| Intel Core i7-2600K Sandy Bridge CPU |
| Reviews - Featured Reviews: Processors | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Written by David Ramsey | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Monday, 03 January 2011 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Intel Core i7-2600K Sandy Bridge CPU ReviewIntel's processor development follows a regular "tick-tock" cycle. The "tick" is the refinement of an existing architecture; the "tock" is a new architecture. Proceeding at a roughly yearly pace, the "tick-tock" model brought us the 45nm Nehalem architecture processors (the original Core-i3, -i5, and -i7 CPUs) as a "tock", and the subsequent 32nm Westmere processors as the "tick" part of the cycle. Now, Intel introduces their new Sandy Bridge architecture as the latest "tock", and Benchmark Reviews checks out the new Sandy Bridge-based Core i7-2600K. This unlocked, 3.4GHz, Hyper-Threading, quad-core CPU is the top of the Sandy Bridge line, and we'll see how it performs against the best AMD processors and Intel's own as well. Intel's first quad-core processors were merely two dual-core dies on a single chip; the CPUs had to communicate across the front-side bus. Later iterations put all four CPU cores on a single slice of silicon. In a similar fashion, Intel's Clarksdale processors had on-chip video, but it was simply a separate GPU placed on the same chip as the CPU cores— even the process used was different: 32nm for the CPU cores and 40nm for the GPU. But the Sandy Bridge CPUs make the transition to a truly integrated product, with all four CPU cores and a GPU core on the same silicon. There's even a shared Level 3 cache that's used by both the CPU and GPU cores.
The integrated graphics core is the Intel HD Graphics 3000core, and Intel promises about twice the performance of the graphics core in the Clarksdale-architecture Core i5-661 processor. Like the CPU cores, the GPU core uses Intel's Turbo Boost technology to increase its power draw and performance when thermal and power headroom permit. Since the Cougar Point motherboards I had available for this test were all based on the P67 Express chipset, which doesn't support the integrated graphics of the Sandy Bridge processors, I wasn't able to test the graphics features of this CPU.
Manufacturer: Intel Corporation Full Disclosure: The product sample used in this article has been provided by Intel Corporation. Intel Sandy Bridge FeaturesThe following information is courtesy of Intel Corporation
Core i7-2600K SpecificationsThe following chart shows the specifications of all the desktop-level Sandy Bridge processors.
In the next section we'll take a look at the Sandy Bridge architecture this processor is based on. Sandy Bridge CPU ArchitectureSandy Bridge CPUs represent the "tock" on Intel's famous "tick-tock" product development cycle, wherein new architecture features are introduced with a "tock" and process refinement comes roughly a year later with the "tick". In the previous generation, the original Core i3, Core i5, and Core i7 Nehalem processors were the "tock", and the Westmere processors were the "tick". Here's Intel's official chart: 
Based on this chart, we'll see Ivy Bridge CPUs built on a 22nm process in about a year; this will doubtless bring additional clock speed and power consumption improvements. But what are the new features of the Sandy Bridge architecture? Internal changes aside, Intel touts these new features:
Sandy Bridge CPUs come in three varieties: there are versions for mobile (laptop) use, "lifestyle PC" use, and "desktop enthusiast" use. The P67 Express/Core i7-2600K platform I tested represents the high end of the "desktop enthusiast" line. Intel claims that overall performance improvements (presumably at similar clock speeds) are in the range of 30%. Note that this includes graphics performance improvements (which I was unable to test since the P67 Express-based motherboards I had do not support the Sandy Bridge integrated graphics), as well as improvements from programs specifically coded to use the new AVX instructions. 
As an "enthusiast" platform, the P67 Express/Sandy Bridge combination suffers from a dearth of PCIe lanes. The P67 has the same 8 lanes as does the P55, but at least upgrades them to the full 2.0 (5Gb/s) specification from the P55's 1.0 (2.5Gb/s) spec. This will provide better performance for SuperSpeed USB 3.0 and SATA 6Gb/s ports (see Benchmark Reviews' coverage of the ASUS P7H55D-M EVO motherboard for the difference this can make in USB 3.0 performance), as well as other PCIe devices you might put in your system. Still, combined with the sixteen PCIe lanes on a Sandy Bridge CPU, that's only 24 lanes total, which pales in comparison to the 40 lanes available on an X58 system or the 42 on an AMD 890FX system. Enthusiasts should think carefully about their current and future system configurations and determine if the available PCIe lanes on a P67 system will be sufficient. Processor Testing MethodologyThe Intel Cougar Point P67 Express chipset and Sandy Bridge processors will eventually replace the older P55 chipset. However, Intel promises much greater levels of performance from their new platform. For this test I had access to three P67 Express-based motherboards for the Core i7-2600K CPU: an Intel DB67BG, an ASUS P8P67, and an ASUS P8P67 EVO. For comparison platforms, I used an AMD 890FX-based system with an AMD 1100T six-core CPU and an X58-based system with an Intel Core i7-950 processor (whose price is a rough match for the Core i7-2600K CPU used on the P67 systems). At $269, the AMD 1100T Black Edition is cheaper, but it's the most expensive consumer processor AMD makes. Intel released a new BIOS for the DP67BG motherboard as I was partway through testing, but despite numerous attempts I was unable to flash the board to the new BIOS- the Windows-based update utility would report "Success" every time without actually updating the BIOS, and attempting to use the flash utility in the existing BIOS merely rebooted the motherboard, then shut it down after a few seconds. For each P67 motherboard, I tested at both stock settings as well as the highest overclock I could achieve. Overclocking a Cougar Point/Sandy Bridge system is different from what you're used to, as I'll detail in the "Overclocking" section after the test results. I tried for the maximum overclock I could achieve with all four processor cores running under load; the results for each P67 motherboard are summarized in this chart.
The labeling of the results in the charts may be slightly confusing: the first two columns are the 890FX/AMD 1100T and the X58/Intel Core i7-950 platforms; the next six columns are all the Intel Core i7-2600K processor, at stock and overclocked speeds in the Intel DP67BG, ASUS P8P67, and ASUS P8P67 EVO motherboards, respectively. Intel P67 Test Platforms
Intel X58 Test Platform
AMD 890FX Test Platform
Benchmark Applications
AIDA64 Extreme Edition TestsAIDA64 Extreme Editionis the evolution of Lavalys' "Everest Ultimate Edition". Hungarian developer FinalWire acquired the rights to Everest in late November 2010, and renamed the product "AIDA64". The Everest product was discontinued and FinalWire is offering 1-year license keys to those with active Everest keys. AIDA64 is a full 64-bit benchmark and test suite utilizing MMX, 3DNow! and SSE instruction set extensions, and will scale up to 32 processor cores. An enhanced 64-bit System Stability Test module is also available to stress the whole system to its limits. For legacy processors all benchmarks and the System Stability Test are available in 32-bit versions as well. Additionally, AIDA64 adds new hardware to its database, including 300 solid-state drives. On top of the usual ATA auto-detect information the new SSD database enables AIDA64 to display flash memory type, controller model, physical dimensions, and data transfer performance data. AIDA64 v1.00 also implements SSD-specific SMART disk health information for Indilinx, Intel, JMicron, Samsung, and SandForce controllers. All of the benchmarks used in this test— Queen, Photoworxx, ZLib, hash, and AES— rely on basic x86 instructions, and consume very little system memory while also being aware of Hyper-Threading, multi-processors, and multi-core processors. Of all the tests in this review, AIDA64 is the one that best isolates the processor's performance from the rest of the system. While this is useful in that it more directly compares processor performance, readers should remember that virtually no "real world" programs will mirror these results. 
The Queen and Photoworxx tests are synthetic benchmarks that iterate the function many times and over-exaggerate what the real-world performance would be like. The Queen benchmark focuses on the branch prediction capabilities and misprediction penalties of the CPU. It does this by finding possible solutions to the classic queen problem on a chessboard. At the same clock speed theoretically the processor with the shorter pipeline and smaller misprediction penalties will attain higher benchmark scores. Despite its comparable clock speed and two extra cores, the AMD 1100T falls well behind the Intel processors in the Queen test. Even the slower-clocked Core i7-950 beats it, and the Core i7-2600K, especially when overclocked, dominates the results. Here we see a pattern that will be similar throughout all these tests: at stock clock speeds, the Intel DP67BG motherboard and the two ASUS motherboards return virtually identical performances, while the higher overclocks the ASUS boards can reach provide greater performance than the relatively limited overclock the DP67BG was capable of. Like the Queen benchmark, the Photoworxx tests for penalties against pipeline architecture. The synthetic Photoworxx benchmark stresses the integer arithmetic and multiplication execution units of the CPU and also the memory subsystem. Due to the fact that this test performs high memory read/write traffic, it cannot effectively scale in situations where more than two processing threads are used, so quad-core processors with Hyper-Threading have no real advantage. The AIDIA64 Photoworxx benchmark performs the following tasks on a very large RGB image:
The Photoworxx test rankings are identical to the Queen test rankings, but the AMD 1100T drops even further behind the Intel results, which are clustered together with only a 16% difference separating the Core i7-950 from the overclocked Core i7-2600K. The overclocked 2600K results are much closer to the stock-clocked results than was the case with the Queen test. 
The Zip Library test measures combined CPU and memory subsystem performance through the public ZLib compression library. ZLib is designed as a free lossless data compression library for use on virtually any computer hardware and operating system. The ZLib data format is itself portable across platforms and has a data-independent footprint that can be reduced at some cost in compression. The AES integer benchmark measures CPU performance using AES data encryption. It utilizes Vincent Rijmen, Antoon Bosselaers and Paulo Barreto's public domain C code in ECB mode and consumes 48 MB of memory. Both of these tests are much more applicable to the "real world" than the previous tests. In the ZLib test, the AMD 1100T surges ahead of the Intel 950, posting scores less than 10% slower than the stock-clocked 2600K. Overclocking the 2600K on the top-performance ASUS P8P67 EVO motherboard improves its score by over 36%. The AES test isn't really a fair one: the Advanced Encryption Standard New Instructions (AES-NI) feature in the latest Intel processors dramatically accelerate AES code. Although the AMD 1100T returns a better score than the Intel 950, the stock-clocked Core i7-2600K is still 560% better. Oddly, overclocking the 2600K doesn't yield significantly better results in the AES test. 
Finally, a win for the 1100T. As we've seen in our review of the AMD 1100T Black Edition, AMD processors dominate in this particular benchmark. PCMark Vantage TestsPCMark Vantage is an objective hardware performance benchmark tool for PCs running 32- and 64-bit versions of Microsoft Windows Vista or Windows 7. It's well suited for benchmarking any type of Microsoft Windows Vista/7 PC: from multimedia home entertainment systems and laptops, to dedicated workstations and high-end gaming rigs. Benchmark Reviews has decided to use a few select tests from the suite to simulate real-world processor usage in this article. Our tests were conducted on 64-bit Windows 7, with results displayed in the chart below. TV and Movies Suite
Gaming Suite*
Music Suite
* EDITOR'S NOTE: Hopefully our readers will carefully consider how relevant PCMark Vantage is as a "real-world" benchmark, since many of the tests rely on unrelated hardware components. For example, per the FutureMark PCMark Vantage White Paper document, Gaming test #2 weighs the storage device for 100% of the test score. In fact, according to PCMark Vantage the video card only impacts 23% of the total gaming score, but the CPU represents 37% of the final score. As our tests in this article (and many others) have already proven, gaming performance has a lot more to do with the GPU than the CPU, and especially more than the hard drive or SSD (which is worth 38% of the final gaming performance score). 
The TV and Movies suite concentrates on video playback and transcoding, but only uses two threads at a maximum, so the Intel processor's Hyper-Threading and AMD 1100T's six cores shouldn't be an advantage. Still, the Intel processors are all faster than the 1100T, and the results seem to scale almost directly with clock speed, with the Sandy Bridge architecture seeming to provide little advantage. The Gaming benchmark relies on the hard disk and video card for over 50% of its score (see the Editor's Note above), and we're using the same HDD and video card for all platforms, so the Intel processor's decisive win in this test simply means that Vantage's gaming code is more optimized for Intel processors. Bear in mind, however, that most "real world" games will not show this difference; generally, in games, your video card matters most, followed by the clock speed (not number of cores) of your processor. The PCMark Vantage gaming test can use up to 16 threads, so Hyper-Threading gives the Intel CPUs a real advantage, but very few commercial games will take full advantage of multicore processors. Unlike the Gaming test, the Music test results have more real-world relevance, since multi-threading is much more common in music transcoding applications than it is in games. What's strange here is the exceptional performance of the Nehalem-based Core i7-950 proc, which beats the 2600K's stock results and comes close to its overclocked results. This is something you should be aware of: when Intel (or AMD) change a processor's instructions or architecture, it's not a given that existing code will take full, or any, advantage of it. This is the only benchmark I ran in which the Intel DP67BG motherboard with the stock-clocked 2500K CPU performed noticeably worse than the ASUS boards at stock clock speeds. Futuremark's weighing of the various system components in each test is the subject of some debate; and some of their choices (such as the Gaming test's use of a 1024x768 resolution with no anti-aliasing or texture filtering being "representative" of the "consumer experience") seem odd to me, but the TV and Movies and Music benchmarks are arguably reasonable predictors of overall system performance. CINEBENCH R11.5 BenchmarksMaxon CINEBENCH is a real-world test suite that assesses the computer's performance capabilities. CINEBENCH is based on Maxon's award-winning animation software, Cinema 4D, which is used extensively by studios and production houses worldwide for 3D content creation. Maxon software has been used in blockbuster movies such as Spider-Man, Star Wars, The Chronicles of Narnia, and many more. CINEBENCH Release 11.5 includes the ability to more accurately test the industry's latest hardware, including systems with up to 64 processor threads, and the testing environment better reflects the expectations of today's production demands. A more streamlined interface makes testing systems and reading results incredibly straightforward. The CINEBENCH R11.5 test scenario comprises three tests: an OpenGL-based test that models a simple car chase, and single-core and multi-core versions of a CPU-bound computation using all of a system's processing power to render a photo-realistic 3D scene, "No Keyframes", the viral animation by AixSponza. This scene makes use of various algorithms to stress all available processor cores, and all the rendering is performed by the CPU: the graphics card is not involved except as a display device. The multi-core version of the rendering benchmark uses as many cores as the processor has, including the "virtual cores" in processors that support Hyper-Threading. The resulting "CineMark" is a dimensionless number only useful for comparisons with results generated from the same version of CINEBENCH. First, let's look at the OpenGL results.
Although this test relies on the graphics card and its OpenGL driver, we still see the top-clocked ASUS P8P67 EVO/2600K combination returning 44% more frames per second than the bottom-scoring Core i7-950. It's a reminder that while your graphics card matters the most in games and tests like this, the processor still contributes a lot. 
The single-core rendering test results might seem very close, but that's an artifact of the scaling in this chart: the stock-clocked Intel Core i7-2600K is 33% faster than the Core i7-950, and that's a substantial difference any way you look at it. The difference (at stock clock speeds) drops to 24% with the multi-core rendering test, and the AMD 1100T's performance here is very impressive given that it can only render six tiles at once, rather than the 8 tiles the Hyper-Threaded Intel processors can manage. Remember: the "virtual cores" provided by Hyper-Threading are not the same as "real cores". CPU-Dependent 3D GamingBenchmark Reviews continually evaluates the various tests and benchmarks we use, and we have switched from Ubisoft's Far Cry 2 benchmark to CAPCOM's Street Fighter IV benchmark. Street Fighter IV uses a new, built-from-scratch graphics engine that enables CAPCOM to tune the visuals and performance to fit the needs of the game, as well as run well on lower-end hardware. Although the engine is based on DX9 capabilities, it does add soft shadows, High Dynamic Range lighting, depth of field effects, and motion blur to enhance the game experience. The game is multi-threaded, with rendering, audio, and file I/O all running in different threads. The development team has also worked to maintain a relatively constant CPU load in all parts of the game so that on-screen performance does not change dramatically in different game scenarios.
I ran the Street Fighter IV benchmark at low-resolution, low settings as well as high-resolution, high settings. Low-resolution settings were 1024x768, no AA, with all other settings set to minimum; high resolution tests were run at 1920x1200 with 8xAA and all other settings maxed out. Low-resolution gaming tests make the video card less of a factor since any high-end video card like the NVIDIA GTX280 used in these tests can easily handle them; differences here are more biased towards processor horsepower. The AMD 1100T brings up the rear here, but the real surprise is that the 3.06Ghz, last-generation Core i7-950 performs identically with the spiffy new 3.4Ghz Core i7-2600K. Again, the latest new processor doesn't necessarily mean better performance. In the high-resolution tests, as expected, the results are all similar, since the performance of the graphics card becomes the primary factor. Still, the AMD 1100T beats the i7-950 and stock-clocked 2600K by about 7.5%. PassMark PerformanceTest 7.0The PassMark PerformanceTest allows you to objectively benchmark a PC using a variety of different speed tests and compare the results to other computers. PassMark comprises a complete suite of tests for your computer, including CPU tests, 2D and 3D graphics tests, disk tests, memory tests, and even tests to determine the speed of your system's optical drive. PassMark tests support Hyper-Threading and systems with multiple CPUs, and allow you to save benchmark results to disk (or to export them to HTML, text, GIF, and BMP formats). Knowledgeable users can use the Advanced Testing section to alter the parameters for the disk, network, graphics, multitasking, and memory tests, and created individual, customized testing suites. But for this review I used only the built-in CPU tests, which aren't configurable. PassMark computes a "CPU Marks" score based on the scores of the individual tests:
The Cougar Point/Sandy Bridge systems show a dramatic advantage over the older 890FX/X58 systems, with scores at least 40% higher. But this score is a composite of the scores returned by the other, individual tests...let's take a look at them.
Integer and floating point operations are the basic things modern CPUs do. Integer operations are everything except floating point; technically, even instructions like comparisons, branches, and bit rotates are integer instructions. Floating point instructions deal specifically with floating point math operations. For example, an integer division of 2 into 7 will return "3" as the result, whereas a floating point division of 2 into 7 will return "3.5" as the result. While most program code is comprised of integer instructions, floating point instructions are important in modeling and rendering applications Intel CPUs utterly dominate in the integer tests, with even the mid-range Core i7-950 beating the high-end AMD 1100T by more than 140%. On the floating point side of things, though, the order reverses, with the AMD processor beating even the overclocked Core i7-2600K. The excellent floating point results of the AMD CPUs help explain how AMD processors keep up in the rendering benchmarks. 
The Compress and String Sort benchmarks are both integer-based, and thus the Intel CPUs dominate here. Overclocking the 2600K provides more of a boost with compression than sorting, apparently. 
The AMD 1100T wins (barely) against stock-clocked Intel CPUs in the Encryption test, beating the 950 and matching the 2600K. This benchmark also responds particularly well to the overclocked 2600K, with almost 40% better performance at the highest level. Intel jumps back into the lead in the Physics test, though, with the 1100T falling behind every Intel processor. 
The Primes test shows all the processors clustering together at their stock clock speeds, with the 2600K showing about an 8% advantage, but this is another test where overclocking the 2600K yields dramatic results, with more than a 30% improvement in the scores. The P67 systems surge ahead in the matrix multiplication tests, though, with the stock-clocked Sandy Bridge CPUs more than twice as fast as the 950, and more than three times as fast when overclocked. Handbrake Media EncodingIt's a truism that consumer-level computer performance reached the "fast enough" point years ago, where increases in system performance don't make thing any faster for most people. Web browsing, e-mail, word processing, and even most games won't benefit dramatically from a super-fast CPU. There are some exceptions, though, and media encoding is one of them: transcoding video, especially high-definition video, can bring the strongest system to its knees. Fortunately, media transcoding is one of those things that (depending on the design of the code, of course) that scales really well with both clock speed and the number of cores, so the more you have of both, the better your results will be. The free and open-source Handbrake 0.94 video transcoder is an example of a program that makes full use of the computational resources available. For this test I used Handbrake 0.94 to transcode a standard-definition episode of Family Guy to the "iPhone & iPod Touch" presets, and recorded the total time (in seconds) it took to transcode the video.
As the only six-core CPU in the test, the AMD 1100T tries its best, but it can only beat the four-core i7-950 by about 7%, and is badly spanked by the Cougar Point/Sandy Bridge systems. Intel identified video transcoding as one of the prime targets for performance improvements with the Sandy Bridge processors, and although this version of Handbrake does not make use of the Intel Quick Sync Video Technology implemented in these CPUs, it's telling that the four-core 2600K matches the six-core i7-980X, which, although not shown in this chart, required 132 seconds to encode the same video. Upcoming encoders that do use this feature will show even greater performance. x264 HD Benchmark 3.19Tech ARP's x264 HD Benchmark comprises the Avisynth video scripting engine, an x264 encoder, a sample 720P video file, and a script file that actually runs the benchmark. The script invokes four two-pass encoding runs and reports the average frames per second encoded as a result. The script file is a simple batch file, so you could edit the encoding parameters if you were interested, although your results wouldn't then be comparable to others.
Again, the 2600K dominates, turning in 980X-matching performances (the 980X returned 89.6 and 89 frames per second on these two runs) for about a third the price. Overclocking the Sandy Bridge CPU returns performance increases that scale almost linearly with the increase in clock speed. 
Although the frames-per-second numbers are different, the results of runs 3 and 4 are virtually identical to the results of runs 1 and 2, when considered on a processor-to-processor comparison basis. There's no doubt about it: the Intel Core i7-2600K processor is a video transcoding monster. SPECviewperf 11 testsThe Standard Performance Evaluation Corporation is "...a non-profit corporation formed to establish, maintain and endorse a standardized set of relevant benchmarks that can be applied to the newest generation of high-performance computers." Their free SPECviewperf benchmark incorporates code and tests contributed by several other companies and is designed to stress computers in a reproducible way. SPECviewperf 11 was released in June 2010 and incorporates an expanded range of capabilities and tests. Note that results from previous versions of SPECviewperf cannot be compared with results from the latest version, as even benchmarks with the same name have been updated with new code and models. SPECviewperf comprises test code from several vendors of professional graphics modeling, rendering, and visualization software. Most of the tests emphasize the CPU over the graphics card, and have between 5 and 13 sub-sections. For this review I ran the Lightwave, Maya, and Seimens Teamcenter Visualization tests. Results are reported as abstract scores, with higher being better. LightwaveThe lightwave-01 viewset was created from traces of the graphics workloads generated by the SPECapc for Lightwave 9.6 benchmark. The models for this viewset range in size from 2.5 to 6 million vertices, with heavy use of vertex buffer objects (VBOs) mixed with immediate mode. GLSL shaders are used throughout the tests. Applications represented by the viewset include 3D character animation, architectural review, and industrial design. MayaThe maya-03 viewset was created from traces of the graphics workload generated by the SPECapc for Maya 2009 benchmark. The models used in the tests range in size from 6 to 66 million vertices, and are tested with and without vertex and fragment shaders. State changes such as those executed by the application- including matrix, material, light and line-stipple changes- are included throughout the rendering of the models. All state changes are derived from a trace of the running application. Siemens Teamcenter Visualization MockupThe tcvis-02 viewset is based on traces of the Siemens Teamcenter Visualization Mockup application (also known as VisMockup) used for visual simulation. Models range from 10 to 22 million vertices and incorporate vertex arrays and fixed-function lighting. State changes such as those executed by the application- including matrix, material, light and line-stipple changes- are included throughout the rendering of the model. All state changes are derived from a trace of the running application.
The SPECviewperf suite is a good example of a real-world test of applications that would normally be the province of a high-end workstation: the individual tests comprise code and models from real applications, running scripts that do real work. The Intel CPUs dominate the Lightwave and TCVIS tests, but it's the Maya test that's the real surprise: in previous testing, AMD processors have done better than Intel processors in the Maya test, with the AMD 1100T beating even the Core i7-980X Extreme Edition CPU, and we can see echoes of that here with the i7-950's lower score compared to the 1100T. But the P67/Sandy Bridge systems are much faster than anything else. This is another example where the 2600K scores are better than the 980x scores...in fact, they're about twice the score of the 980X in this same test. SPECapc LightwaveSPECapc (Application Performance Characterization) tests are fundamentally different from the SPECviewperf tests. While SPECviewperf tests incorporate code from the various test programs directly into the benchmark, the SPECapc tests are separate scripts and datasets that are run against a stand-alone installation of the program being benchmarked. SPECapc group members sponsor applications and work with end-users, user groups, publications and ISVs to select and refine workloads, which consist of data sets and benchmark script files. Workloads are determined by end-users and ISVs, not SPECapc group members. These workloads will evolve over time in conjunction with end-users' needs and the increasing functionality of PCs and workstations. For this test, I ran the SPECapc "Lightwave" benchmark against a trial installation of Newtek's Lightwave 3D product. The benchmark, developed in cooperation with NewTek, provides realistic workloads that simulate a typical LightWave 3D workflow. It contains 11 datasets ranging from 64,000 to 1.75 million polygons and representing such applications as 3D character animation, architectural review, and industrial design. Scores for individual workloads are composited under three categories: interactive, render and multitask. The benchmark puts special emphasis on processes that benefit from multi-threaded computing, such as animation, OpenGL playback, deformations, and high-end rendering that includes ray tracing, radiosity, complex textures and volumetric lighting. The test reports three scores: Animation (multitasking), Animation (interactive), and Rendering. The numeric scores represent the time it took to complete each section of the benchmark, in seconds, so lower scores are better. I've found the SPECapc Lightwave 3D test to be an excellent indicator of overclock stability. In many cases, overclocked systems that will make it through every other benchmark here will crash in this test. 
Although this test stresses system components other than the processor (the video card's OpenGL implementation, for example), it still shows obvious performance differences in the CPUs. The AMD 1100T ekes out a couple of very narrow victories over the Core i7-950 in two of the three tests, but it can't compete with the 2600K. In the Animation (Multitasking) section, we see a very nice performance scaling with frequency for the Intel processors, with results following clock speed almost perfectly. This pattern is repeated in the Animation (Interactive) and Rendering sections, although the differences are less notable. The 1100T's relatively poor showing here prove that six physical cores don't always beat four physical cores. Core i7-2600K OverclockingThe Cougar Point/Sandy Bridge platform brings major changes to the overclocking process. Here are the bullet points:
Overclocking by raising the motherboard's base clock is now all but impossible. Of the three P67-based motherboards I tried to overclock, the highest increase to the 100MHz base clock that I could get to run through stress testing was...103MHz. The Intel DP67BG motherboard wasn't stable above 102MHz, and couldn't run above 101MHz in stress testing. This limited overclocking ability is apparently because the P67's base clock is used to derive almost every other clock in the system, including the SATA and USB clocks. While having a single clock be the base for every other clock in the system probably means cheaper, more reliable motherboards, it removes an overclocking mechanism enthusiasts have used for many years. Intel compensates for this by giving all Sandy Bridge processors unlocked multipliers: K-series processors get "fully unlocked" multipliers with no limits, while non-K series processors are "limited unlocked" CPUs that can only have their multipliers increased by a maximum of 4. All Sandy Bridge processors have fully unlocked video cores, RAM multipliers, and power settings, so you can tweak your RAM and on-board video with any motherboard, but for actual CPU core overclocking, you'll want a P67 Express motherboard. Overclocking Sandy Bridge CPUs is different in another way, too. While everyone has their own overclocking techniques, I generally like to disable "turbo" features and run all processor cores as fast as I can under stress by raising the base multiplier. Well, you can't do this with the Intel Core i7-2600K: in fact, you can't increase the base multiplier at all! This was true in both the Intel DP67BG motherboard and ASUS motherboards I used, so I suspect this limitation is built into either the processor or the P67 chipset. Your only option is to increase the multiplier that will be used by Turbo Boost, and you can set individual multipliers to be used when 1, 2, 3, or all 4 cores are in use. Thus, if you disable Turbo Boost technology, you can't overclock the processor at all. Amusingly, Intel's press kit describes the the P67 Express/Sandy Bridge platform as having "flexible overclocking options"... I had the best overclocking results from the ASUS P8P67 EVO motherboard. In the snazzy ASUS graphical EFI BIOS, you set the base clock and Turbo frequency in the "AI Tweaker" screen as shown below: 
You can adjust the Turbo Boost settings for the Core i7-2600K in two ways: by all cores, or by individual cores. The first setting uses the same Turbo Boost ratio whether 1, 2, 3, or 4 cores are active; the second setting allows you to specify distinct multipliers for each situation. I'm interested in how much performance I can get when everything's active, so I used the "By All Cores" setting. This has the added advantage of being adjustable on the fly from inside Windows using ASUS' "Turbo V EVO" utility. However, you'd probably achieve overall better performance by hand-tweaking the multipliers used in each of the four possible Turbo Boost core configurations. My best results came with a base clock of 103Mhz and an "all cores" ratio of 46, for a final clock speed of 4.738Ghz. This is better than it seems since the 3.8Ghz "maximum Turbo Boost" frequency Intel touts for the Core i7-2600K is when only one core is being stressed; the 4.738Ghz overclock achieved here is when all cores are being stressed. Under these circumstances the stock-clocked 2600K is only running at 3.5Ghz.
This overclock represents a solid 25% overclock from the standard 3.8Ghz Turbo Boost frequency, and applies to all four cores under load rather than the single core the stock 3.8Ghz applies to. This performance differential was reflected in the benchmarks. This is the highest "on air cooling" frequency I've seen with an Intel quad-core processor. I was initially disappointed that I wasn't able to break the 5Ghz barrier, but apparently only about 10% of the initial run of Core i7-2600K processors can do this. Sandy Bridge Final ThoughtsThe Sandy Bridge architecture is the future for Intel, and we can expect to see similar changes wrought to the X58 replacement LGA2011-based systems start to appear later this year. Whether that means integrated graphics cores for the high end or not (hopefully not), the 32nm process and instruction improvements are worth looking forward to. Although the P67/Sandy Bridge combination doesn't compete directly with the X58/Nehalem platform, it's interesting to see how the performance compares to a similarly-priced Socket 1366 processor, the Core i7-950:
On the average, on these benchmarks, the stock-clocked Intel Core i7-2600K is 54% faster than the Core i7-950. Eliminating the AES result from the mix results in an average improvement of 20%. As you can see from the chart, the differences are all over the place, with the relative performance of these processors varying wildly depending on the test used. The 2600K's biggest wins (other than the AES score) are in CINEBENCH, SPECviewperf, and video transcoding, and the performance delta will doubtless increase as programs that use the new vector instructions become available. Intel has made media transcoding performance a major design goal of the Sandy Bridge architecture, and this makes sense given the preponderance of digital audio, photo, and video devices in the consumer market these days. The Core i7-950 processor is near the top of Intel's Socket 1366 line; the next step up, the 960, brings only a trivial 200Mhz clock speed increase accompanied by a price of $562. As I mentioned, in some of the tests the Core i7-2600K at stock speeds even beat the scores of the $999 Core i7-980X Extreme Edition. The 32nm fabrication process brings with it excellent overclocking abilities, as well as much lower processor temperatures: at am ambient temperature of 24 degrees Celcius, the core temperature of the overclocked, overvolted, and under-load 2600K never exceeded 74 degrees Celcius, much lower than you'd see when overclocking a Nehalem processor like the 950. OK, so it's fast and it's cool. Then why am I somewhat disappointed in the Cougar Point/Sandy Bridge platform? Because of the way Intel has locked out overclocking on all but the high end: unless you've a P67 Express-based motherboard and a "K"-series Sandy Bridge CPU, you're not going to be able to get much if any extra performance out of your system. Restricting overclocking to only the highest-end parts may serve Intel in that it prevents lower-end parts from stealing sales from the high end, but it's at the expense of the consumer. After all, the whole point behind overclocking is to wring extra performance from the lower-end, lower-cost parts. Because of this artificial limitation, the more versatile overclocking options and the ability to grow into triple- or quad-CrossFireX systems will keep the AMD 890FX platform competitive in the mid-range market. It will be interesting to see how AMD's upcoming "Fusion" systems compare with Sandy Bridge. Intel Core i7-2600K ConclusionBenchmark tests should always be taken with a grain of salt. It's difficult to try and isolate the performance difference a single component in a computer system makes, especially when it's necessary to compare across different manufacturers and platforms. Complicating the matter is the fact that benchmarks change, a manufacturer may change the technical details of a product, and the retail price may change as well. So please use this review as just one piece of information, and do your research before making a buying decision. Considered on its own, the Intel Core i7-2600K is a very impressive CPU. Its performance compares favorably to the mighty Core i7-980X, and is vastly superior to any of the Socket 1156 processors. At $329.99, the bang for the buck is amazing, and it will only get better as programs that can make full use of the processor's new features become available. Overclocking it to 4.5Ghz or more is easy and leads to massive increases in performance...if it's paired with a P67 Express-based motherboard. If you choose an H67-based motherboard because your application doesn't require the performance of a separate GPU, then you'll be locked out of any overclocking options for the processor.
And things are still a little confused in the Cougar Point/Sandy Bridge world: despite numerous press reports that Intel had licensed NVIDIA SLI technology for the P67 chipset, the ASUS P8P67 motherboard used in this test does not support SLI...although the Intel DP67BG and ASUS P8P67 EVO motherboards do. And that just reinforces the point: what you get out of this processor is very dependent on the motherboard it's paired with. You can have integrated video but no overclocking, or overlocking but no integrated video, or SLI or no SLI...you get the picture. If you plan to build a Sandy Bridge system, some up-front research could save you some grief later. Pros:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Comments
Of course y entire argument relies on AMD finally getting up enough tech power to challenge Intel in the higher end market--something it hasn't done wince 2004.
Ok, ok, it's really easy to break my argument down: If Intel locks us out of OCing, and AMD does not and can come even close to Intel performance, it's a no brainer--AMD will win the day.
So, there's only one option for us: Buy Intel CPUs and give up OCing, or buy AMD cpus and continue OCing. And this is where my second key premise comes in, that AMD will need to bring a lot faster cpu to market than it has today, since even the stock 950 beats it handily. In other words, if I had to buy a processor, and it was a choice between an OC-able AMD that was always going to be slower than a stock, locked Intel, I'd choose the stock, locked Intel.
So the argument, when broken down looks like this:
AMD OCed if and only if Intel is locked and AMD can provide near equal speed by OCing for a fraction of the cost (e.g., The Intel 920 did this).
This all is a factor of Intel being able to beat AMD to a point that they can, again, start dictating the market. For instance, if AMD came out with an OC-able CPU that could nearly match the new 2600 Intel, but only cost 1/3 the price, then OCers would buy that CPU and OCers would have nothing to do with Intel. Again, this harks back to the key point that the only reason I can see Intel doing this is that OCing does cut into their bottom line enough to try and force the sale of the more expensive CPU. If that is a true premise, then again, if AMD can rise to the occasion (like the 920 did), Intel will get smoked out of that market share.
Intel has no high-end Desktop competition today. That's why we're seeing this.
The fastest OC'g AMD chips top out at ~ 4.0, Phenom 4 or 6
Intel can do near 4.5 on their stock cooler
The Intel SB chips are cheaper.
And last but not, their chips are just plane faster clock for clock and just another 25%
Chips are then set to their maximum stable settings, some(ones with less impurities) can be used for over clocking and are sold with an unlocked multiplier.
There is quite a bit more left to how this works, but basically chips with locked multipliers are locked for a reason, not due to companies trying to make more money.
Nice article!
Looks like a separate spread sheet will be needed when all the mobo players get their version of 1155 spec'd out and sorted to the various processors. And yet another for the on-board GPU cycles vs. discrete graphics.
I'm tired just thinking about it.
Happy New Year !
So here we are, at a place, or nearly a place, where we can buy a 300USD CPU and have it be so fast that unless we are doing very specific tasks, and lots of them, we will not see or enjoy more speed from the CPU, and except for a hobby and fun, have no need to OC for practical reasons.
We've had this discussion before: Things are getting so fast and so small, that they are becoming mainstream and uninteresting. It's only a matter of time before you can buy an entire system as fast as any top of the line OCed desktop, and have it fit into your hand.
I'm all for it. Saves me space and time.
Then in the next sentence, you state that processors have gotten so fast that it doesn't make a difference.
So which one is it?
AMD guys will do anything they can to justify the fact that AMD will always be 2nd best to Intel, and there are only 2 companies in competition. LOL
Gotta love the AMD guys for supporting their sub-par performance for the past 6 years.
If you consider real-world performance and saving money more important that bragging about your scores in AIDA64 or CINEBENCH, an AMD hex-core processor might be your best bet.
Now, personally, I'm one of those people who must have the fastest, even if it makes no difference. So my personal machine is an overclocked 980X with 12G of memory, a 256G SSD main drive, and two spanking new GTX580 cards. But you know what? When I drag out my overclocked 1090T AMD system with Radeon 5770s, I never notice any difference in gameplay at 1920x1200 in games like Crysis, Metro 2033, and Bad Company 2. So what did all that extra money for the 980X system really buy me?
The point is that unless you're the kind of person who brags about their car's 1/4 mile times (utterly irrelevant to real-world driving), you should plan where to spend your performance dollars.
For example the old Voodoo2
Max 640*480 refresh = 120 Hz
Max 800*600 refresh = 120 Hz
Max 1024*768 refresh = 85 Hz
I`m currently cpu limited with my Q9650 at 4.1ghz and a GTX 580 and this is why i`ve just ordered the i7 2600k along with 8gb of ram and an asus P8P67 deluxe.Money well spent imo seeing as this lot cost much less than your cpu alone and will definitely improve MY gaming along with the encoding i do.
"And this is where my second key premise comes in, that AMD will need to bring a lot faster cpu to market than it has today, since even the stock 950 beats it handily.
If you research into current advancements and and productural timelines, you should begin to see consumer based entry level systems within the next 5-10 years. Intel Knows this, as well as AMD, NVidia/ATI..
One only has to view some the time lines and review current tech reviews and product spec's to see it begin to unfold very quickly. You can blame SSD's, they are what bandwidth is needed to make the all happen, Hard drives were always the bottleneck, you can blow the doors off a better processor and slow drive system with a slower processor and the fastest drive system. Especially if if that drive system can handle multiple reads and writes per clock cycle.
Oh the expense of it all, I could have probably bout an Island in the Bahama's for the money I've spent over the years on PC's
Danny
:)
What the fanboys from Team Orange never want to admit is that when you overclock you pay quite a bit for a beefier power supply, a 3rd party cooler, and a bigger case with more fans to fit the cooler into. Those are not insignificant expenses and can easily exceed $200. So when I see someone trying to compare their OC'd-to-the-max 1090T to an i7-920 running stock I just laugh. Are they really that desperate?
Sandy Bridge is going to be a game changer for sure, and I doubt if Bulldozer will even be enough to catch up to the edge Intel holds just with their current 1156 and 1366 products.
with RamDisk Plus version 11:
##superspeed.com/desktop/ramdisk.php
We already have a 16 GB XP/Pro workstation working
well with a 10 GB ramdisk using an earlier version
of that software:
##supremelaw.org/systems/superspeed/RamDiskPlus.Review.htm
32 GB total - 4 GB for 32-bit OS = 28 GB for a storage partition!
NIIICE!
I think AMD could "leap ahead" by enhancing their
integrated memory controllers to support quad-channel mode.
(I suggested this to them, many weeks ago, but
received no acknowledgment.)
If Intel's Sandy Bridge will only get "shrunk"
at the next "tick-tock" iteration, Sandy Bridge
is "stuck" with dual-channel mode.
AMD, are you reading this?
/s/ Paul A. Mitchell
Systems Development Consultant
Just Saying.
And to think AMD now own ATI, which for years I loved, but crippled those video cards too, thats why NVidia is way better, double precision FPU, parrallel processing etc....Not to mention NVidia eqquips their cards with (PPU), Physics Processing Unit, and a GPU, (Graphics Processing Unit)...Which has not only been the fastest Gamming cards always, but performed Direct Processing for quite some time, which is why BIONIC, folding at home etc., use them for Physics as well as high speed calculations....
I could go on and on, but my point is you get what you pay for period..
If you want the best of the best stay with Intel and NVidia all the way and you will have the fastest, most accurate, stable and reliable computing platform period. Very few BSOD's and computation errors.
Peace out people
Danny
Variables being temperature, voltage, clock speed, etc.
If run properly, they will both return correct results (unless that piece of silicon is actually faulty, which would get detected before sold)
If I remember correctly, the only processor that ever actually computed numbers incorrectly was an old pentium version that had a division error.
This is the major point I am trying to get across....
For me, this has been my experience, run for a month or so and the AMD's return Divide by Zero and compute errors, BSOD's or a host of other issues.
Durability under other than normal working conditions, you'd be wise to Invest into Intel/NVidia...Think about what I said, over seven years, no errors, nothing, they just keep on running at 100%. That in of itself speaks volumes to me about product quality. I spent a thousand dollars a piece for those Pentium 4 Extreme processors and believe me when I tell you that I have gotten more than my monies worth from them. I cant think of anything else you can buy that will run that long at full throttle non-stop with out a meltdown well before 7 years..It's amazing, It's called Intel, It's what should be inside everyones box.
:)
What a shame Intel could not take that one little lesson from what they copied from AMD. I would have gladly purchased one of these new processors if it would have fit my MOBO. This year?s upgrade will put into AMD hardware to go with the 5870 upgrade from 2010.
The next time I upgrade, I may well buy AMD for those reasons, if they can get close to Intel CPU speeds. It would be nice to only have to pop in another CPU and you're done, rather than replacing your RAM, MB, and cpu.
Pretty damn perky Linux system actually.
Thanks.
I am a Comp Sci major and have had to research and study such things as Computer Hardware Architecture, Operating systems etc. I constantly build computers to study them, old to new. I have both AMD and Intel computers, I have a closet full of parts I play with all the time, build computers, run software to check performance etc. take it back apart.
The idea that a CPU would not support an Operating system is plain ridiculous that is absolutely not possible. And there are scientific reasons Intel changes their pin patterns not just to create a new line of cpu's to sell. And Currently AMD cpu speeds not clock speeds but actual measured performance speed of calculations etc. are still pretty far behind Intels chips.
I am tempted to sit here and reply to all of these with facts, but really do not have the time, not sure what I was researching when I found this site.
See my earlier Blog, 28 years of computing experience with both platforms and AMD has yet to live up to its promise to Knock Intel to the wayside.
AMD produces nothing but compute errors, and BSOD's even at stock clocks, which is why none of their motherboards support ECC RAM, why would you spend money to verify compute errors. 1+1=2 not 3 or 4 or 5. Thats why my main rig is Intel, and the AMD is just a file server. I dont like mistakes, And when I do game, I like the fact that my main rig based on Intel/Nvidia smoke the hell out of anything AMD/ATI has ever had to offer. Dont believe me, just look at the history of 3Dmark, SYSMark etc. and see the truth for yourself..
Danny
I run both Intel and AMD rigs, and neither seems to have a stability advantage over the other.
You claim that AMD has "yet to live up to its promise to knock Intel to the wayside." I don't recall AMD ever "promising" such a thing, outside of the usual advertising hyperbole both companies engage in, but when the original Athlon 64 X2 processors came out in late 2003, they decisively smoked everything Intel had at the time, and remained on top of the performance heap for at least a year.
AMD cannot compute high end math and physics very well, period...These arent assertions, theyre plain fact... And believe me if I could I would tell you why, but I am not at liberty to tell you why....
I would love for AMD to be a better processor, but they arent, otherwise I would buy them...
So I'm stuck spending more money to get real world accuracy....
Danny
But I really value your honest opinion and killer benchmark reviews, the are some of the most thorough and accurated and unbiased I have seen to date, in my entire experience with computing, way to go and thank you.
Danny
Danny
I'm talking real world heavy use, Don't tell me no half trut's and I'll tell you no lies my friend...You go ahead and believe in AMD/ATI, thats your right, I run both AMD/Intel, NVidia/ATI, IMO Intel/NVidia combo, hands down, wins period.....
I can be rather pushey by nature, as well as brash, I'm sorry if you took offense, as it wasn't intended that way..
Danny
And your idea that Intel creates nothing but compute errors when over clocking is wrong as well, I also have an AMD computer sitting right here, it overclocks just fine. To properly overclock you have to know what you are doing, many things have to be set inside your BIOS and can't just be directly controlled by a piece of software inside the operating system.
It is depends on your motherboards BIOS, whether you can over clock, what you can over clock and to what extent. You also need to know the math to figure out what settings you should set and to what they should be set. I build computers constantly. I spend days researching motherboards, RAM, CPU's, etc. on manufacturers sites, reading white papers, manuals and other specs, before I decide on a good combination of hardware to build a system with.
It trips me out to see amateur hobbiests on sites like this passing along faulty information based on their weekend trials and errors with computers.
... like integrated memory controllers?
... like native 64-bit logic backwards compatible
with 32-bit instructions?
How quickly we forget!
MRFS
If you want an entry level low tech way to game AMD is what I run.
But I'm 48, and I like to play with movies and editing and audio as well as gamming, and when everything is idle it defaults to Seti@home, BIONIC projects etc. My main entertainment system PC does it all so very very well and so much more reliable. 2600k, Dual SLI GTX 580's...
Those AMD's are religated to file servers, and print servers etc., as well as some of those BIONIC chores etc during their idle time. But they fail misserably at that chore. Oh by the way every single one of those have crossfire going on, so they were'nt exactly cheap eigther...
more to follow:
I'm still Configuring and tweaking the new hardware and software, but initial impression, hell yeah and allright Intel....
Real world computing and serious damn gamming for men with bigger toys..
You get what you pay for, Enterprise quality, speed, perfomance and accuracy, Thank you Intel, and a special thanks to NVida on the SLI GTX 580's..BOOM, BOOM, A, ZOOM, ZOOM!!!!!
more to follow:
One thing I will conclude with is faster RAM, means faster, everything..Running SSD's do better with the fastest board memmory, everything completes so, so, much faster, This machine is a monster, my god, AMD you'd better take note, because the dust you see was left by Intel..
Me happy now, and high David
:)
Intel Community Forum:
#communities.intel.com/index.jspa?iid=hdr+communities
nVidia Community Forum:
#forums.nvidia.com/
Other than that, why is it that I can find no information on processor errors when googling AMD for that information? If you have evidence that AMD is producing faulty CPUs please do post it.
but proven claims about defects in Intel's push-pins are NOT OK??
Surely this will prove your systems overall integrity, not to mention possible problems will usually occur during this burn in run. Really, you should try this out, I learned alot of things about the whole system in general, as I prepare PC's extreme heavy load computational environments, that's all I am trying to convey here..
Danny
Danny
1. What does it mean exactly for a motherboard to "support" Sandy Bridge integrated graphics?
2. Which boards currently do so?
3. Do no P67 motherboards offer this support? That would rule out the ASUS P8P67 DELUXE and the Intel DP67BG motherboard, each of which I was considering to use with the i7-2600K processor.
4. What criteria should one use to determine whether a motherboard can exploit the full capabilities of the i7-2600K processor?
Thanks,
Peter
You need a P67 motherboard to overclock the CPU, but then you can't use the integrated graphics.
An H67 motherboard supports the integrated graphics, but doesn't allow any overclocking.
Don't like that? Complain to Intel. I sure don't.
1. Does this really mean that if you use a P67-based motherboard with the i7-2600K processor, the graphics capabilities of the processor really go entirely unused altogether?
2. I realize it's a very powerful processor in other respects as well, but doesn't that in part defeat the purpose of getting the i7-2600K?
3. I saw a website that suggested a new Intel chipset, the Z68, would be released in the near future (2nd Quarter 2011, possibly). Does anyone know about this, and whether it would be more appropriate for the i7-2600K?
Thanks!
Peter
2. If you use a P67 motherboard and use a video card (presumably with HDMI/DVI output ports), how much worse video performance would you experience compared to using the integrated graphics capabilities that an H67 motherboard would offer?
Thanks,
Peter
No, you don't. Probably.
"If you use a P67 motherboard and use a video card (presumably with HDMI/DVI output ports), how much worse video performance would you experience compared to using the integrated graphics capabilities that an H67 motherboard would offer?"
I haven't tested the integrated Sandy Bridge video, but Intel states it provides roughly twice the performance of their previous generation integrated video, which was unusably slow for modern gaming. So it's a pretty safe bet that almost any separate graphics card costs more than $50 will handily outperform the Sandy Bridge integrated video.
Yes, that's what it means. It's more than just the connectors available on the board; the P67 Express chipset does not have the circuitry required to support Sandy Bridge video.
"...doesn't that in part defeat the purpose of getting the i7-2600K?"
In part, yes, although one could argue that most people buying the top-end Sandy Bridge would want better video. Of course, if you buy that argument, why should the 2600K have integrated video at all? Good question.
Any Z68 stuff is rumor at this point.
"one could argue that most people buying the top-end Sandy Bridge would want better video. Of course, if you buy that argument, why should the 2600K have integrated video at all? Good question."
This is very interesting and also helpful. If a $50 video card can significantly outperform the on-board graphics capabilities of the Sandy Bridge (i7-2600K), it really does raise the question of why Intel bothered to put such a capability on the chip in the first place. Perhaps just "because they could." I suppose this is just a stepping stone to some future processor whose on-chip graphics capabilities WILL be as good or better as those on a separate video card. Maybe. In the meantime, I guess the obvious choice is to choose a P67 motherboard over a H67 motherboard, because
1. P67 motherboards can overclock and H67 motherboards can't
2. a P67 motherboard using a separate video card offers better graphics than an H67 motherboard using the on-chip integrated graphics.
Is this correct? If so, why would anyone opt for an H67 motherboard?