 |
 |
 |
||
|
|
||||
 |
|
|
| Introduction |
If the recent release of Intel's X58 chipset has taught us anything, it's that change can be good. In our first two forays with these new boards and Intel's Core i7 processor, we've been firmly impressed by the performance gains seen over legacy Core 2 architectures, much less AMD's recent advancements and new offerings. Now, with some time under our belts, we're going to take a look at two new Intel X58 Express chipset based boards from ASUS and MSI that we've put through the test ringer. Focusing on the upper-tier of the PC market, both come fully stocked with additional components, fancy board designs, and all of the other benefits that the X58 provides, with higher-end prices to match.
First, we'll start off with an overview of the Intel X58 Express chipset. As shown in the block diagram below, the major changes revolve around the CPU's system and memory interfaces. Considering Intel's relatively slower evolution over their past few chipset updates, it's almost as if they sat back, looked at what needed changing, and then just went ahead and did them all in one fell swoop.
The first and most obvious change is the new LGA1366 socket, which nearly doubles the number of pins in order to support the Core i7. Why all the extra connections? Well, one of the biggest changes to the CPU design was to move the memory controller from the North Bridge to the CPU itself, which now supports Triple Channel DDR3. Also to supplement this new memory controller with more bandwidth, Intel created a new way to connect the CPU and the rest of the board. Dubbed QPI (Quick Path Interconnect), it consists of a serial point-to-point interconnect that offers up to 25.6GB/s of bandwidth - more than double that of the previous X48 chipset. Another big change brought about is the native support of not only ATI's CrossFireX, but 2 or 3-way SLI for NVIDIA's graphics cards as well, provided the motherboard has the necessary BIOS hooks. As the X58 is currently the only chipset that supports the Core i7, it means you won't have to debate where to go to support not only the top processors, but the top graphics solution as well (depending on which side is currently residing on top). The South Bridge sees very little change, actually, although the ICH10/R originates from a previous coupling to the mainstream P45 chipset rather than the ICH9/R found on the X48s. Both of today's entries feature all of the above, and then throw in their own special accoutrements to make them stand out on their own. For our testing purposes, we'll pit these two boards against each other in head-to-head combat spanning synthetic benchmarks, some real-world rendering and encoding, and then finally the all-important gaming benchmarks including some SLI frame rates. |
| ASUS Rampage II Extreme - Specifications | ||||||||
Although the box seems slim in comparison to the MSI Eclipse, there was still plenty to be found within. Three sets of SATA cables, each containing one straight end and another that's angled, plus a single SATA power splitter cable and flat ribbon IDE cable cover all of the standard connections. A bracket with two USB and IEEE-1394 ports adds additional flexibility. We also found a collection of items that ASUS has been providing for some time now, including active cooling fans for the MOSFET heatsinks, Q-Connector pins for easier front panel installation, and the LCD Poster.
There are also a handful of new or updated additions that we noted. As the X58 supports not only CrossFire but SLI as well, a standard SLI cable as well as a 3-Way SLI bridge are included for those not operating on a tight budget. The ProbeIt cables as well as the smallish red and black wires pictured above, can be used along with a multitester to more accurately gauge voltages and thermal sensor cables can be attached to certain spots on the board to get a better readout of on-board temps. Finally, an EL I/O shield lights up the back plate. With the clearly labeled ports, this eases connection of devices, especially in dimly lit areas. The final piece is the SupremeFX X-Fi riser card, updated with Creative's X-Fi technology. Although X-Fi is featured prominently, it should be noted that the audio still comes from an ADI CODEC, similar to that used on past boards. It does not use a discrete Creative chip, as does the MSI Eclipse. As such, the SoundMax ADI Audio driver gets installed, but used along with Creative's applications. |
| ASUS Rampage II Extreme - Layout and Features | ||||
Along with the package, the layout of the Rampage gets a complete facelift. The cooling system is more and more designed as part of the overall look and feel of the board rather than an afterthought to getting the job done. Probably as a result of trying to fit everything in, the Rampage II Extreme is an inch larger than the standard ATX form factor. That extra inch could cause an issue with some cases, and in fact made our installation into an Antec Nine Hundred a little bit of a challenge.
Black and red heatsinks sit directly over the Southbridge with an oversized footprint, yet it purposely was designed as low profile as can possibly be leaving ample clearance for the graphics cards that will pass over them. Cooling for the Northbridge is found directly adjacent to this and has a passive gunmetal-colored heatsink placed on top with a lit-up RoG logo on it. A single heatpipe leads away from here to two sets of heatsinks placed on top of the MOSFETs around the CPU socket area. As these heatsinks overhang the shielded chokes, they also pull a little heat away from them as well. Overall, the board has a really classy look that even provides a signature area so you can put your name down for bragging rights.
Although red and black seems to be the dominant color scheme of the Rampage II Extreme, a few components get treated in blue, including two of the PCI-E X16 slots and one set of DIMM channels (the other channel is colored in boring beige). IDE and SATA ports get pushed down into one corner with both sets of ports angled forward, again to prevent connected cables from obstructing longer video cards. In theory, it's a great idea, but in practice getting cables installed after the motherboard has been set, especially in tight quarters inside of the chassis, can be a bit of a pain.
TweakIt and ProbeIt create a whole new level of interaction with the motherboard. Start and reset buttons are larger than on past models, and are easier to locate with three new buttons making an appearance for the first time. These buttons have a great tactile feel to them, letting you know when you've pressed something. The new buttons are the 'toggle', which lets you cycle through system options, 'confirm', used to enter an option or to confirm your choice, and 'select' - a mini joystick used to decrease or increase values. Used in conjunction with the LCD Poster, a builder can view voltages, temperatures, and fan speeds or change said voltages and CPU frequencies all without entering the BIOS. ProbeIt consists of a series on contact points that can be connected to a multitester to get more accurate voltage measurements, especially helpful while overclocking.
LEDs are placed around the board to give another voltage readout based in normal, high, or crazy ranges (as set in the BIOS). CPU is up high, nearest to the 8-pin power connector, with North Bridge, South Bridge, and memory LEDs all nearest their associated component. Hard Drive and power LEDs also keep the user informed as to their current state of activity, but are mostly helpful only when the chassis does not offer them. Seven 3-pin fan headers combined with three temperature readout connectors really allow the user to monitor temps and attach fans, or increase their output, as necessary.
A floppy port is placed far down the board, near the expansion slots while the TweakIt/ProbeIt controls are nearest the drive cages. This switch-up is not as favorable in our eyes, as the area around the drives is typically cluttered with various cables while the floppy drive cable will have to stretch the length of the board--if it's used at all that is. A single SATA port for external SATA HD is placed in the corner of the board, controlled by a JMicron JMB363 chip. Three USB headers provide for an additional six USB ports, bringing the total number of USB devices natively supported to 12. Rear I/O consists of a PS/2 port for a keyboard, six USB 2.0 ports, 2 RJ-45 LAN jacks, IEEE-1394a, eSATA, and a Clear CMOS button, which can be disabled using jumpers to prevent accidental usage, but got a thorough workout in our testing. The SupremeFX X-Fi riser card consists of six audio jacks, as well as coaxial and optical S/PDIF audio out. Although a bit hard to make out in the last photo, the backplate lights up make things easier to see at night, although it's not so bright as to make it obtrusive.
|
| MSI Eclipse - Specifications | ||||||||
MSI's latest board utilizes a new naming scheme, somewhat reminiscint of ASUS' Republic of Gamers boards in that the name itself seems completely isolated from anything computer-centric, in this case, the MSI Eclipse. The box is thick and heavy, with a stylized solar eclipse on the cover. Opening the box flap gave us a quick rundown of the some of the main selling points of the Eclipse, including their GreenPower technology which uses a more polished version of the GreenPower Genie, SoundBlaster X-Fi Xtreme Audio, and Hi-C CAP - highly conductiove polymerized capacitors, which can reportedly handle higher over-voltages, and thus, offer greater overclocking possibilities.
MSI packs the Eclipse with a complete assortment of not only manuals and driver CDs, but cables and assorted paraphenelia as well. Separate discs are used for Windows XP and Vista owners, with proprietary drivers for the X-Fi Xtreme audio. Manuals include a quick installation quide, a full user's guide with all the little details, and HDDBackup and X-Fi Xtreme setup guides as well. SATA, IDE, and floppy cables, addtional brackets with USB, IEE-1394 and eSATA ports - you'll find it all in this high-end entry. And, like the "green-powered" P45 Platinum and Diamond boards that we reviewed last year, the Eclipse also supports the GreenPower Genie, which helps conserve energy by shutting down unused components.
One new addition is the D-LED2, a small daughterboard with display which replaces the cryptic LED readouts of the past. Installing the D-LED2 seems to be a no-brainer, until you run into a cautionary note in the manual that states that improper placement could damage the unit and/or the board itself. However, the unit is large when compared to the pins, and making sure you're plugging them in correctly becomes a hassle, especially after the board has been installed into the chassis. We ran into various issues with the board not starting, or the D-LED2 not being active when trying to install it after the fact (luckily, we did not run into any damaging situations). On a subsequent build, we made sure to connect this item first before placing it into the chassis, but even then found that temperatures were not being reported correctly. Any way you look at it, we'd appreciate if MSI took another route in the future. We love the idea of concise troubleshooting readouts and informational displays, but without the hassle that the D-LED2 created. |
| MSI Eclipse - Layout and Features | ||||
Although the X58 brings with it some new advancements that necessitate more board real estate, the Eclipse itself does not seem to be overly crowded in any one area, nor do we find any obvious spacing issues. Due to changes within the CPU, the socket area is slightly larger than in the past to accomodate not only the new LGA-1366 socket but the larger coolers that come with the Core i7. Some solid capacitors are found on one side of the socket, while the Hi-C capacitors are found lengthwise between the socket itself and a row of chokes, cooled by the heatsinks placed directly on top. Cooling measures seem somewhat tame with a low-rise heatsink over the South Bridge, connected to the the larger heatsink over the North Bridge by way of two heat pipes. Even this cooler seems rather slim in profile with a round set of fins that rise vertically away from the PCB. These remain separate from the other nearby heatsink, all of which rely on passive cooling only. There are no active measures taken at all, as well as no ventilation access to the exterior of the chassis. Spring-loaded clips help keep each of the heatsinks in place, although one board we received had a few of these broken off in transit rendering the board inoperable.
Of course, one of the main things that sticks out is the 6 DIMM slots, which offer support for three channels of DDR3. Joined channels are colored either blue or black, and unlike builds of the past, you start from the outside first. That is, the black slots must be populated before the blue ones. There's not much room between these slots and the edge of the board, so nothing else gets placed here other than the 24-pin power connection. With little space to give, the floppy drive connector finally meets its end here. Directly to the left of these slots are six front-angled SATA ports and a single IDE port. Four other SATA ports, colored in blue, rise vertically away from the board and, with the IDE port, are controlled separately of the rest by the nearby JMicron JMB363 chip. These are placed just ever so slightly to the right of the PCI-E X16 slots and should allow larger graphic cards to fit in around them without a conflict.
The primary PCI-E X16 slot is sandwiched in between two PCI-E X1 slots, one of which will mostly likely be used with the X-Fi Xtreme riser card included in the package. Two other standard PCI slots are intermixed with the other PCI-E X16 slots. Most users will probably settle in with only a single or dual-graphic card configuration which is a good thing as far as the Eclipse is concerned - going down the Triple SLI route runs the risk of covering nearly all of the headers placed along the edge of the board with the possible exception of the front panel pins. That list includes two yellow USB headers, the IEEE-1394 header in green, a fan header, and the handy when necessary power, reset, and D-LED2 buttons. Of these, the loss of the buttons is a minor concern as they are only used to cycle through displays on the D-LED2 readout, while power and reset will probably only be used during the build process or by testers like ourselves. These buttons also have a tendency to stick and hold, and don't bounce back nicely as we expected. |
| MSI Eclipse - BIOS & Overclocking | ||||
Our first attempt at updating the Eclipse's BIOS met with complete failure; using the built-in MFlash update tool, we rendered the board inoperable as the revision we used apparently was not compatible with this feature. This kink seems to be ironed out now with the newer BIOS we have used, but just points out the relative immaturity of the platform, and what pitfalls can befall a new purchase. When we received our replacement board, we were able to get back to business and take some photos of the AMIBIOS that the Eclipse uses. Again, nothing earth-shattering to report and frequent MSI adopters will find some comfort in these familiar confines. Standard and Advanced BIOS sections report on the basics: what drives and CPU are installed, which drive to boot from, etc. There is a wealth of information regarding the specifics of the CPU and Memory installed, just in case you need to refer to it later. CPU Specifcations will display the frequency, ratio, and amount of cache memory as well as a listing of supported extensions and other technologies. Listed on the Memory-Z page are the populated DIMM slots, with the SPD timings displayed for each stick of RAM.
Cell Menu for the Eclipse is what Extreme Tweaker is to the ASUS Rampage II Extreme. Although the options listed here are not as vast as what we saw with ASUS' BIOS, the Cell Menu is very straightforward and easy to use. To keep things simple and neat, major sub-sections are highlighted with arrows. To configure the QPI settings hit Enter on the QPI Configuration. Want to check on the memory timings? Open up MEMORY-Z, then enter the Advance DRAM Configuration to tweak them even further. Of course, the Cell Menu isn't the only place you'll want to check out to maintain a stable environment. A big part of overclocking is keeping your components cool. H/W Monitor is where you'll get instant readouts of board and CPU temperatures as well as current fan speeds and voltages. Fans installed on the indivdually marked headers on the motherboard can be set at 50%, 75%, or 100% depending on the temperature / noise ratio you're looking for. In our overclocking process, we set all fans to 100% for maximum effect.
Memory ratios are oddly listed solely by a single number, thus you need to click on each number to see what effect it has on the memory speed. Displaying a ratio value or a speed setting would definitely streamline the process. Up to four overclocking profiles can be saved or loaded in, preserving one's accomplishments for future use. With a thorough understanding of the tools presented to us, we started in earnest on our overclocking.
We were able to only push a few MHz forward, winding up at a 206 MHz base clock as our final overclock. That translates to a 4.11 GHz overclock for our 2.66GHz Core i7 CPU - not bad considering we were using the stock cooler from Intel. In general, we spent very little time tweaking the system as it appears that the board is quite capable of tweaking itself for the most part. Our attempts mostly consisted of simply raising the base clock and trying to keep memory speeds in check, although the lack of available memory dividers made it kind of hard to continue. Higher speed memory would probably have helped us in this regard. |
| Test Setup and SANDRA Results | |||||||||||
How we configured our test systems: When configuring our test systems for this article, we first entered their respective system BIOSes and set each board to its "Optimized" or "High performance Defaults". We then saved the settings, re-entered the BIOS and set the memory for DDR2-1066 with 5-5-5-18 timings (AMD Phenom), DDR3-1333 with 7-7-7-20 timings (Intel Core 2) or DDR3-1066 with 7-7-7-20 timings (Intel Core i7). The hard drive was then formatted, and Windows Vista Ultimate was installed. When the Windows installation was complete, we updated the OS, and installed the drivers necessary for our components. Auto-Updating and Windows Defender were then disabled and we installed all of our benchmarking software, defragged the hard drives, and ran all of the tests.
We started off our testing with SiSoftware's SANDRA 2009, the System ANalyzer, Diagnostic and Reporting Assistant. We ran three of the built-in subsystem tests that partially comprise the SANDRA 2009 suite on the test motherboards (CPU Arithmetic, CPU Multi-Media, and Memory).  
As reported by SANDRA, both of the X58 boards used in this review were operating well within expected levels in respect to both CPU and memory performance. MSI's Eclipse fared a bit better during the CPU tests, while the memory tests seemed to favor ASUS' Rampage II Extreme. We then went back into the BIOS to disable Hyperthreading and re-ran the same tests to see the difference. Some tests, such as Multimedia Float, saw as high as a 60% increase with HT enabled, while CPU Drhystone ALU saw a more meager 13% gain. As there's no penalty for having HT enabled in most circumstances, it surely makes sense to check that this is enabled when setting up your X58-based system. |
| PCMark Vantage and 3DMark06 | ||||||||
We then ran our test motherboards through PCMark Vantage, Futuremark’s latest system performance metric built especially for Windows Vista. PCMark Vantage runs through a host of different usage scenarios to simulate different types of workloads, including High Definition TV and movie playback and manipulation, gaming, image editing and manipulation, music compression, communications, and productivity. Most of the tests are multi-threaded as well, so they can exploit the additional resources offered by multi-core CPUs.
What really jumps off of the page in the PCMark Vantage results is the wide margin between the older boards and the two X58s in regards to the gaming performance. A lot of this can be explained by the new architecture: with an on-die memory controller, lateny is decreased and increased bandwidth is offered by that extra memory channel. As expected, all other scores increased as well, but to a lesser degree. The overall PCMark scores gains were almost 10% higher than the similarly clocked Core 2 Quad and Phenom CPUs, but about on par with the much faster Core 2 Extreme.
3DMark06's built-in CPU test is a multi-threaded DirectX gaming metric that's useful for comparing relative performance between similarly equipped systems. This test consists of two different 3D scenes that are processed with a software renderer that is dependent on the host CPU's performance. Calculations that are normally reserved for your 3D accelerator are instead sent to the CPU for processing and rendering. The frame-rate generated in each test is used to determine the final score.
Solely looking at the CPU performance with 3DMark06, we saw the two X58 boards running nearly parallel with each other. The extra cores in the Core i7 920 combined with the Hyperthreading technology allow 3DMark to churn through this benchmark in record time. The Core i7 920 in our X58 boards was also to pull out just in front of the Core 2 Extreme QX9770, and easily handled the Core 2 Quad and AMD Phenom X4. |
| Cinebench and POV-Ray Rendering | ||||||||
Cinebench R10 is an OpenGL 3D rendering performance test based on Cinema 4D. Cinema 4D from Maxon is a 3D rendering and animation tool suite used by 3D animation houses and producers like Sony Animation and many others. It's very demanding of system processor resources and is an excellent gauge of pure computational throughput. This is a multi-threaded, multi-processor aware benchmark that renders a single 3D scene and tracks the length of the entire process. The rate at which each test system was able to render the entire scene is represented in the graph below.
The Core i7 matched up with the Core 2 Quad Q9400 in the single-threaded cinebench rendering run, but was no match for the sheer speed of the Core 2 Extreme, which finished with a score roughly 16% higher. However, due to the efficiency of the Core i7 and the re-addition of HyperThreading, these CPUs managed to pull ahead of the Core 2 Extreme in the multi-threaded benchmark.
POV-Ray, or the Persistence of Vision Ray-Tracer, is an open source tool for creating realistically lit 3D graphics artwork. We tested with POV-Ray's standard included benchmarking model on all of our test machines and recorded the scores reported for each. Results are measured in pixels-per-second (PPS) throughput.
POV-Ray Rendering results had the Core i7 920 / X58 slipping behind the Core 2 Extreme by a decent margin, but faster than both the Core 2 Quad and AMD Phenom X4. As with the Cinebench rendering, we were able to get the most out of the MSI Eclipse as far as the X58 boards go, although the margin of difference this time over the Rampage II Extreme was far less, amounting to less than a percent. |
| Low-Res Gaming Framerates | ||||
For our next set of tests, we moved on to some in-game benchmarking with ET Quake Wars and Crysis. When testing processors and motherboards in these games, we drop the resolution to 800x600 and reduce all of the in-game graphical options to their minimum values to isolate CPU and memory performance as much as possible. However, any in-game effects, which control the level of detail for the games' physics engines and particle systems, are left at their maximum values, since these actually do place some load on the CPU rather than GPU. Both game engines produced excellent frame rates when fueled by the Core i7 and its triple channel DDR3, while the graphics card remained the same. And, even though we've seen most, if not all of the synthetic and rendering benchmarks go MSI's way, the Rampage II Extreme pulled out on top in both ET: Quake Wars and Crysis. |
| High-Res Gaming and SLI Testing | ||||
We started our high-resolution 3D testing with Crysis again. This time we have the resolution dialed up in the game to 1600X1200. We also turned up all quality settings and anisotropic filtering for textures to maximum levels.
Crysis is completely GPU bound with a single card, as is evidenced by the near identical frame rates on all systems. Performance gains with a second graphics card added, running in SLI mode, were appreciable with Crysis. We saw a 50% increase on the 780i, and nearly a 67% increase on the X58 boards. |
| LameMT MP3 Encoding and Power Consumption | ||||||||
In our custom LAME MT MP3 encoding test, we convert a large WAV file to the MP3 format, which is a popular scenario that many end users work with on a day-to-day basis to provide portability and storage of their digital audio content. LAME is an open-source mid to high bit-rate and VBR (variable bit rate) MP3 audio encoder that is used widely around the world in a multitude of third party applications. We created our own 223MB WAV file and converted it to the MP3 format using the multi-thread capable LAME MT application in both single and multi-threaded modes. Processing times are recorded below, listed in seconds. Shorter times equate to better performance
Our two X58s were in sync with each other, producing the exact same times for both the single-threaded and multi-threaded encoding runs. The Core 2 Extreme led the tandem of X58s by finishing 6 seconds faster using a single core, but only a mere second when all cores are getting utilized. Again we see the Core 2 Quad Q9400 falling behind the Core i7 920, even though the rated speeds are exactly the same.
We'd like to cover a few final data points before bringing this article to a close. Throughout all of our benchmarking and testing, we monitored how much power our test systems consumed using a power meter. Our goal was to give you all an idea as to how much power each configuration used while idling and while under a heavy workload. Please keep in mind that we were testing total system power consumption at the outlet here, not just the power being drawn by the processors alone.
|
| Performance Summary and Conclusion | ||||||||
Performance Summary: The MSI Eclipse and ASUS Rampage II Extreme performed at near identical levels according to our benchmark results. Throughout our testing, the boards traded victories and typically finished within a percentage point of one another. What we hope you take from of this article, however, is how well the X58 paired with Intel's Core i7 performs compared to quad-core Core 2 CPUs and other platforms. As the benchmarks show, these boards in combination with Intel's latest desktop processors are truly in a league of their own at this time.
ASUS RAMPAGE II EXTREME
|
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;){kind=link}
;){kind=link}
;){kind=link}
;){kind=link}
;){kind=link}
;){kind=link}