| Intel Core i5-2500K Sandy Bridge CPU |
| Reviews - Featured Reviews: Processors | |
| Written by Hank Tolman | |
| Monday, 03 January 2011 | |
Intel Core i5-2500K Sandy Bridge CPU ReviewOne of the processors in Intel's Sandy Bridge line-up that is being released in early January 2011 is the Intel Core i5-2500K. Intel is calling Sandy Bridge the 2nd Generation of Intel Core Processors. That also happens to be the significance of the 2 in the name of the Core i5-2500K. Sandy Bridge has promised to bring a lot of new ideas and technologies to the computer hardware industry, including improved media and gaming performance. The Intel Core i5-2500K is a quad-core, non-hyper-threaded, 3.3GHz processor equipped with the latest version of Intel's Turbo Boost. Set to be priced at $205 dollars, the Intel Core i5-2500K competes directly with the top end AMD Phenom-II X4 processors in price. In this article, Benchmark Reviews is bringing you the scoop on the Intel Core i5-2500K, including its performance against AMD's newest flagship quad-core, the Phenom-II X4-975BE. With their second generation core processor line, Intel made a few changes. Core i5 processors have previously been found with anywhere from two to four cores and with or without hyperthreading technology. With the release of Sandy Bridge, that changes a little. The basic series Sandy Bridge Core i5 CPUs are all quad-core processors and none of them are equipped with hyperthreading. To take advantage of an extra thread per core in Sandy Bridge, you'll have to invest in a Core i7. Some of the off-series Core i5 CPUs (like the S or T series with lower clock and turbo frequencies) may come with hyperthreading, as may the Core i3 processor line, but you'll be hard pressed to find one of those at a retail location. 
The Sandy Bridge CPU offers a combination of CPU, GPU, and IMC all housed on the same die. This allows for increased speed in processing as well as lower power consumption. The newest installment of Intel HD Graphics housed with the CPU was made with media consumers in mind. Intel also says that the Intel HD Graphics will be able to play most mainstream games. The Sandy Bridge is supposed to be a real game changer. For that reason, we are testing the Intel Core i5-2500K against a whole slew of AMD CPUs. In the past, Core i5 Lynnfield processors have competed pretty well performance-wise with Athlon-II X4 CPUs, which cost a whole lot less. At about $200, the new i5-2500K fits in the same price range as the soon to be released Phenom-II X4-975BE, AMD's newest flagship quad-core processor. Can the i5-2500K outperform the Phenom-II X4-975BE? And how does it match up against older X58 hardware, like the extremely popular i7-920? These questions will be answered as Benchmark Reviews brings you all the details on the Core i5-2500K.
Manufacturer: Intel Corporation Full Disclosure: The product sample used in this article has been provided by Intel Corporation. Sandy Bridge FeaturesIntel Microarchitecture
Processor Graphics
Intel Turbo Boost Technology 2.0
Up to 8MB Shared Intel Smart Cache
Integrated Memory Controller
Halogen Free
Supported Technologies
Core i5-2500K Specifications
Closer Look: Core i5-2500KThe launch of Sandy Bridge brings a whole new architecture to the table. This was also the case when Clarkdale/Arrandale launched and we moved to the LGA1156 socket. With those processors, the 32nm CPU and 45nm GPU were combined onto the same die. This was the first large-scale release of on-die graphics, though they had been talked about for a long time. In the Sandy Bridge architecture, the on-die GPU moves to a 32nm process as well and the IMC moves onto the die. The smaller process lets Intel use roughly the same amount of space for the new Sandy Bridge die, but it can fit a lot more transistors.
Another difference is that Sandy Bridge CPUs will come with one of two versions of the on-die graphics; either the 2000 or the 3000 version. Also, while Clarkdale/Arrandale processors took advantage of turbo boost, Sandy Bridge boasts turbo boost usage for the GPU independent of the CPU. This is good news for gamers who choose to use the on-die graphics. If their game, like most, requires heavy GPU loads, the turbo boost can leave the CPU at stock speeds and clock up the GPU to improve performance. The Intel Core i5-2500K comes equipped with a 6MB shared L3 cache. Since the Nehalem architecture, where each core had a dedicated 2MB of cache, Intel has gone towards shared cache. In the Sandy Bridge architecture, the L3 cache is not only shared across all four cores of the Core i5-2500K, but also with the integrated GPU.
Another interesting cache feature of the Sandy Bridge CPUs is a tiny L0 instruction cache integrated with the L1 cache known as a Decoded Uop Cache. This caches instructions as they are decoded without discrimination. All decoded instructions run through this ~6KB cache. Old instructions are overwritten as new ones come along. This cache is designed to help alleviate front-end operation for commonly performed tasks.
Intel has also added a physical register file to their Sandy Bridge CPUs. AMD has talked about doing the same thing and it's a product of the growing size of Out-Of-Order execution hardware. It's grown quite a bit and with the inclusion of Advanced Vector Extensions (AVX) in the Sandy Bridge, the size of operands running through the OoO hardware would have become 256-bits. So Intel decided to put in a physical register file that will store the micro-op operands allowing the OoO to only carry pointers, rather than the data itself. Additionally, Sandy Bridge keeps the AES-NI introduced in Westmere and enhances Large Number Arithmetic Throughput. We should notice this in our CPU tests, especially the arithmetic processor tests and compression tests.
In order to speed up processing, Intel changed the way the parts of the die interconnect by creating what they call a ring bus. Older Intel architecture had each core with a separate path to the L3 cache. This isn't so bad if every core has its own cache, but when you have a shared cache between all the cores and the GPU, there isn't a lot of room to give everything its own path. The ring bus works basically how it sounds and each component of the die has its own location along the bus. Doing things this way gives the Sandy Bridge L3 cache great bandwidth and helps to reduce latency.
With everything on die now, we are moving away from older terminology like "Northbridge". The Sandy Bridge CPUs have a "System Agent" that controls the 16 PCI Express Lanes (they can be split into two x8 lanes), the DMI interface, the Memory Controller, and the Display Engine. The system agent also has a power control unit that controls all power management and reset functions, has separate voltage and frequency than the CPU cores, and offers power and thermal management for the PCI Express.
CPU Testing and ResultsBefore I begin any benchmarking or overclocking, I thoroughly stress the CPU and memory by running Prime95 on all available cores for 12 hours. If no errors are found, I move on to a gaming stress test. To do this, I use Prime95 again to stress the processor, while running an instance of FurMark's stability test on top of this. If the computer survives this test for 2 hours without lockup or corruption, I consider it to be stable and ready for overclocking. After achieving what I feel is stable overclock, I run to these tests again for certainty. The goal of this stress testing is to ensure the clock speeds and settings are stable before performing any benchmarks. I adopted this method from another writer here at Benchmark Reviews and it seems to do a great job of flushing out what only seem to be stable overclocks.
Once the hardware is prepared, we begin our testing. Each benchmark test program begins after a system restart, and the very first result for every test will be ignored since it often only caches the test. This process proves extremely important in the many gaming benchmarks, as the first run serves to cache maps allowing subsequent tests to perform much better than the first. Each test is completed five times, with the average results displayed in our article. We are going to focus here mainly on comparing the test results from the Core i5-2500K against the Phenom-II X4-975BE since that is the AMD processor that the Core i5-2500K relates to most closely in price. Also, it will be interesting to see the results of the Core i5-2500K in comparison to the Core i7-920, an extremely popular Nehalem CPU that still holds a large market share. Intel H67 Test Platforms
Intel X58 Test Platform
AMD 890GX Test Platform
Benchmark Applications
AIDA64 Extreme Edition Benchmark TestsIn November, 2010, FinalWire acquired and discontinued Lavalys EVEREST, updated it, and released it as AIDA64. AIDA64 is an industry leading system diagnostics and benchmarking solution for enthusiasts PC users, based on the award-winning EVEREST Technology. During system optimizations and tweaking it provides essential system and overclock information, advanced hardware monitoring and diagnostics capabilities to check the effects of the applied settings. CPU, FPU and memory benchmarks are available to measure the actual system performance and compare it to previous states or other systems. Furthermore, complete software, operating system and security information makes AIDA64 a comprehensive system diagnostics tool that offers a total of 100 pages of information about your PC. All of the benchmarks used in our test bed rely on basic x86 instructions and consume very low system memory while also being aware of HyperThreading, multi-processors, and multi-core processors. While the AIDA64 CPU tests really only compare the processor performance more than it measures platforms, it still offers a glimpse into what kind of power each platform possesses. Queen and Photoworxx tests are synthetic benchmarks that operate the function many times and over-exaggerate by several magnitudes 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.
The Intel Core i5-2500K comes out swinging in our AIDA64 Queen Test suite, topping the charts. Not only does the show a 22% better performance than the similarly priced AMD Phenom-II X4-975BE, but it also narrowly edges out the aging Core i7-920 Nehalem processor. The gains over the i7-920 are only just over 1%, so it's ok to say that the two are on par here. 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. The AIDA64 Photoworxx benchmark performs the following tasks on a very large RGB image:
I have noticed over time that the Photoworxx test, unlike most of the other AIDA64 tests, depends a lot on the L3 cache. As you can see from the AMD CPU results, the CPUs without an L3 cache perform worse than those with an L3 cache. Still, while pulling ahead strongly in the Queen tests, the Core i5-2500K falls behind here in the Photoworxx tests. Seeing as how Intel is marketing the Sandy Bridge lineup as media-centric, a lack of performance here is disappointing. The Core i5-2500K falls behind the i7-920 by over 34.5% and behind the Phenom-II X4-975BE by almost 9%. 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 footprint independent of input data that can be reduced at some cost in compression.
Compression is an area in which newer Intel CPUs have been benefitting from new design techniques. Although AES has been the major boost, zip functions have been improved as well. We can see this clearly from the results as the Core i5-2500K beats the Phenom-II X4-975BE by about 5% and the i7-920 by about 8%. 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. 
While I normally like to put both of the Everest integer performance tests on one graph, the Core i5-2500K made that impossible this time. With the new Sandy Bridge and Clarkdale/Arrandale series of processors, Intel made some major changes to the way their CPUs handle AES compression. This new processing is a boon to webmasters everywhere, as well as anyone who deals with compressed files on a regular basis. With that in mind, the Core i5 processor completely destroys the competition in the AES test, boasting gains of 281% over the Phenom-II X4-975BE and 339% over the i7-920. 
The floating point tests show us a very strong performance from the Core i5-2500K as well. All three tests, the 32-bit, 64-bit, and 128-bit keep almost exactly the same level of increase over the i7-920, lending validity to the tests. All three floating point numbers are increased between 21% and 23% when using the new Sandy Bridge CPU over the i7-920. Similar gains are seen when compared to the Phenom-II X4-975BE, with 24% gains in the 32-bit Julia tests and 19% gains in the 64-bit Mandel tests. When moving into the 128-bit floating point SinJulia tests, the Core i5-2500K realizes a 100% gain in performance over the Phenom-II X4-975BE. Passmark Performance TestPassMark Performance Test is a PC hardware benchmark utility that allows a user to quickly assess the performance of their computer and compare it to a number of standard 'baseline' computer systems. The Passmark Performance Test CPU tests all benchmark the mathematical operations, compression, encryption, SSE, and 3DNow! instructions of modern processors. In our tests there were several areas of concentration for each benchmark, which are combined into one compound score. This score is referred to as the CPU Mark, and is a composite of the following tests: Integer Math, Floating Point Math, Find Prime Numbers, SSE/3DNow!, Compression, Encryption, Image Rotation, and String Sorting. For this review, we've also decided to run the memory benchmark, which results in a composite score based on the following tests: small block allocation, cached read, uncached read, write performance, and large block allocation.
In the Passmark tests, the Intel Core i5-2500K once again outperforms its AMD rival, the Phenom-II X4-975BE and this time by almost 9% in the CPU tests. In the memory tests, the gains are closer to 38%. This is likely due, at least in part, to the integration of the memory controller onto the die for the Sandy Bridge CPU. This will allow faster communication with the memory. The same memory was used for both systems. The only system with different memory, the X58 system, outperforms the Core i5-2500K by about 12% in the CPU tests and 14% in the memory tests. PCMark Vantage Benchmark TestsPCMark Vantage is an objective hardware performance benchmark tool for PCs running 32- and 64-bit versions of Microsoft Windows Vista or Windows 7. PCMark Vantage is 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 demonstrate simulated 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 relative PCMark Vantage is as "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) has 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 PCMark Vantage test are all over the place. This is likely due to the fact that much of the testing relies on components other than the CPU. While we kept the test systems as similar as possible, because of the different hardware required, it wasn't possible to only change the CPU. Interestingly enough, the Core i5-2500K outperformed the i7-920 by a consistent margin of about 3.5% in the TV and Movies suite, the Gaming suite, and in the overall PCMark score. In the music tests, it fails to overcome the i7-920. The same is not true with performance in comparison to the Phenom-II X4-975BE. The Core i5-920 marks gains of between 8.5% and almost 38%. The scores vary widely between AMD CPUs. Remember to consider the editor's note from the beginning of the PCMark Vantage tests carefully.SiSoftware Sandra Benchmark TestsSiSoftware Sandra (the System ANalyser, Diagnostic and Reporting Assistant) is an information & diagnostic utility. It should provide most of the information (including undocumented) you need to know about your hardware, software and other devices whether hardware or software. It works along the lines of other Windows utilities, however it tries to go beyond them and show you more of what's really going on. Giving the user the ability to draw comparisons at both a high and low-level. You can get information about the CPU, chipset, video adapter, ports, printers, sound card, memory, network, Windows internals, AGP, PCI, PCI-X, PCIe (PCI Express), database, USB, USB2, 1394/Firewire, etc. The SANDRA DhryStone and Whetstone tests are CPU tests that run completely within the CPU + cache memory itself. These tests are perfect for seeing general efficiency per processing core. Dhrystone is basically a suite of arithmetic and string manipulating programs and is an older CPU tests. Even so, it remains a simple and accurate way to show RAW CPU processing performance. The whetstone benchmark primarily measures floating-point arithmetic performance. 
The two SiSoftware Sandra tests we used are both Arithmetic tests. As we mentioned before, Intel put a lot of effort into revamping the way the Sandy Bridge CPUs handle arithmetic processes. We should see their effort pay off here. The Dhrystone test suite results in quite a large gap in performance between the Core i5-2500K and the Phenom-II X4-975BE with the Sandy Bridge CPU holding more than a 65% advantage over the AMD CPU. When compared to the Core i7-920, the i5-2500K holds onto a lead of almost 4% in the Dhrystone tests. The Whetstone tests put the Phenom-II X4-975BE a lot closer to the i5-2500K, this time only trailing by about 18%. Alternately, the Core i5-2500K fails to supersede the Core i7-920 in the Whetstone tests, trailing by about 16%.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 uses all of a system's processing power to render a photorealistic 3D scene, "No Keyframes" the viral animation by AixSponza. This scene makes use of various algorithms to stress all available processor cores. The OpenGL graphics card testing procedure uses a complex 3D scene depicting a car chase with which the performance of your graphics card in OpenGL mode is measured. During the benchmark tests the graphics card is evaluated by way of displaying an intricate scene that includes complex geometry, high-resolution textures, and a variety of effects to evaluate the performance across a variety of real-world scenarios.
In the CPU Single Core tests, Cinebench focuses on a single thread of processing, so all of the AMD CPU scores are based nearly completely on their clock speeds. You can see this clearly from the results if you exclude the Core i7 and i5 processors. The Core i7-920 sits about at the same point as the Phenom-II X2-560BE which runs at 3.3GHz. The similarly clocked Core i5-2500K, however, scores much higher than any CPU on the charts. This is undoubtedly due in part to the turbo boost ability of the Sandy Bridge CPU, which allows it to overclock a single core up to 3.7GHz if no other cores are active. The Phenom-II X4-975BE, while beating the Core i7-920 CPU, falls behind the i5-2500K by 37% in the Single Core test. In the Multi-Core Cinebench test the Intel Core i7 can take advantage of multi-threading so the Core i7's four cores turn into eight threads, giving it a higher score than all but the Core i5-2500K. The 2500K doesn't have hyperthreading capabilities, but still destroys even the 8-threaded Core i7-920. Since Cinebench focuses on isolating the CPU for testing, this multi-core test really shows the power of the new Sandy Bridge CPU. Gains over the Phenom-II X4-975BE are at almost 28%, with 12.5% gains over the hyperthreaded i7-920.Street Fighter IV BenchmarkPC-based video games can depend heavily on the CPU if the attached GPU (Graphics Processing Unit) is less powerful, or the graphics settings are configured so low that they create no strain on the video card and must rely purely on system processing speed; a phenomenon known as CPU-dependence. The opposite is true when the video game has a powerful video card installed, and can handle all graphical demands without receiving assistance from the CPU. Benchmark Reviews has proven consistently that, with a high end GPU in use, frame rates are not often noticeably impacted by changes in processor or RAM. Capcom's Street Fighter IV is part of the now-famous Street Fighter series that began in 1987. The 2D Street Fighter II was one of the most popular fighting games of the 1990s, and now gets a 3D face-lift to become Street Fighter 4. The Street Fighter 4 benchmark utility was released as a novel way to test your system's ability to run the game. It uses a few dressed-up fight scenes where combatants fight against each other using various martial arts disciplines. Feet, fists and magic fill the screen with a flurry of activity. Due to the rapid pace, varied lighting and the use of music this is one of the more enjoyable benchmarks. Street Fighter IV uses a proprietary Capcom SF4 game engine, which is enhanced over previous versions of the game. In terms of 3D graphical demand, Street Fighter IV is considered very low-end for most desktop GPUs. While modern desktop computers with discrete graphics have no problem playing Street Fighter IV at its highest graphical settings, integrated and mobile GPUs have a difficult time producing playable frame rates with the lowest settings configured. While PC games are generally playable regardless of CPU, the Street Fighter IV game is surprisingly dependent on the CPU. That is why it is included here.
It looks like Street Fighter IV really gives the advantage to the old Intel processor, with the Core i7-920 taking the cake here. You can see, though, how CPU performance makes a significant (though probably not recognizable by the human eye) difference in the frame rates of SFIV. Here, the Core i5-2500K is neck and neck with the Phenom-II X4-975BE. This gives the AMD CPU some credence as a gaming processor while showing that Intel was serious about their claim that they made the Sandy Bridge CPUs to be more media-centered processors.
|
|
Comments
Either way I am most likely going to get the i5 2500k since it is very strong. Also the price is rather cheap at $211 when currently the i5 750 is $200 and the i5 760 is about $209 dollars. Since I did not upgrade following the first i series I would need to get a new motherboard anyway.Integrated graphics means little to me since as a gamer I would get a higher end GPU regardless. I just have to make sure the motherboard is p67 not H67 for overclocking and such.
Anyway nice read, thanks for the article.