates, there were no major problems. When it comes to LM_Sensors compatibility, many of the latest motherboards have experienced faults with the v2.9.2 version of the Linux monitoring sensors (as we have shown in various articles), and in the event of the ASRock 775XFire-eSATA2, sensors-detect was unable to detect the various integrated sensors. Unfortunately, there is also presently no CrossFire support under Linux to the public, so at this time we are unable to publish ATI multi-GPU results. Upon the deliverance of CrossFire support, we will test the technology on multiple motherboards under Linux, including the 775XFire-eSATA2, and will deliver that in an individual article. Onto ASRock's eSATA 2 features, we had tested the Serial ATA interface at the I/O panel using the above mentioned Western Digital 160GB SATA hard drive as well as enclosing the drive in a Data-Tec DS350 3.5" external SATA enclosure. In both situations, the SATA drive performed flawlessly under Linux with no reported signs of trouble. Due to the eSATA ports relying upon the motherboard's main SATA ports powered by the Intel ICH-7R, and simply using a data cable as the connection, we refrained from running any independent benchmarks in this setup as the ports do not rely upon any separate ASICs. In our examination section we had also mentioned there was a "SLI" marking on the motherboard's PCB near the power connection, however, when we had installed two NVIDIA GeForce 6600GT parts, connected the 4-pin molex connection, and adapted an ASUS SLI bridge, we were unsuccessful in our attempts of running SLI with the 1.0-8178 display drivers. For benchmarking these various motherboard configurations, our usual platter of Linux tests were called in, which consist of Enemy Territory, Doom 3, Quake 4, SPECViewPerf, HDparm, diskWriggler, Gzip Compression, LAME Compilation, LAME Encoding, BlueSailSoftware Opstone Sparse-Vector Scalar Product, BlueSailSoftware Opstone Singular Value Decomposition, and FreeBench. During which our usual testing methodology was enforced.


Related Articles