What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

Intel Haswell-E i7 5960X Review

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
Only 18 months ago, Intel’s enthusiast platform was on the chopping block. While X79 and Sandy Bridge-E did well and Ivy Bridge-E was on the horizon, the desktop PC market was in a downturn and the continuity of interest for ultra-high-end platforms was in doubt. And yet here we are reviewing Haswell-E and Intel’s first 8-core 16-thread enthusiast processor, the Core i7-5960X. So what happened? A resurgence in the desktop space has created a rapid succession of roadmap updates, adding processors like Devil’s Canyon, the Pentium Anniversary Edition and a now full Haswell-E lineup.


While Intel may have been a bit cautious in bringing Haswell-E to the market, there’s no denying that we have seen a relatively constant progression in terms of technology, if not performance. However, due in no small part to a lack of true competition from AMD, the actual speed of change has slowed down a bit. While the trademark tick / tock process is still very much alive we are no longer seeing massive performance uplifts from one generation to the next. Rather, Intel is focusing on both processing and TDP efficiency while also optimizing their instructions per clock ratio.

In many ways the original Haswell architecture moved things to another level in nearly every respect and Haswell-E alongside the upcoming server-oriented Haswell-EP will capitalize on its internal changes. Naturally, those original Haswell processors will remain around at lower price points since Haswell-E is supposed to offer something they can’t: a ridiculous amount of parallel processing horsepower and a platform that’s been designed for high end computing.


While the Haswell-E processors are still based on the same internal architecture and utilizes the same 22nm Tri-Gate 3D transistor manufacturing process as Intel’s other CPUs like the i7-4790K and i7-4770K, there are some fundamental differences between these two lineups. First and foremost is the move away from an internal graphics co-processor which took up a good amount of die space on Haswell and the addition of an advanced quad channel DDR4 memory controller. With that being said, if you are wondering what other changes have been rolled into Haswell-E in comparison to Ivy Bridge-E, look no further than here.

The process of ditching the graphics engine may have saved some on-die space but that doesn’t mean Haswell-E is a small processor. On the contrary. With its eight cores and massive 20MB of L3 cache, the i7-5960X weighs in at 2.6 billion transistors. Compare and contrast this with an i7-4770K’s 1.4 billion and you can begin to understand how complex a chip Intel have created.


Bringing Haswell-E to the desktop while maintaining a reasonable TDP was certainly a challenge which is obvious given the clock speeds of these processors. In short, they lag behind (in some cases significantly) the SKUs they replace but we can’t forget that the Haswell architecture does feature noteworthy IPC improvements over IVB-E so some of the shortfall will be made up by the CPU’s back end.

At the top of this new lineup sits the $1000 i7-5960X which is Intel’s 8-core, 16-thread processing monster which also comes with a full 20MB of cache allotment. Unfortunately, in order to hit a 140W TDP its frequencies had to be dialed down to 3.0GHz and 3.5GHz for the Base and Turbo clocks respectively. This means on paper at least it will be fighting an uphill battle against the i7-4960X and perhaps even the i7-3960X, in any application or game that can’t access all 16 threads.

The i7-5930K will likely be the CPU most Haswell-E buyers gravitate towards since it seems to offer a good blend of clock speeds and the potential to run twelve concurrent processing threads. With six physical cores and a base clock of 3.5GHz, we wouldn’t be surprised to see it come close to its big brother in some multi-threaded applications as well. When compared against the outgoing i7-4930K the i7-5930K stacks up quite well with similar frequencies alongside Haswell’s numerous processing improvements.

Finally there’s the i7-5820K which is a bit of an oddball addition since it plays the part of a bridge between Haswell-E and the less expensive Devil’s Canyon i7-4790K. While it boasts six physical cores and the potential to process up to twelve threads (a definite benefit over the i7-4820K’s 4/8 configuration), clock speeds are about 10% lower than the comparable IVB-E part. Then there are the 28 PCI-E lanes which had us scratching our collective heads since this is a layout which won’t allow buyers to run two graphics cards at full x16 speeds. Without that capability and unless having those 12 threads is an absolute necessity, gamers may as well turn towards the Z97 platform which is less expensive and whose higher end processors will run circles around Haswell-E in games. More on that later but for the time being we have to question the elimination of one of Haswell-E’s primary selling points for enthusiasts.


Haswell-E’s frequencies and specifications will likely comes as a surprise for those who are accustomed to the raw clock speeds offered by the likes of Devil’s Canyon and even Ivy Bridge-E but there’s more here than what first meets the eye. This platform is meant for professionals, gamers who also want to multi task and anyone who appreciates getting tasks done quicker.

One interesting thing we did notice during testing was the i7-5960X’s apparent lack of a Base and Turbo mode. Regardless of how many threads we threw at it (from one all the way to 16 concurrent threads), it remained steadfastly at 3.3GHz with some millisecond fluctuations to 3.5GHz. This is basically the way every Haswell chip behaves but in this case it seems there is very little overhead, even when the CPU is working on a single thread.

Truth be told, the days of single threaded applications are almost over, even on the gaming side. Microsoft has already demonstrated how DirectX 12’s dynamic resource distribution will be better able to utilize multi core processors for physics, AI and general GPU communications. With Intel being a significant member of the DX12 planning board, you can bet that Haswell-E will be well positioned for the future of gaming even though its specs may not show it now. Plus, the i7-5960X has been designed for power users above all else and that means tasks like decrypting, transcoding and rendering will be its primary uses.
 
Last edited by a moderator:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
The X99 Platform; Enthusiasts Rejoice

The X99 Platform; Enthusiasts Rejoice


One of the main critiques leveled at Intel’s X79 was its similarity to the old-as-the-hills X58. As a matter of fact, from a specifications standpoint, that’s exactly what it was: an X58 chipset with a new coat of paint in the form of PCI-E 3.0 support. Since it didn’t feature current technologies like native USB 3.0 and only had two SATA 6Gbps ports, motherboards required third party controllers to attain those functions, and support wasn’t the greatest especially for key features like RAID and high speed USB throughput. That caused a serious problem for a so-called enthusiast platform when Intel’s own Z87 incorporated those elements into boards that often cost hundreds less than their X79 cousins.

X99 changes this equation in a big way towards compatibility that many thought should have been incorporated into X79 in the first place. Nonetheless, we are now (finally!) going to see native support for USB 3.0, SATA Express, and Thunderbolt 2 on Intel’s enthusiast motherboards.


Starting with the most obvious thing first: X99 chipsets will still use the LGA2011 socket but it has been updated for Haswell-E compatibility. This not only means new microcode but also support for the processors’ fine grain power distribution needs and higher current capability. In short, older LGA2011 boards will not be forwards compatible with these new processors, nor will this so-called LGA2011-v3 socket be backwards compatible with Ivy Bridge-E CPUs.

The X99 platform is of course headlined by the Haswell-E CPU which provides up to 40 PCI-E 3.0 lanes (the i7-5820K will only come with 28 lanes enabled) which can be distributed via up to three integrated slots. This means a x16 / x16 / x8 setup is possible as is a 5x8 setup via third party controllers should motherboard vendors decide to go that route. The processor also houses the quad channel DDR4 memory controller.

As with all Intel platforms, the PCH is where all the I/O fun happens and it is connected to the processor via a x4 DMI interface providing up to 4GB/s of aggregate upstream / downstream bandwidth. In this case the X99 supports up to 14 USB ports spread across six USB 3.0 and 8 USB 2.0 along with ten native SATA 6Gbps ports. Through the use of Intel’s refreshed architecture these can be paired with additional PCIe 2.0 lanes for SATA Express or 4x M.2 compatibility without needing to resort to a so-called “FlexIO” interface. Naturally, those lanes can also be used for additional controllers as well which typically provide Bluetooth, secondary LAN and WiFi features.

Past the obvious continuity of an integrated Intel LAN, all of the SATA 6Gbps ports are backstopped by Intel’s RST 13.1 infrastructure should a motherboard vendor choose to include it (most will be). Extreme Tuning Utility compatibility is also a requirement here whereas on Z97 it’s considered an optional feature.


Since this is considered Intel’s high end platform, motherboard manufacturers are pulling out all the stops when it comes to board design. Take the ASUS X99-Deluxe for example; it features a laundry list of must-have features for enthusiasts. It has 3-way SLI / Crossfire, two SATA Express ports, an add-in Thunderbolt II card, a x4 M.2 storage slot, a high end sound solution with Nichicon Muse caps, integrated AC wireless support and the list goes on.

X99 boards are supposed to be the best around and the Deluxe’s $399 price reflects exactly that. However, when the cost of DDR4 memory is also factored into the equation, upgrading to Haswell-E may be rewarding but it won’t be an inexpensive proposition.
 
Last edited by a moderator:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
Test Setups & Methodology

Test Setups & Methodology


For this review, we have prepared a number of different test setups, representing many of the popular platforms at the moment. As much as possible, the test setups feature identical components, memory timings, drivers, etc. Aside from manually selecting memory frequencies and timings, every option in the BIOS was at its default setting.


For all of the benchmarks, appropriate lengths are taken to ensure an equal comparison through methodical setup, installation, and testing. The following outlines our testing methodology:

A) Windows is installed using a full format.

B) Chipset drivers and accessory hardware drivers (audio, network, GPU) are installed.

C)To ensure consistent results, a few tweaks are applied to Windows 7 and the NVIDIA control panel:
  • UAC – Disabled
  • Indexing – Disabled
  • Superfetch – Disabled
  • System Protection/Restore – Disabled
  • Problem & Error Reporting – Disabled
  • Remote Desktop/Assistance - Disabled
  • Windows Security Center Alerts – Disabled
  • Windows Defender – Disabled
  • Screensaver – Disabled
  • Power Plan – High Performance
  • V-Sync – Off

D) Windows updates are then completed installing all available updates

E) All programs are installed and then updated.

F) Benchmarks are each run three to eight times, and unless otherwise stated, the results are then averaged.

G) All processors had their energy saving options / c-states enabled
 

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
System Benchmarks: AIDA64 / Cinebench R15

System Benchmarks


In this section, we will be using a combination of synthetic benchmarks which stress the CPU and system in a number of different domains. Most of these tests are easy to acquire or are completely free to use so anyone reading this article can easily repeat our tests on their own systems.

To vary the results as much as possible, we have chosen a selection of benchmarks which focus upon varied instruction sets (SSE, SSE3, 3DNow!, AVX, etc.) and different internal CPU components like the floating point units and general processing stages.



AIDA64 Extreme Edition


AIDA64 uses a suite of benchmarks to determine general performance and has quickly become one of the de facto standards among end users for component comparisons. While it may include a great many tests, we used it for general CPU testing (CPU ZLib / CPU Hash) and floating point benchmarks (FPU VP8 / FPU SinJulia).


CPU ZLib Benchmark

This integer benchmark measures combined CPU and memory subsystem performance through the public ZLib compression library. CPU ZLib test uses only the basic x86 instructions but is nonetheless a good indicator of general system performance.




CPU Hash Benchmark

This benchmark measures CPU performance using the SHA1 hashing algorithm defined in the Federal Information Processing Standards Publication 180-3. The code behind this benchmark method is written in Assembly. More importantly, it uses MMX, MMX+/SSE, SSE2, SSSE3, AVX instruction sets, allowing for increased performance on supporting processors.




FPU VP8 / SinJulia Benchmarks

AIDA’s FPU VP8 benchmark measures video compression performance using the Google VP8 (WebM) video codec Version 0.9.5 and stresses the floating point unit. The test encodes 1280x720 resolution video frames in 1-pass mode at a bitrate of 8192 kbps with best quality settings. The content of the frames are then generated by the FPU Julia fractal module. The code behind this benchmark method utilizes MMX, SSE2 or SSSE3 instruction set extensions.

Meanwhile, SinJulia measures the extended precision (also known as 80-bit) floating-point performance through the computation of a single frame of a modified "Julia" fractal. The code behind this benchmark method is written in Assembly, and utilizes trigonometric and exponential x87 instructions.





CineBench R15 64-bit


The latest benchmark from MAXON, Cinebench R15 makes use of all your system's processing power to render a photorealistic 3D scene using various different algorithms to stress all available processor cores. The test scene contains approximately 2,000 objects containing more than 300,000 total polygons and uses sharp and blurred reflections, area lights and shadows, procedural shaders, antialiasing, and much more. This particular benchmarking can measure systems with up to 64 processor threads. The result is given in points (pts). The higher the number, the faster your processor.

 

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
System Benchmarks: Civ V / PCMark 8

System Benchmarks (pg.2)



Civilization V: Gods & Kings Unit Benchmark


Civilization V includes a number of benchmarks which run on the CPU, GPU or a combination thereof. The Unit Benchmark simulates thousands of units and actions being generated at the same time, stresses multi core CPUs, system memory and GPU We give the non-rendered score below as it is more pertinent to overall CPU performance within the application.



PCMark 8


PCMark 8 is the latest iteration of Futuremark’s system benchmark franchise. It generates an overall score based upon system performance with all components being stressed in one way or another. The result is posted as a generalized score. In this case, we didn’t use the Accelerated benchmark but rather just used the standard Computational results which cut out OpenCL from the equation.

 

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
System Benchmarks: WPrime

WPrime


wPrime is a leading multithreaded benchmark for x86 processors that tests your processor performance by calculating square roots with a recursive call of Newton's method for estimating functions, with f(x)=x2-k, where k is the number we're squaring, until Sgn(f(x)/f'(x)) does not equal that of the previous iteration, starting with an estimation of k/2. It then uses an iterative calling of the estimation method a set amount of times to increase the accuracy of the results. It then confirms that n(k)2=k to ensure the calculation was correct. It repeats this for all numbers from 1 to the requested maximum. This is a highly multi-threaded workload. Below are the scores for the 32M and 1024M benchmarks.


 

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
System Benchmarks: Single Thread Performance

System Benchmarks: Single Thread Performance


Even though most modern applications have the capability to utilize more than one CPU thread, single threaded performance is still a cornerstone of modern CPU IPC improvements. In this section, we take a number of synthetic applications and run them in single thread mode.



 

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
Productivity Benchmarks: 7-Zip / MediaCoder

7-Zip


At face value, 7-Zip is a simple compression/decompresion tool like popular applications like WinZip and WinRAR but it also has numerous additional functions that can allow encryption, decryption and other options. For this test, we are avoiding its built-in benchmark and once again only focus upon real world testing by compressing a 2.6GB folder of various files and adding an AES-256 encryption layer for good measure. The test is timed until it is complete.



MediaCoder x64


Due to the varying compatibility of certain mobile devices, video transcoding performance has become something of a big deal. Transcoding allows one type of video / audio file to be converted into a different format and it typically takes up a huge amount of system resources. The MediaCoder application brings multi format transcoding to an accessible level with numerous options and acceleration for Intel’s QuickSync and NVIDIA’s CUDA technologies. In addition, its CPU support allows for full multi core utilization. In this test, we use the MediaCoder i-devices edition to convert a 600MB AVCHD file to an iPhone 4S friendly MPEG-4 format.

 

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
Productivity Benchmarks: Photoshop CS6 / POV Ray 3.7

Productivity Benchmarks (pg.2)



Adobe Photoshop CS6


For the image editing portion of this section, we use Photoshop CS6 in coordination with a custom benchmark script. This script automates the application of 20 different image manipulation functions to a 120MB PNG image, acting as an excellent test of CPU power and memory bandwidth. For this test, we have disabled GPU acceleration so it won’t play a factor in the areas where it would typically be used. We use Photoshop’s built-in timing feature to provide a result at each test stage.



POV Ray 3.7 RC6


POV Ray is a complex yet simple to use freeware ray tracing program which has the ability to efficiently use multiple CPU cores in order to speed up rendering output. For this test, we use its built-in benchmark feature which renders multiple passes of a high definition scene. In order to get the most accurate results, the second pass of the first test is logged, resulting in a benchmark score showing the average amount of pixels rendered per second.

 

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
13,421
Location
Montreal
Productivity Benchmarks: TrueCrypt 7.1 / x264HD

Productivity Benchmarks (pg. 3)



TrueCrypt 7.1


Truecrypt is another freeware gem which allows for on-the-fly disk encryption. More importantly, it fully supports AES-256 encryption methods and multi core processors. For this test, we used the built-in benchmark tool are logged the data throughput for TrueCrypt’s AES-256 encryption method.



x264HD Benchmark


x264 is quickly becoming the new codec of choice for encoding a growing number of H.264/MPEG-4 AVC videos. Think of it as the new Divx of HD and you can understand why we felt it critical to include. Tech Arp's recent development of the x264 HD Benchmark takes a 30 second HD video clip and encodes it into the x264 codec with the intention of little to no quality loss. The test is measured using the average frames per second achieved during encoding, which scales with processor speed and efficiency. The benchmark also allows the use of multi-core processors so it gives a very accurate depiction of what to expect when using encoding application on a typical full length video. We use the second pass of the first test for this benchmark as it fully loads all multi core processors.

 

Latest posts

Twitter

Top