ASUS F2A85-V Pro Review: A Look at FM2 with A85Xby Ian Cutress on October 10, 2012 11:20 AM EST
ASUS F2A85-V Pro In The Box
What we get in the box usually is the cream of the package - something that little bit extra that makes the whole product a more enjoyable experience. Despite this, the price of the board usually indicates what level of goodies we get in the box, even if we have had a few surprises in the past. With this product selling for $140 at launch, more expensive than the high end Trinity processors, we should feel that there's something in the box to make it worth while. What we actually get in the box is:
Four SATA Cables
The box could be a little more filled with joy - either a USB 3.0 panel or something similar. As mentioned before, I would have liked to see the ASUS WiFi/Bluetooth module on the 'Pro' board, which would have included antenna and the module in the box.
Unfortunately we are unable to bring you the results of our OCCT test, due to a level of incompatibility between OCCT and current FM2 boards we have discovered that was not correctly reading the voltage. This test in the future will require an OCCT update.
ASUS F2A85-V Pro Overclocking
Experience with ASUS F2A85-V Pro
Overclocking on a new platform is always a little frustrating – even if the principles of overclocking have not changed, or the architecture has not changed, a manufacturer may slightly confuse with different names for voltages, and there is no experience guiding how these processors may interact under voltage.
As such we have to take a methodical view to overclocking. For almost all 24/7 CPU overclocks all we ever need to adjust is the CPU multiplier and the CPU voltage, so starting with the load multiplier and voltage that comes with the processors is never a bad thing. In the case of our test bed we had an A10-5800K which has a maximum turbo multiplier of 42x and a load voltage of around 1.40 volts according to OCCT. Using this information I performed overclock testing starting at an underclock of 35x and attempted to find the minimum voltage needed to be set in the BIOS to make this stable. The system was then raised slowly with the multiplier, each time finding the minimum voltage required to be stable.
The third option that an overclocker may play with is Load Line Calibration. This adjusts the voltage drop across the processor when under load (as causing the processor to do work causes a droop in the voltage reading) – a low LLC uses less energy overall in the system and is often dictated in part by the processor manufacturer. However a high LLC often has the benefit of making an overclock stable. In the case of ASUS motherboards we get a variety of options for LLC, but for the purposes of testing here it was left on automatic.
The experience of overclocking on the F2A85-V Pro was fairly standard for a top tier motherboard – we get two automatic overclock options in the OS in the form of ‘Fast’ and ‘Extreme’, as well as one in the BIOS and the TPU switch on board which both perform the ‘Fast’ overclock. Manual overclock involves either playing with the AI Suite software until the system is unstable, then making permanent adjustments in the BIOS as required. Without comparing against other motherboards yet I cannot say how well this board performs relative to others, but having the temperature reading issue does not help much.
Our standard overclocking methodology is as follows. We select the automatic overclock options and test for stability with PovRay and OCCT to simulate high-end workloads. These stability tests aim to catch any immediate causes for memory or CPU errors.
For manual overclocks, based on the information gathered from previous testing, starts off at a nominal voltage and CPU multiplier, and the multiplier is increased until the stability tests are failed. The CPU voltage is increased gradually until the stability tests are passed, and the process repeated until the motherboard reduces the multiplier automatically (due to safety protocol) or the CPU temperature reaches a stupidly high level (100ºC+).
Our test bed is not in a case, which should push overclocks higher with fresher (cooler) air. We are using a beQuiet Dark Pro CPU cooler with its stock fan. This is a high-end air cooler, designed to tackle up to 150W of CPU power without issue.
Using the AI Suite software, we navigated to the TurboV Evo Automatic Tuning menu. It offers two options – ‘Fast’ and ‘Extreme’. Here are our results with these options.
With the ‘Fast’ option, the system rebooted indicating the CPU had been boosted to 43x and 100 MHz (4300 MHz total), with the IGP also boosted to 950 MHz. This overclock passed both OCCT and PovRay.
With the ‘Extreme’ option the system rebooted and initialized stress testing with the processor. The software started adjusting the CPU multiplier in the OS, rebooted then adjusted the APU frequency. When all was said and done, the final result was a CPU overclock to 44x and 100 MHz (4400 MHz total), with the IGP also boosted to 1013 MHz. This overclock passed PovRay but led to CPU errors in OCCT.
With the manual overclock we left LLC on automatic, started at a CPU voltage of 1.1 volts and multiplier of 35x. On a failed boot or unstable system, the voltage was raised by +0.025 and retested. If a settings passed both PovRay and OCCT then the multiplier was raised. To show the tests going into this, here is a direct screenshot from my results file:
The best way to represent these results is with the following graph:
At 4.5 GHz I was unwilling to go much further without any clear indication of the temperature of the processor. Every setting would give a max reading of ~62C. Judging by the results of overclockers online, these processors on air could potentially go up to 5.1 GHz with the correct settings or a better processor – I have seen 4.8 GHz on 1.50 volts stable enough to run simple benchmarks. Overclocking a processor is like opening a packet of chocolate chip cookies – some cookies have a lot of chocolate chips and some have none. You hope the cookie you get is full of chocolate chips. In this case, I may have one without any. If I took this result in isolation, I would say that AMD are really pushing these chips to the limit on clock speed – getting 300 MHz more than stock is not representative of recent processor releases.