Core i7-980X CPU Review

[nextpage title=”Introduction”]

Intel will be launching very soon the world’s first six-core CPU for desktops, Core i7-980X (codenamed “Gulftown”), a socket LGA1366 processor running at 3.33 GHz, and we had the privilege of benchmarking this beast. Let’s see what kind of performance level it will deliver.

Core i7-980X is manufactured under the latest 32 nm manufacturing technology (this manufacturing process is codenamed Westmere). Internally this CPU still uses Nehalem microarchitecture (click here to learn about this microarchitecture). The next microarchitecture to be released (probably in 2011) by Intel is codenamed Sandy Bridge and the first CPUs based on it will also be manufactured under 32 nm process.

This CPU also has Hyper-Threading technology, which simulates an extra core on each CPU core. Therefore it is seen by the operating system and programs as a 12-core CPU. Holy Moly!

The official internal clock for this CPU is 3.33 GHz, but with TurboBoost technology enabled (which allows the CPU to automatically overclock itself when needed), internal clock rate can reach up to 3.60 GHz.

This CPU is based on socket LGA1366 like the first Core i7 units (models starting with “9”). This means that its integrated memory controller supports only DDR3 memories up to 1,066 MHz under triple-channel architecture. At first the support of only DDR3-1066 may be seen as a drawback since several mainstream CPUs support DDR3-1333. However the triple channel configuration more than overcomes this limitation, as it provides 50% more bandwidth for memory access. Some numbers for you. A CPU with two DDR3-1066 memory modules in dual-channel configuration has a 17,056 MB/s maximum theoretical transfer rate available for memory communication, while a CPU with three DDR3-1066 memory modules in triple-channel configuration has a 25,584 MB/s bandwidth available (50% increase). A CPU with two DDR3-1333 modules in dual-channel configuration has a maximum theoretical transfer rate of 21,328 MB/s available, so the triple-channel configuration with DDR3-1066 memories still has 20% more bandwidth than two DDR3-1333 modules under dual channel.

The L3 memory cache was also increased from 8 MB (on the other “9xx” Core i7 CPUs) to 12 MB on this model. This 12 MB cache is shared by all six cores.

Core i7-980X supports AES-NI instructions, a set of 12 new instructions to improve performance of  encryption and decryption operations (if the software makes use of these new instructions, of course). Like all other CPUs based on the Nehalem microarchitecture, this new Core i7 supports both SSE4.1 and SSE4.2 instruction sets.

Intel will ship Core i7-980X with a new stock CPU cooler, called DBX-B. This cooler will be only available with this CPU and won’t be sold separately. It uses the now popular tower design with four U-shaped copper heatpipes, very polished copper based (mirrored finishing) and a Quiet/Performance switch, so you can choose between automatic control or maximum speed all the times (and maximum noise level), respectively. Below you can see some pictures of this new cooler.

Figure 1: Intel DBX-B CPU cooler.

Figure 2: Intel DBX-B CPU cooler.

Figure 3: Intel DBX-B CPU cooler.

The price tag of this monster? The same as Core i7-975: USD 999 (price for distributors in the USA buying in 1,000-unit lots; the price for the end-user will be higher). It will be a very expensive CPU, but look at the bright side: it could be even more expensive.

Now let’s make a quick comparison between Core i7-980X and the other CPUs we included in our benchmarking.

[nextpage title=”The Tested CPUs”]

On the tables below you can see a comparison between the CPUs we included in our review. AMD CPUs do not support SSE4 instructions (they have a proprietary instruction set called SSE4a, which is not the same thing as SSE4). We only included one CPU from AMD because Phenom II X4 965 is the fastest CPU AMD has available and it is positioned at a lower price range (see table). AMD has announced that they will be also launching a six-core CPU this year, called Phenom II X6.

CPU	Cores	HT	Internal Clock	Turbo Clock	QPI or HT	Base Clock	Core	Technology	TDP	Socket	Price
Core i7-980X	6	Yes	3.33 GHz	3.60 GHz	6.4 GB/s	133 MHz	Gulftown	32 nm	130 W	1366	USD 999
Core i7-965	4	Yes	3.20 GHz	3.46 GHz	6.4 GB/s	133 MHz	Bloomfield	45 nm	130 W	1366	USD 999
Core i7-870	4	Yes	2.93 GHz	3.60 GHz	2 GB/s	133 MHz	Lynnfield	45 nm	95 W	1156	USD 562
Core i5-750	4	No	2.66 GHz	3.20 GHz	2 GB/s	133 MHz	Lynnfield	45 nm	95 W	1156	USD 196
Phenom II X4 965	4	No	3.4 GHz	–	8 GB/s	200 MHz	Deneb	45 nm	140 W *	AM3	USD 185

TDP stands for Thermal Design Power which advises the user of the maximum amount of heat the CPU can dissipate. The CPU cooler must be capable of dissipating at least this amount of heat.

* Newer models are coming with a TDP of 125 W. The tested model was from the older version, with a TDP of 140 W.

The prices listed are the official prices for distributors based on 1,000 quantities. The end-user price is higher than the prices listed.

CPU	L1 Cache	L2 Cache	L3 Cache	Memory Support	Memory Channels
Core i7-980X	32 KB + 32 KB per core	256 KB per core	12 MB total	DDR3 up to 1066 MHz	Three
Core i7-965	32 KB + 32 KB per core	256 KB per core	8 MB total	DDR3 up to 1066 MHz	Three
Core i7-870	32 KB + 32 KB per core	256 KB per core	8 MB total	DDR3 up to 1333 MHz	Two
Core i5-750	32 KB + 32 KB per core	256 KB per core	8 MB total	DDR3 up to 1333 MHz	Two
Phenom II X4 965	64 KB + 64 KB per core	512 KB per core	6 MB total	DDR3 up to 1333 MHz	Two

Socket LGA1366 CPUs talk to the external world (i.e., the chipset) through a bus called QuickPath Interconnect (QPI), which has the same goal as t
he HyperTransport bus used with AMD CPUs. For a detailed explanation on how QPI bus works, read our Everything You Need to Know About The QuickPath Interconnect (QPI) tutorial. Socket LGA1156 CPUs, however, use the DMI (Digital Media Interface) bus to talk to the chipset, which is the interface previously used to make the connection between the north bridge and the south bridge chips on Intel chipsets. At a first look this solution may seem worse than using the QPI bus, because the DMI interface provides a maximum transfer rate of 2 GB/s while QPI provides a maximum transfer rate of 4.8 GB/s or 6.4 GB/s, depending on the CPU. However, on socket LGA1156 the CPU has an integrated PCI Express 2.0 controller, so these CPUs talk directly to the main video card without using their external bus and without using the chipset.

Our tests have a known flaw. Socket LGA1366 Core i7 processors support triple-channel memory configuration and with them we used three 1 GB DDR3-1066 modules, so these CPUs had 3 GB available. With all other CPUs we used two 1 GB DDR3-1333 modules, so these CPUs had 2 GB available. Unfortunately due to the different memory configuration supported by each CPU, we had to decide which methodology to use, and we chose to use one that would provide the “best” memory configuration for the tested system.

[nextpage title=”How We Tested”]During our benchmarking sessions, we used the configuration listed below. Between our benchmarking sessions the only variable was the CPU being tested and the motherboard, which had to be replaced to match the different CPU sockets.

Hardware Configuration

• Motherboard (Socket LGA1156): MSI P55-GD85 (1.10 BIOS)
• Motherboard (Socket LGA1366): ASUS P6T Deluxe OC Palm Edition (1904 BIOS)
• Motherboard (Socket AM3): ASUS M4A79XTD EVO (0704 BIOS)
• CPU Cooler (Socket LGA1156): Intel stock
• CPU Cooler (Socket LGA1366): Intel DBX-B
• CPU Cooler (Socket AM3): AMD stock
• Memory (Socket LGA1366): Two Qimonda IMSH1GU03A1F1C-10F modules (DDR3-1066/PC3-8500)
• Memory (Sockets 1156 and AM3): Two 1 GB Crucial CT12864BA1339 modules (DDR3-1333/PC3-10600, CL9, 1.5 V), configured at 1,333 MHz
• Hard Disk Drive: Western Digital Caviar Black 1 TB (WD1001FALS, SATA-300, 7,200 rpm, 32 MB buffer)
• Video Card: EVGA GeForce GTX 285 FTW
• Video Monitor: Samsung Syncmaster 305T
• Power Supply: SilverStone Element ST75EF
• Optical Drive: Lite-On LH-20A1L

Operating System Configuration

• Windows 7 Ultimate 64-bit
• NTFS
• Video resolution: 2560×1600 @ 60 Hz

Driver Versions

• NVIDIA video driver version: 195.62
• Intel Inf chipset driver version: 9.1.1.1019

Software Used

• PCMark Vantage Professional 1.1.0
• VirtualDub 1.9.5 + MPEG-2 Plugin 3.1 + DivX 6.8.5
• Adobe Photoshop CS4 Extended + GamingHeaven Photoshop Benchmark V3
• Adobe After Effects CS4
• WinRAR 3.92
• Cinebench 11.5
• Call of Duty 4 – Patch 1.7
• Fallout 3 – Patch 1.7
• Crysis Warhead – Patch 1.1 + HOC Bench Crysis Warhead Benchmark Tool 1.1.1
• Far Cry 2 – Patch 1.03

Error Margin

We adopted a 3% error margin; thus, differences below 3% cannot be considered relevant. In other words, products with a performance difference below 3% should be considered as having similar performance.

[nextpage title=”PCMark Vantage”]

PCMark Vantage simulates the use of real-world applications and gives scores for the following categories:

• PCMark
• Memories
• TV and Movies
• Gaming
• Music
• Communications
• Productivity
• HDD

For a detailed description of each one of these tests, please download and read the PCMark Vantage Reviewer’s Guide.

You can see the results for each category below. We are not going to compare the results for the Memories and HDD suites.

Core i7-980X achieved an overall score 17% higher than Core i7-870, 20% higher than Core i7-965, 23% higher than Core i5-750 and 35% higher than Phenom II X4 965.

On the TV and Movies benchmark Core i7-980X achieved a score 10% higher than Core i7-965, 12% higher than Core i7-965, 12% higher than Core i7-870, 30% higher than Core i5-750 and 31% higher than Phenom II X4 965.

On the Gaming set Core i7-980X achieved a score 13% higher than Core i7-965, 16% higher than Core i7-870, 28% higher than Core i5-750 and 71% higher than Phenom II X4 965.

On the Music benchmark Core i7-980X achieved a score similar to Core i7-870 and 7% higher than Core i6-750 and 30% higher than Phenom II X4 965. Core i7-965 achieved a score 3% higher than the reviewed CPU.

On the Communications tests Core i7-980X achieved a score 62% higher than Core i7-965, 64% higher than Core i7-870, 68% higher than Phenom II X4 965 and 79% higher than Core i5-750.

And finally on the Productivity benchmark Core i7-980X achieved a score 5% higher than Core i7-965, 6% higher than Core i7-870, 16% higher than Phenom II X4 965 and 17% higher than Core i5-750.

[nextpage title=”VirtualDub + DivX”]

With VirtualDub we converted a full-length DVD movie to DivX format and saw how long it took for this conversion to be completed. The DivX codec is capable of recognizing and using not only more than one CPU (i.e., more than one core), but also the SSE4 instruction set (feature not available on the reviewed CPUs).

The movie we chose to convert was Star Trek – The Motion Picture: Director’s Cut. We copied the movie to our hard disk drive with no compression, so the final original file on our HDD was 6.79 GB. After compressing it with DivX, the final file was only 767.40 MB, which is quite remarkable.

The results below are given in seconds, so the lower the better.

On DivX encoding Core i7-980X was the fastest CPU, being 5% faster than Core i7-965, 9% faster than Core i5-750, 24% faster than Phenom II X4 965 and 27% faster than Core i7-870.

[nextpage title=”Photoshop CS4″]

The best way to measure performance is by using real programs. The problem, though, is creating a methodology using real software that provides accurate results. For Photoshop CS4, there is a methodology created by the folks at GamingHeaven that is very accurate. Their script applies a series of 15 filters to a sample image, and we wrote down the time taken for each filter to run. At the end, we have the results for each individual filter and we simply added them up to have the total time taken to run the 15 filters from the GamingHeaven batch. The results below are given in seconds, so the lower the number the better.

On Photoshop CS4 Core i7-980X was again the fastest CPU, being 4% faster than Core i7-965, 12% faster than Core i7-870, 21% faster than Phenom II X4 965 and 23% faster than Core i5-750.

[nextpage title=”After Effects CS4″]

After Effects is a very well-known program for video post-production that is used to add animation and visual effects in videos. To evaluate the performance of each CPU running this program, we ran a workload consisting of 25 compositions that applied several filters and effects to a variety of input file types such as PSD (Photoshop), AI (Illustrator), EPS, and TIF. After each filter was applied, the composition was rendered to an uncompressed AVI file with the same resolution as the input files. The results below are the time each CPU took to finish the whole batch, given in seconds, so the lower the number the better.

On After Effects CS4 Core i7-980X was the fastest CPU and here we can clearly see the difference in performance the extra cores can bring on an application that can take advantage of them: the reviewed CPU was 28% faster than Core i7-965, 30% faster than Core i7-870, 47% faster than Core i5-750 and 59% faster than Phenom II X4 965.

[nextpage title=”WinRAR”]

We measured the time each CPU took to compress five high-resolution 48-bit uncompressed TIF images, each one with around 70 MB, to RAR format with the popular WinRAR application. The results below are given in seconds, so the lower the number the better.

Core i7-980X was again the fastest CPU, being 11% faster than Core i7-965, 15% faster than Core i7-870, 23% faster than Core i5-750 and 41% faster than Phenom II X4 965.

[nextpage title=”Cinebench 11.5″]

Cinebench 11.5 is based on the 3D software, Cinema 4d. It is very useful to measure the performance gain given by having more than one CPU installed on the system when rendering heavy 3D images. Rendering is one area in which having more than one CPU helps considerably, because usually, rendering software recognizes several CPUs. (Cinebench, for instance, can use up to 16 CPUs.)

Since we were interested in measuring the rendering performance, we ran the test called “Rendering x CPUs,” which renders a “heavy” sample image using all available CPUs (or cores – either real or virtual, as on CPUs with Hyper-Threading technology, each core is recognized as two cores by the operating system) to speed up the process.

On Cinebench we could see a big difference in performance between Core i7-980X and the other CPUs: it was 56% faster than Core i7-965, 63% faster than Core i7-870, 122% faster than Phenom II X4 965 and 172% faster than Core i5-750.

[nextpage title=”Call of Duty 4″]

Call of Duty 4 is a DirectX 9 game implementing high-dynamic range (HDR) and its own physics engine, which is used to calculate how objects interact. For example, if you shoot, what exactly will happen to the object when the bullet hits it? Will it break? Will it move? Will the bullet bounce back? It gives a more realistic experience to the user.

We ran this game under 2560×1600, maxing out all image quality controls (i.e., everything was put on the maximum values on the Graphics and Texture menus). We used the game internal benchmarking feature, running a demo provided by NVIDIA called “wetwork.” We are putting this demo for downloading here if you want to run your own benchmarks. The game was updated to version 1.7. We ran this test five times, discarding the lowest and the highest scores. The results below are an arithmetic average of the three remaining values, given in frames per second (FPS).