|
|
| Introduction and Related Information | |||||||
Intel hasn't exactly kept many details regarding Nehalem and the X58 Express chipset a well guarded secret this past year or so. In fact, we've already posted a number of articles related to both, dating back to March of last year, in which we write about many of the main features and specifications of Nehalem, its naming convention and the supporting X58 Express chipset. Here is a breakdown of a few of the most pertinent stories...
We're going to summarize many of the main details again here, but if you'd like to check out our complete coverage of the Core i7 (Nehalem) and the X58 Express chipset, the list of articles above covers just about all there is to know. |
| Introducing the X58 Express |
As they typically do when launching a line of processors based on a new micro-architecture, Intel will also be releasing the X58 Express chipset to support Core i7 processors. Due to the fact that Core i7 processors have their memory controllers integrated into the CPU die, however, the X58 isn't so much a brand new chipset as it is an adaptation of existing technologies to support the Core i7.
|
| Core i7 Motherboards |
For the purposes of this article, we got our hands on a number of X58 Express-based motherboards. We've got a trio of them pictured below, and we'll be showing you a couple more in a video spotlight that will accompany this article.
First up, we have the Intel "Smackover" DX58SO. This will be Intel's premiere desktop motherboard for Core i7 processors. It features 4 DIMM slots, a 6-phase power array, and obviously an LGA1366 socket. The expansion slots on the board consist of dual PCI Express x16 PEG slots, dual PCI Express x1 slots, a single PCI slot, and a notched PCI Express x4 slot that features a retention clip to accommodate a standard graphics card. As of this time, the DX58SO supports ATI's CrossFire multi-GPU technology, but it does not support SLI.
Next up we have the elaborate Gigabyte GA-EX58 Extreme. Of the three motherboards featured here, this one has the most outrageous cooling apparatus, which includes an optional bolt-on, slot mounted array of heat-pipes and heatsink fins and an integrated water-block. The GA-EX58 Extreme is a member of Gigabyte's new Ultra Durable 3 family of products, and as such it supports the company's Dynamic Energy Saver (DES) technology and sports all solid capacitors. The board's slot configuration consists of three PCI Express x16 PEG slots, a single x1, a notched x4 slot, and dual PCI slots. We should note, the GA-EX58 Extreme supports both CrossFireX and SLI.
|
| Core i7 Memory and Coolers |
Over the past few weeks, a number of memory and accessory manufacturers have announced products for the Core i7 / X58 Express chipset in anticipation of the platforms arrival.
To date, we have already received a triple-channel DDR3-1600 memory kit from Corsair, that features an updated heatsink assembly, reminiscent of the company's Dominator DHX line of products. These Corsair TR3X6G1600C8D DIMMs feature 8-8-8-20 timings, a 1.65v voltage, and a 2GB capacity each, for a total of 6GB in a triple channel configuration. We also received a set of Qimonda DDR-1066 modules that operate at 1.65v as well. The Qimonda triple channel kit consists of three 1GB modules, for a total of 3GB, with 7-7-7-20 timings.
We figured this would also be a good time to show you the size difference of the stock Core i7 cooler as it relates to a Core 2 cooler, and a new high-end air-cooler from Thermalright. As you can see in the images above, the new stock Core i7 LGA 1366 cooler has a significantly larger diameter than the Core 2 LGA 775 cooler. There use a similar plastic push-pin locking mechanism, but the pins are spread further out on the Core i7 cooler. Current LGA 775 coolers will not fit the new LGA 1366 socket.
Enthusiasts need not worry though, as Thermalright, and others, are either new high-end heatsinks or new mounting hardware for existing heatsinks to support Intel's new socket. The Thermalright Ultra-120 eXtreme RT pictured here is what we used to overclock the Core i7 Extreme 965 to over 4.1GHz, and it did so with minimal noise output thanks to its dense array of heatsink fins, six heatpipes, and 120mm fan. |
| Our Test Systems and SANDRA | |||||||
Test System Configuration Notes: 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 memory timings for either DDR2-1066 (AMD) with 5,5,5,15 timings, 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 drives were 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, performed a disk clean-up, defragged the hard drives, and ran all of the tests.
|
|
We began our testing with SiSoftware's SANDRA 2009 SP1, the System ANalyzer, Diagnostic and Reporting Assistant. We ran six of the built-in subsystem tests that partially comprise the SANDRA 2009 SP1 suite with Intel's new Core i7 Extreme 965 processor (CPU Arithmetic, Multimedia, Multi-Core Efficiency, Memory Bandwidth, Memory Latency, and Cache and Memory). All of the scores reported below were taken with the processor running at its default clock speeds of 3.2GHz with 3GB of DDR3-1066 RAM running in triple-channel mode on the Intel DX58SO motherboard.
In all of the SANDRA tests we ran, the new Core i7 platform significantly outperformed all of the other systems in SiSoft's reference database. In the CPU Arithmetic and Multi-Media benchmarks the Core i7 finished way out in front of the competition. The same held true in the Multi-Core efficiency benchmark, where the Core i7 blew everything away until hitting the 4x128kb block size, at which point the graphs leveled out.
The Core i7 also shined in all of the various memory-related benchmarks. In its stock form using a DDR3-1066 triple-channel configuration, the Core i7 Extreme 965 put up an impressive 18.33GB/s of memory bandwidth, with a latency score of 76ns, and over 66GB/s of combined cache / memory bandwidth. All three results were superior to every non-Core i7 based system in the database by a wide margin.
| PCMark Vantage Performance | ||||
Next up, we ran a number of different test systems through Futuremark’s latest system performance metric built especially for Windows Vista, PCMark Vantage. 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 the tests can exploit the additional resources offered by a quad-core CPU.
The first thing that jumps right out in these PCMark Vantage test results is the huge advantage the Core i7 powered systems have in the Gaming test versus the competition. All of the Core i7 powered systems blow right past every other rig in the gaming test. In the other tests, the Core i7 remains in the lead, but by smaller margins. The affects of Hyper-Threading on the Core i7 Extreme 965's performance is mixed, with HT helping in some tests and hurting in others. |
| LAME MT and Kribibench | ||||||||
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.
In this test, we created our own 223MB WAV file (a hallucinogenically-induced Grateful Dead jam) and converted it to the MP3 format using the multi-thread capable LAME MT application in single and multi-thread modes. Processing times are recorded below, listed in seconds. Shorter times equate to better performance.
The new Core i7 processors were the fastest of the bunch according to our custom LAME MT benchmarks. Clock for the clock, the Core i7 Extreme 965 was 4 and 5 seconds faster than the Core 2 Extreme QX9700 in the multi- and single-threaded tests, respectively. The 2.66GHz Core i7 920 performed even better than the similarly clocked Core 2 Quad Q9400, beating the Q9400 by 5 and 9 seconds.
For this next batch of tests, we ran Kribibench v1.1, a 3D rendering benchmark produced by the folks at Adept Development. Kribibench is an SSE aware software renderer where a 3D model is rendered and animated by the host CPU and the average frame rate is reported. We used two of the included models with this benchmark: a "Sponge Explode" model consisting of over 19.2 million polygons and the test suite's "Ultra" model which is comprised of over 16 billion polys.
Intel's new Core i7 processors--regardless of their clock speed--were also the fastest of the bunch in the Kribibench rendering tests. The 2.66GHz Core i7 920 was able to outperform the 3.2GHz Core 2 Extreme QX9770, and the flagship Core i7 Extreme 965 finished way out in front. Hyper-Threading also gave the new Core i7 processors a speed boost in this multi-threaded benchmark, to the tune of about 24%. |
| Cinebench R10 and POV-Ray | ||||||||
Cinebench R10 is an OpenGL 3D rendering performance test based on Cinema 4D from Maxon. Cinema 4D 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 new Core i7 processors were once again the fastest of all the test systems, regardless of CPU frequency. The 2.66GHz Core i7 920 finished just behind the 3.2GHz Core 2 Extreme QX9770 in the single-threaded portion of this test, but due to the Core i7's advantages in multi-core efficiency as evidenced by SANDRA, and the addition of HT, the lower clocked Core i7 920 was able to outpace the fastest quad-core Core 2 based processor here. Hyper-Threading also gave the Core i7 965 a performance boost here, due to the multi-threaded nature of this benchmark, and the various X58-based motherboards we tested all put up similar numbers--with a slight edge going to the ASUS board.
POV-Ray , or the Persistence of Vision Ray-Tracer, is a top-notch open source tool for creating realistically lit 3D graphics artwork. We tested with POV-Ray's standard 'all-CPU' benchmarking tool on all of our test machines, and recorded the scores reported for each. Results are measured in pixels-per-second throughput; higher scores equate to better performance.
Update November 5, 2008: Intel informed us of a bug with POV-Ray v3.7 Beta 29 and the Core i7 that causes a threading issue on systems with 4 logical cores. So, we re-ran the tests with POV-Ray v3.7 Beta 28 and have updated the results above. As you can see, performance with all of the Core i7 processors is vastly increased, and they finish well ahead of the competition. |
| 3DMark06 and Vantage CPU Tests | ||||||||
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.
The new Core i7 processors returned to their winning ways, outpacing all other test systems in this test. The Core i7 920 was once again able to pull out in front of the Core 2 Extreme QX9770, and the higher-clocked Core i7 940 and Core i7 Extreme i965 only extend the lead.
3DMark Vantage's CPU Test 2 is a multi-threaded test designed for comparing relative game physics processing performance between systems. This test consists of a single scene that features an air race of sorts, with a complex configuration of gates. There are aircraft in the test that trail smoke and collide with various cloth and soft-body obstacles, each other, and the ground. The smoke spreads, and reacts to the planes as they pass through it as well and all of this is calculated on the host CPU. The results with 3DMark Vantage's CPU Test 2 essentially mirror those of the 3DMark06 results above, with all Core i7 processors outperforming the competition. However, in this test, disabling Hyper-Threading gave the Core i7-based system resulted in measurable increase in performance. |
| Low-Res Gaming: Crysis and F.E.A.R. | ||||
For our next set of tests, we moved on to some in-game benchmarking with Crysis and F.E.A.R. When testing processors with Crysis or F.E.A.R., 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, the 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.
The low-resolution Crysis and F.E.A.R. benchmarks tell somewhat different stories. In Crysis, all of the Core i7-based systems finished ahead of the competition, regardless of clock speed. And the ASUS P6T Deluxe motherboard in particular put up some strong numbers. In F.E.A.R., however, the Core 2 Extreme QX9770 finished strong, besting the Core i7 processors that were running in the Intel DX58SO motherboard. But in the more enthusiast-friendly ASUS and Gigabyte boards the Core i7 Extreme 965 comes out on top. It appears Intel still has some motherboard tuning to do with the DX58SO according to these results. |
| Low-Res Gaming: UT3 and ETQW | ||||
Next, we continued with some more low-res, in-game benchmarking with Unreal Tournament 3 and Enemy Territory Quake Wars. As was the case with the Crysis and F.E.A.R. on the previous page, here we also dropped the resolution to 800x600, and reduced all of the in-game graphical options to their minimum values to isolate CPU and memory performance as much as possible.
Chalk up two more clear victories for the new Core i7 processors. In both Unreal Tournament 3 and Quake Wars Enemy Territory, even the lowest clocked Core i7 920 processor was able to outperform the highest clocked Core 2 Extreme. Disabling Hyper-Threading resulted in a marginal performance increase in both games. And once again the ASUS motherboard we tested put up relatively strong numbers in light of its competition. The Gigabyte board also performed well, besting Intel's offering in both tests, but ASUS seems to have done a bit more tweaking at this early stage of the X58 Express chipset's lifecycle. |
| High-Res Gaming: 3DMark Vantage and Crysis | ||||
For our next set of tests, we moved on to some high-resolution graphics benchmarking with 3DMark Vantage and Crysis. For these tests, we've compared a Core i7 Extreme 965 powered system against two Core 2 Extreme QX9770 powered machines--one running a Radeon HD 4870 X2 CrossFireX configuration (X48), the other a 3-Way SLI configuration (nForce 790i SLI Ultra) consisting of three Zotac GeForce GTX 280 AMP! Edition cards. Our goal was to see if upgrading to a Core i7 CPU had an impact on performance in more GPU bound circumstances.
NVIDIA has clearly done some optimizing for the Core i7 / X58 Express platform. As you can see the 3-Way Zotac GeForce GTX 280 SLI configuration put up much better numbers on the Core i7 platform than it did on the Core 2 Extreme. The Radeon HD 4870 X2 CrossFireX configuration also showed a slight performance increase in 3DMark Vantage, but nowhere near as large as NVIDIA's. And in Crysis, the CrossFireX configuration actually performed a bit lower than the Core 2 Extreme. During the course of writing this article, ATI had actually released two sets of beta / hotfix drivers that incorporated tweaks for Intel's new platform, but it appears there is still some work to be done. We know Intel and ATI are working together to optimize performance for CrossFireX and the Core i7, so we're confident more performance will be coming soon. We should point out that our numbers for the CrossFireX config, while reproducible, do not jibe with ATI's own internal benchmarks. Once we get to the bottom of the issue, however, we will re-run these tests and update the graphs where necessary. |
| High-Res Gaming: UT3 and ETQW | ||||
In this next set of tests, we continued with some more high-resolution in-game benchmarking with Enemy Territory Quake Wars and Unreal Tournament 3. We maximized the in-game graphics options with both game and cranked the resolution up 1920x1200. With ETQW, 4X anti-aliasing and 16X anisotropic filtering were enabled as well.
Both Enemy Territory Quake Wars and Unreal Tournament 3 showed a marked performance improvement when running on the Core i7 platform with either 3-Way GeForce GTX 280 SLI or a Radeon HD 4870 X2 CrossFireX configuration. As we stated on the previous page, however, although both graphics configurations were faster overall, expect even more performance with future driver releases. |
| Total System Power Consumption | ||||
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.
|
| Our Summary and Conclusion | ||||
Performance Summary: We have a number of different things to consider in this summary--the performance of the Core i7 as it compares to the Core 2 and Phenom, the performance of the Core i7 with and without Hyper-Threading enabled, and the performance of the three X58-based motherboards we tested. Thankfully, it's rather easy to summarize the Core i7 processor's performance. In the majority of the benchmarks we ran, the new Core i7 processors outperformed the fastest Core 2 Extreme, the 3.2GHz QX9770. Clock for clock, the Core i7 is markedly faster than the Core 2, which allowed the 2.66GHz Core i7 920 to outperform the 3.2GHz QX9770 on a number of occasions, and the overall the Core i7 Extreme 965 / X58 combo put up the best performance we have seen from a desktop PC platform to date. At no time did the Phenom X4 9950 threaten to overtake the similarly clocked Core i7 920 and the higher clocked 940 and Core i7 Extreme 965 were simply in another league.
What can we say about the Core i7 and X58 Express that the benchmarks haven't already told us? Looking back through the results, it is obvious Intel has raised the desktop CPU performance bar yet again and in a big way. The Core i7 920, 940 and Core i7 Extreme 965 put up some of the best benchmark scores we have seen to date. In a few instances, the Core 2 Extreme is able to come close clock-for-clock, but overall there is no denying the Core i7 represents a significant step up in performance. Even at this early stage, we were also impressed by the X58 Express chipset-based motherboards we tested. Throughout our entire battery of tests, which took place over the better part of week, we did not experience any instability whatsoever and everything we connected to the boards "just worked". We also had a good experience overclocking with the Core i7, which bodes well for the enthusiasts and modders out there contemplating a purchase of one of these new processors.
|