12/07/2010 | 06:05 PM
Besides ordinary mainboards, ASUS is known to produce a special Republic of Gamers series which features enhanced functionality. What is less known, the company launched another special series of mainboards in the last year. It goes under the name of The Ultimate Force or TUF. Besides a characteristic exterior design with military camouflage colors, this series features a special cooling system and high-quality components that make the warranty period as long as 5 years. The TUF series began with the Sabertooth 55i model based on the Intel P55 Express chipset but now it’s got an addition. As its name suggests, the new Sabertooth X58 is based on the Intel X58 Express chipset and supports LGA1366 processors. Let’s check out if it differs from its predecessor in terms of PCB design, BIOS settings, and performance in our tests.
The box with the ASUS Sabertooth X58 is colored in such a way that it seems to be metallic although it is actually plain cardboard.
Besides the mainboard, there are also the following accessories:
We’ve already mentioned the characteristic color scheme of the TUF series. Its winged emblem might do quite well for some special troops whereas the combination of black, green and brown is not unlike military camouflage.
The heatsinks look unusual. They look coarse and are indeed so when you touch them. The manufacturer says that the ceramic coating of the heatsinks, called CeraM!X, dissipates heat better than ordinary metallic heatsinks, thanks to its larger surface area.
The mainboard’s functionality is largely determined by its Intel X58 Express chipset. It supports all modern LGA1366 processors and triple-channel memory. There are six memory slots for a maximum of 24 gigabytes. The two graphics card slots work in full-speed PCI Express 2.0 x16 mode and support CrossFireX and SLI configurations. The third PCIe slot works in x4 mode. The Intel ICH10R South Bridge offers six SATA 3 Gbps ports with the option of uniting the attached hard disks into RAID 0, 1, 5 or 10 arrays.
The capabilities of the flagship, yet already rather old, chipset are complemented by a number of onboard controllers: the Marvell 88SE9128 adds two SATA 6 Gbps ports, the VIA VT6308P is responsible for IEEE1394 (FireWire), the NEC D720200F1 supports the two USB 3.0 ports at the back panel, and the JMicron JMB362 is for the mainboard’s eSATA and Power eSATA ports. By the way, we have seen Power eSATA connectors on other mainboards but they were all combined ones. That is, the power for the external SATA device was provided via USB and you could use the combo eSATA/USB port not only for eSATA but also for USB devices. The Power eSATA connector of ASUS mainboards only supports external SATA devices.
Here is a full list of the mainboard’s back-panel connectors:
We can note a few exclusive features of ASUS mainboards in general that are implemented in the Sabertooth X58: jumpers to expand the CPU and memory voltage range, a MemOK! button to solve memory-related problems that might prevent your system from starting up, and Q-LED indicators for diagnostic purposes. The following table lists the mainboard’s specifications:
The interface, structure and available options of BIOSes of ASUS mainboards are very much alike, yet there are certain differences between the ordinary and ROG series products. We might also expect to see some differences in the TUF series but the BIOS of the Sabertooth X58 doesn’t show any.
The majority of performance tuning options can be found in the Ai Tweaker menu. By default, each parameter is set up by the mainboard automatically. You can set the Ai Overclock Tuner parameter at D.O.C.P. and the mainboard will adjust the base clock rate and the CPU frequency multiplier in such a way as to increase the memory frequency to the specified level. In the X.M.P. mode the mainboard will apply the parameters written in the memory modules’ SPD if they support this technology. The CPU Level Up parameter can automatically overclock the CPU.
The numbers are all hidden behind “Auto” but the values of frequencies and voltages will be shown when you change them.
This BIOS section offers all the settings to change system frequencies, multipliers and voltages in order to set everything up in the most optimal way. You can also play with memory timings here (they are so numerous that are allotted a separate page).
The Advanced section is for managing the chipset, onboard controllers and interfaces.
The CPU Configuration subsection contains a number of important CPU-related options.
Some options can be found on a separate page titled “Intel PPM Configuration”. You should certainly go there to change the Intel C-STATE Tech parameter. It is turned off by default, which means the CPU won’t switch into deep power-saving modes when idle. Besides, it won’t be able to increase its frequency multiplier by x2 at single-threaded load. In other words, the Turbo Boost technology won’t work to its full extent and will only increase the multiplier by x1 irrespective of CPU load and power consumption.
The Onboard Devices Configuration page may be interesting, too.
The Power section contains the EuP Ready parameter which lowers the mainboard’s power consumption when the latter is turned off but not disconnected from the mains.
The Hardware Monitor section is for checking out the key voltages, temperatures and fan speeds.
Fans attached to the Chassis Fan connectors can be managed by choosing one of the three predefined modes: Standard, Silent and Turbo. For the CPU fan you can also manually specify the correlation between speed and temperature.
The Boot section is for specifying the boot device order and some other parameters related to the boot-up process. The Tools section contains a few special utilities.
EZ Flash 2 is a user-friendly and informative tool for updating the mainboard’s BIOS.
OC Profile is for saving and managing a few profiles with BIOS settings.
AI NET 2 is a tool that helps diagnose a network cable whereas Drive Xpert Configuration is for managing hard disks connected to the SATA 6 Gbps ports.
The next table summarizes the mainboard’s BIOS capabilities:
All performance tests were run on the following test platform:
We used Microsoft Windows 7 Ultimate 64 bit (Microsoft Windows, Version 6.1, Build 7600) operating system, Intel Chipset Software Installation Utility version 188.8.131.525, ATI Catalyst 10.9 graphics card driver.
We had no problems running our Sabertooth X58 in its default mode. We easily assembled our testbed around it and installed the OS. The CPU would lower its frequency and voltage in idle mode.
To make the most of the power-saving modes and Intel Turbo Boost technology we enabled the Intel C-STATE Tech option in the mainboard’s BIOS. As a result, the CPU multiplier was increased to x23 at single-threaded load and to x22 at higher loads.
It is when we tried to overclock our CPU that we had a problem. The mainboard would start up and pass the POST normally, but then it would restart every time it tried to boot the OS up. We thought that the CPU voltage might be set up incorrectly in the Offset mode (when it is not fixed at a certain level, like in the Manual mode, but is added to the default level, which keeps Intel’s power-saving technologies up and running).
But in fact, we didn’t even have to overclock our CPU to experience the described problem. If we but increased the CPU voltage by a mere 0.00625 volts and left the rest of the parameters at their defaults, the mainboard couldn’t boot the OS up anymore! As it turned out, the mainboard increased the voltage in the Offset mode correctly and the root of the problem was in the Load-Line Calibration option. If you turn it off, the mainboard easily boots the OS up even when the CPU voltage is increased. If this option is turned on or set at Auto, the mainboard will be restarting all the time. We only wonder why the Load-Line Calibration option is enabled when all the parameters are at their defaults except that the CPU voltage is increased by just a bit.
By the way, we guess we should also mention a couple of typical downsides you can find in the BIOSes of ASUS mainboards. First, they do not show you the default CPU voltage explicitly. They can correctly set it in the Auto mode, but you can only see it in some monitoring tool. Second, it is impossible to fix the CPU voltage at its default level. Well, we can do that formally if we enter the required value in the Manual mode, but then this voltage will always be the same irrespective of load and will not be decreased in idle mode. That is, this would effectively turn off all the power-saving technologies implemented in Intel CPUs.
We can alternatively set the CPU voltage up in the Offset mode, but it will be automatically increased by the mainboard if you select the Auto value. Therefore we have to increase the voltage in very small steps, by 0.00625 volts in our case, to keep it as close to the default one as possible. Alas, even this small increase in voltage provoked the problem we’ve described above.
Load-Line Calibration is indeed a handy and useful, but not obligatory option for overclocking. Trying to maintain the same level of voltage on the CPU when the latter is doing some heavy computations, and often exceeding the default voltage level at that, this technology is meant to prevent the CPU voltage from getting dangerously high. When the CPU is idle and enables its power-saving technologies, its voltage is lowered, yet remains high enough even for overclocking because there is no CPU load and the CPU frequency is reduced by those power-saving features. So, we only need an increased voltage when the CPU is working hard, and that’s when Load-Line Calibration comes into play, ensuring stability of the overclocked system. Thanks to that technology we managed to overclock our CPU up to 3.9 GHz on other mainboards without formally increasing its voltage: the voltage was increased by Load-Line Calibration when that was necessary, i.e. at high loads.
Thus, it is clear that we can do without Load-Line Calibration if we increase the CPU voltage manually. As we found out, our CPU could work at 3.9 GHz (with a base clock rate of 177 MHz) when we added 0.075 volts to the default voltage. We increased that parameter in the Offset mode, keeping Intel’s power-saving technologies up and running so that they reduced the voltage and frequency of the CPU when the latter was idle.
We didn’t enjoy our success for long because we had another problem right away. The mainboard could not make the memory modules stable at 1770 MHz with 8-8-8-22-1T timings (we had used these parameters before in our reviews of the Gigabyte GA-X58A-UD5 rev. 2.0 and Gigabyte GA-X58A-UD3R rev. 2.0). Therefore we had to limit ourselves to a memory frequency of 1416 MHz. The final overclocking results are shown in the next screenshot:
We wouldn’t say that the inability of our mainboard to make the memory modules stable at high frequencies was a catastrophe. We made up for the lower frequency by selecting more aggressive timings, namely 7-7-7-20-1T. Therefore we hope the difference from Gigabyte’s mainboards with their memory frequency of 1770 MHz and timings of 8-8-8-22-1T won’t be big. By the way, there is one positive thing to the ASUS mainboard’s failure. We had had to increase the voltage of the CPU-integrated memory controller with the Gigabyte mainboards to achieve stability. And the increased voltage has a negative effect on power consumption, of course. With the Sabertooth X58 we didn’t have to increase the memory controller voltage at all, leaving it at its default 1.2 volts. So, we can expect the ASUS mainboard to be comparable to the Gigabyte ones in terms of performance but better in terms of power consumption. Let’s check this out right now.
As usual, we are going to compare the mainboards speeds in two different modes: in nominal mode and during CPU and memory overclocking. The first mode is interesting because it shows how well the mainboards work with their default settings. It is a known fact that most users do not fine-tune their systems, they simply choose the optimal BIOS settings and do nothing else. That is why we run a round of tests almost without interfering in any way with the default mainboard settings. In this case, however, we did enable “Intel Turbo Boost”. For comparison purposes we are going to use the results of Gigabyte GA-X58A-UD5 (rev. 2.0) mainboard and Gigabyte GA-X58A-UD3R (rev.2.0) tested earlier. The results for Asus Sabertooth X58 are marked with darker color on the diagrams.
We used Cinebench 11.5. All tests were run five times and the average result of the five runs was taken for the performance charts.
We have been using Fritz Chess Benchmark utility for a long time already and it proved very illustrative. It generated repeated results, the performance in it is scales perfectly depending on the number of involved computational threads.
A small video in x264 HD Benchmark 3.0 is encoded in two passes and then the entire process is repeated four times. The average results of the second pass are displayed on the following diagram:
In the archiving test a 1 GB file is compressed using LZMA2 algorithms, while other compression settings remain at defaults.
Like in the data compression test, the faster 16 million of Pi digits are calculated, the better. This is the only benchmark where the number of processor cores doesn’t really matter, because it creates single-threaded load.
There are good and bad things about complex performance tests. However, 3D Mark Vantage has become extremely popular. The diagram below shows the results after three test runs:
Since we do not overclock graphics in our mainboard reviews, the next diagram shows only CPU test from the 3D Mark Vantage suite.
We use FC2 Benchmark Tool to go over Ranch Small map ten times in 1280x1024 resolution with medium and high image quality settings in DirectX 10.
Resident Evil 5 game also has a built-in performance test. Its peculiarity is that it can really take advantage of multi-core processor architecture. The tests were run in DirectX 10 in 1280x1024 resolution with medium image quality settings. The average of five test runs was taken for further analysis:
It is a well-known fact that similar systems working at similar settings will deliver similar performance. This time around the difference between the mainboards is small indeed but the Sabertooth X58 is suspiciously slower than its Gigabyte opponents in any test. However, we must recall here that the Gigabyte mainboards set their base clock rate a little higher by default (at almost 135 instead of 133 MHz), also raising the rest of the related system frequencies, whereas the Sabertooth X58 doesn’t do that. The difference in frequency is about 1%, and the mainboards differ in performance by about 1%, too (the difference in performance is smaller in those applications that depend not only on the CPU or system memory but also on the graphics card). So, we shouldn’t worry about that. The ASUS Sabertooth X58 delivers good performance in its default mode.
Now let’s overclock the systems by increasing the base clock rate and benchmark them once again.
Here, the last place of the ASUS Sabertooth X58 might have been expected because it could not reach as high a memory frequency as the Gigabyte mainboards. When the speed of the test application depends on the CPU frequency or the graphics card, the difference is below 1% or even smaller. But if the test application depends on the memory subsystem, we can see a much bigger difference like over 4% in Resident Evil 5 Benchmark or 7.5% in the 7-Zip test. This difference can be noticed at everyday work. So while we are far from overemphasizing the importance of high memory frequencies and low memory timings (you can find a number of articles on this topic on our site), optimal memory parameters do have an effect on any computer’s performance.
We performed our power consumption measurements using an Extech Power Analyzer 380803. This device is connected before the PSU and measures the power draw of the entire system (without the monitor), including the power loss that occurs in the PSU itself. In the idle mode we start the system up and wait until it stops accessing the hard disk. Then we use LinX to load the Intel Core i7-930 CPU. For a more illustrative picture there are graphs that show how the computer’s power consumption grows up depending on the number of active execution threads in LinX (both at the default and overclocked system settings). We performed the test in four modes: idle mode, single-thread load, four-thread and eight-thread load. The mainboards are sorted in alphabetical order on the diagrams.
The mainboards consume about the same amount of power at their default settings irrespective of load. However, the Sabertooth X58 is far more economical when overclocked, which is due to the default voltage on the CPU-integrated memory controller (we had to increase that voltage above the default level on the Gigabyte mainboards to make the memory modules stable at high frequencies). We should keep it in mind that the lower power consumption is achieved at the expense of performance.
Quality and reliability are crucially important and fundamental characteristics of any mainboard as they determine the service life and stability of the whole computer. This is why MSI emphasizes the use of “military grade” components in its products while Gigabyte’s technologies for higher reliability which go under the name of Ultra Durable are a third version already. It is no wonder then that ASUS offers its own high-reliability mainboard series which is called TUF or The Ultimate Force. There is one important difference of the TUF series from its competitors, though. ASUS not only promotes its products as highly reliable and backs this claim up with test results (there is a certificate with the description of the test methods included into the product box). The most important thing is that ASUS prolongs the warranty period for its TUF products for as many as five years!
According to our tests, the ASUS Sabertooth X58 is no different from ordinary mainboards ASUS offers except for the characteristic color scheme. Its functionality is the same or even somewhat broader than usual. Its packaging and accessories are the same as those of other ASUS mainboards, too. It makes full use of the chipset’s capabilities, supports multi-GPU configurations out of two graphics cards, and has a few additional controllers that implement SATA 6 Gbps, USB 3.0 and IEEE1394 (FireWire). Added to that are the high-quality components, the special heatsink coating CeraM!X, and the 5-year warranty instead of the ordinary 3-year one! We wouldn’t be surprised if such a mainboard cost half as much as an ordinary one, but in fact the Sabertooth X58 is only about $200, which is very modest for an LGA1366 product. This is even less than the price of an entry-level LGA1366 mainboard from ASUS such as the ASUS P6T SE. We just can’t understand why the Sabertooth X58 is so cheap. The only thing that’s missing in it is Express Gate (an integrated Linux-based OS with basic functionality for work and entertainment), but this can hardly make a big difference.
Some users may be disappointed with the problems we had when we tried to overclock our CPU and memory on this mainboard, but they are not too serious and the mainboard can be used for overclocking, too. Besides, those problems may be corrected in future BIOS updates. After all, there are a lot of mainboards from ASUS and other makers that can be overclocked, but the combination of a 5-year warranty, enhanced functionality and low price is a unique offer indeed. In fact, you buy an ASUS Sabertooth X58 for the whole lifetime of your computer. If the mainboard fails in the next 5 years, ASUS will replace it. And if it fails after 6 or 8 years, for example, there will be no reason to repair it. It would be much more reasonable to replace the whole computer then and buy a new Sabertooth. We do hope we will see more TUF series products with the same attractive combination of high functionality, low price and long service life!