ASUS P8Z68-V Pro Z68 Chipset Motherboard Review

When Intel brought out the P67 and H67 chipsets, many enthusiasts complained that the features of one or the other were only available on one or the other, but not both. Intel has heard the community and is therefore launching the Z68 chipset. As usual ASUS is among the first board makers to offer a board featuring this latest chipset. The ASUS P8Z68-V Pro promises to be an interesting first look at this new chipset.

continued...

Subsystem Testing

Intel Test Setup (LGA1155)

NOTE: For all Subsystem Testing, an Intel Core i7 2500K (3.30GHz) and 2 x 2GB Corsair Vengeance (1600MHz DDR3 9-9-9-24-1T @1.50v) memory modules running at DDR3 1600MHz were used. The CPU was cooled with a home brew water cooling setup consisting of a Swiftech MCR320 triple radiator, 3x120mm fans, Swiftech MCP655-B pump, and a Swiftech Apogee GTZ water block modified to work with LGA1156 / LGA1155 boards and CPUs.

Sound Hardware

Like ASUS’ P8P67 family of motherboards, the P8Z68-V Pro uses an integrated Realtek audio solution. Specifically this one uses the Realtek ALC892. ASUS is a bit all over the place on this as many of their boards use either the ALC889 or the ALC892. The specifications listed are as follows:

Realtekآ® ALC892 8-Channel High Definition Audio CODEC

- Absolute Pitch 192khz/24bit True BD Lossless Sound

- BD Audio Layer Content Protection

- DTS Surround Sensation UltraPC

- Supports Jack-Detection, Multi-streaming and Front Panel Jack-Retasking

- Optical S/PDIF out ports at back I/O

Sound quality is almost entirely subjective, so comparisons are as subjective. I found the audio playback of the ALC892 to be passable. It was unremarkable and sounded exactly like the other boards I’ve reviewed that used the same type of solution. So in essence, everything is working precisely as it should be.

Audio آ– Subjective Listening

For subjective listening you want to listen to something that covers a range of sound types. For this portion of the review I went with Disturbed, Indestructible.

CD audio worked as I had expected.

Audio آ– Microphone Port Testing

The onboard audio MIC-IN port was tested using a Logitech Internet Chat Headset. Spoken words were recorded from the Windows Sound Recorder found under the Accessories\\Entertainment folder in the start menu within Windows 7. The recording was made with the Microphone Boost option disabled, then enabled. The Microphone Boost option is found within the advanced menu under the microphone section with in the Volume Control Menu.

The recording sample was audible with the microphone boost option both enabled and disabled, with more distortion being present when the option was disabled. The distortion was in line with other solutions of this type. So I can’t really complain too much.

Drive Performance

To test the capabilities of the on board USB 2.0 connections, we used an ACOMDATA HD060U2FE-72-USB 2.0/FireWire HDD connected first to the USB port. USB 3.0 functionality was tested using a Super Talent USB 3.0 SuperCrypt 32GB Flash Drive. SATA drive tests were performed using Western Digital Caviar Black WD1002FAEX hard drives on all SATA headers. The SATA drives were used for testing in RAID 0 16k block size configurations on all applicable controllers. Testing was also conducted using the same model SATA drives in a stand-alone SATA configuration on all applicable controllers. All drive benchmarks were done using the freely available CrystalDiskMark 3.0 program, run with both 50MB and 100MB sized test sets.

50MB Test Set

Article Image

Article Image

Intel’s chipsets are often the bar by which other chipset features are judged and for good reason. On average these tend to perform better than most other offerings in its respective classes. In the 50MB tests we see excellent USB 2.0 and 3.0 performances, and unsurprisingly the SATA 6G controller tends to dominate in read and write tests when compared to the Marvell controller or even the same controller in 3G mode. We couldn’t test the boards IEEE1394 performance due to a lack of an actual port to connect to. The I/O bracket is unfortunately sold separately on the P8Z68-V Pro.

100MB Test Set

Article Image

Article Image

In the 100MB sequential read tests we see the Marvell controller even the score up quite a bit. In the sequential write tests, that gap gets even narrower. Again we see solid USB 2.0 and 3.0 performance. It is still not surprising that the Intel Z68 Express chipset dominates in both stand alone and RAID modes fairly easily much of the time. That’s pretty much par for the course.

Intel Smart Response Technology

OK, so one of the biggest features in the Z68 chipset over P67 or H67 is the inclusion of the Smart Response Technology feature. As I said in the introduction, you can think of it as Ready Boost or SuperFetch in hardware. Initially I began testing this feature according to guidelines provided by Intel in tons and tons of documents. It turned into a really complex subject matter. So complex in fact that we will be covering the technical aspects of this technology in greater detail in a separate article. However, I want to go ahead and address the broad strokes of the technology a bit here and explain why we are approaching the subject in the fashion we are at this time.

Through this feature, the idea is that you can buy a small SSD which won’t cost you very much and give your larger drive volumes SSD like performance. This can work on single drives, and RAID arrays as well. The feature itself does not use actual file transfers but rather block level transfers of data. The caching algorithm puts frequently used blocks of data on the SSD to provide faster access to the data than a conventional hard drive. It uses the principal of data locality to keep a persistent working set of frequently accessed data on the SSD and retained non-volatile across system shut downs and reboots. Generally for the data to be cached it has to have been accessed previously. In fact Intel cautions that benchmarks must be run multiple times in order for the caching algorithm to actually work. So you can think of this feature as learning what it needs to cache as it is used.

The user doesn't actually see the SSD itself once setup. It is basically hidden and becomes part of your drive array. Even with single drive volumes, it’s basically treated as a RAID array from a software standpoint. SRT operates in one of two modes, write back caching, and write through caching. In the Intel Rapid Storage Technology software the SRT mode has two operational modes. Enhanced and Maximized performance. These are write-thru, and write back modes respectively. They each have their advantages and disadvantages. Essentially the write-thru mode carries no additional risk of data loss after a power failure because all the data is written to both the host drive and the SSD. With the write back or "maximized" performance mode the risk of data loss or corruption after a power loss is very high. Additionally If the system is running with the drive in "maximized mode" and you decide to move the OS drive (or the chosen cached volume) to another computer in which supports the feature, than the caching volume must be moved with it. You can however disable the feature, reset the SSD drive to "available" status and then move them to another machine if you so choose.

While no special firmware is required, Intel is releasing a special SSD designed to make the most of this feature. Intel’s SSD 311 series. These drives formerly code named "Larson Creek" are interesting. They feature a 20GB capacity which is of course too small to really be all that useful by itself in a modern machine. As a caching device however, it can make a larger impact on system performance. This drive features 34nm NAND flash memory. This is SLC memory as well. According to Intel SLC memory was chosen because this memory type is well suited to the caching role and is optimal for longevity. That is not to say that MLC memory based drives can’t be used, they certainly can. Though Intel feels this is a better solution.

Traditionally SLC based drives were ridiculously expensive. However given the size and manufacturing technology the 311 series is based on, the price is more reasonable than you might think. According to Intel MSRP on the SSD 311 series is only going to be around $110 or so.

Without being anymore long-winded than I already have been we ran into some issues with this feature. Intel was still working on the drivers for the chipset and specifically this feature, and ASUS had to give us a new BIOS to try and address some of these issues. In a nutshell, like Superfetch, the system has to take time to learn what applications or data blocks are used most frequently and then cache them. Also the algorithm requires that any data to be cached must have been already used once before it can benefit from caching. Makes sense right? Well it does. The problem is that in the context or rather the time frame of a board review, one does not really get to use the system in a normal everyday type of usage scenario, nor do they get enough time to allow the caching feature to work as intended. Typically we also run drive benchmarks on empty or nearly empty drives. Well the way this technology works that’s not terribly helpful.

In short this is one of those features that isn’t necessarily the easiest to quantify. It’s the type of thing which real world usage will most likely be the most beneficial. In some of our testing we did indeed see some of what Intel was talking about in terms of performance increases. 4K writes and 4K QD32 writes improved dramatically. However we saw dramatic reductions in sequential read and write performance under most circumstances. We looked at the results through ATTO, HD Tune Pro, and of course Crystal Disk Mark. The latter actually showed us the most improvement, or change, but again only under certain circumstances. You also have to run the utility multiple times to see the effect of the caching. The first run usually shows nothing but a drop in performance, and then the results steadily improve with each run.

So stay tuned. We will be covering this feature in greater depth today. I think it’s a cool idea, but sadly not something we can easily benchmark and report on.

Network Utilization Tests

Hagel Technologies’ DU Meter software was used with Windows Task Manager to determine the performance levels of the onboard network interface. DU Meter was used to measure bandwidth and transfer speeds, while Windows Task Manager monitored CPU utilization on the test system. For the testing, a 750MB Archive file consisting of several compressed WMA/MP3 files was used for the large file transfer, and 750MB worth of MP3/WMA files were used ranging in sizes from 3 to 30MB was used for the small files transfer test. The test was performed using a plenum rated category 5e crossover cable to bypass any traffic, routing or other transfer issues and possible packet loss or corruption that can be caused by a router/switch or hub. The cables were connected between two test machines, one using the onboard NIC(s) of the board being reviewed and the other is an Intel EXPI9400PT 10/100/1000Mbps PCI-Express Gigabit Ethernet adapter installed into a test machine using an Intel D5400XS motherboard.

By popular demand, and therefore good business sense, ASUS has embraced Intel’s integrated network solutions. Specifically the P8Z68-V Pro is equipped with the Intel 82579 Gigabit Ethernet controller. It is capable of 10/100/1000mbit Ethernet speeds.

LAN1

Article Image Article Image Article Image Article Image

The small files download test showed an average transfer rate of 90.6MB/s and maxed out at 115MB/s. CPU usage in this test was a modest 3%. The small files upload test had a similar peak transfer rate of 114MB/s but the average fell to 49.4MB/s. CPU usage was almost non-existent at 1%.

Article Image Article Image Article Image Article Image

The large files download test had a maximum transfer rate of 110MB/s and averaged a very nice 97.5MB/s. Again CPU usage hit only 3% during the transfer. The large files upload test gave us similar results with the average hitting 92.4MB/s and the throughput actually reaching 118MB/s. CPU usage was negligible at only 2%.

Test Systems

The following system configurations were used for the Sandra memory benchmark graph, as well as all graphs listed under the Application and Gaming Benchmarks sections:

Article Image

Graphs are labeled as follows: Motherboard - CPU Clock - Memory Clock

Note

We has some issues with the processor running at slightly higher clocks than it should have. Which in turn skews some of the test results in favor of the P8Z68-V Pro. We have consulted with both ASUS and Intel on this but have not come away with any firm conclusions as how to fix the issue. What it comes down to is that the CPU clock is either running at its idle state of 1.6GHz, or it's highest Turbo value of 3.8GHz. There is no scaling between the two clock values. We saw this in the last P67 ASUS motherboard we tested as well. If you leave the BIOS to full defaults, even with Windows 7 running in performance mode, you will see proper Turbo scaling, which we have never seen reach 3.8GHz under any kind of multi-threaded mode. Any changes to the BIOS, such as scaling only the DDR3 clock to 1600MHz (as we test all motherboards) gives us the "3.8GHz" problem. Now this causes two issues. First, and most importantly it skews the benchmark scores compared to the other boards shown that do use the Turbo feature "properly." Second, the system represented is not truly performing as it should in regards to the Turbo feature. The fact of the matter is that you probably should not care seeing as how most of our readers will be overclocking anyway, but it is something to be aware of. The better performing scores see hereafter are not indicative of a "faster" motherboard, but rather this Turbo scaling issue. It came down to use either running default clock RAM, which we never do or showing the board with faster CPU compared to the others. We chose to run it with the faster clock.

SiSoft Sandra 2011

Article Image

Note that all results above were obtained running the installed memory in Dual-Channel mode except for our X58 platform which is running in Triple-Channel mode.

The Sandra memory scores were no surprise here. The Z68 chipset is virtually identical to the P67 in regard to memory performance.

Article Image

Given the issues with Turbo scaling on the P8Z68-V Pro these results are skewed slightly higher than they should be. However the value in showing something like this is in showing that the product works as intended, well aside from the processor not clock scaling like it should.

Hiper Pi

Article Image

Evidently the slight clock speed advantage didn’t have that much of an impact in this test.

wPrime

Article Image

Once again the clock speed advantage seems very clear in this test.