Intel Lynnfield Core i5-750 & Core i7-870 Processor Review

[MAC]

Intel Lynnfield Core i5-750 &
Core i7-870 Processor Review

Over the past year, we have seen several monolithic advances in the world of the processors we all use for our personal computers. First, the Nehalem architecture finally saw the light of day in the guise of the high-end Bloomfield i7 series of CPUs and despite those processors being priced relatively high, they found a footing among enthusiasts and professionals alike. In answer to this widening gap between Intel’s processors and their own, AMD released their revamped Phenom II processors which have taken the market by storm. Now, as this year draws to a close, we have another release to contend with: that of the Lynnfield series mainstream processors from Intel. More than literally any single year over the last decade of personal computer advances, this year has truly has all the makings of the beginning of a CPU renaissance.

So what is Lynnfield? Actually, that question should rather be “what’s in a name?” Lynnfield processors are part of the Nehalem architectural family and as such will carry both the i5 and the i7 designations depending on where they land on both the performance scale and the list of features they include. The i7 800-series will naturally include higher clock speeds and features while others won’t. However, one thing will be constant: even though some Lynnfield processors will be branded as i7, they will only be compatible with P55 (and upcoming lower end) / LGA 1156 motherboards. Bloomfield’s X58 is essentially off-limits for this new family of processors. Other than that, you will see in this review that Lynnfield in general could also be considered the industry’s worst-kept secret with early sales, leaked benchmarks and P55 motherboard previews galore marring what would have been an otherwise smooth launch.

In this review, we will take a somewhat long-winded look at both the i7-870 and the i5-750 processors along with a glance inside of the new P55 platform and everything this pairing will bring to the table. The best part about all of this for you the consumers is the fact that both the CPUs and the accompanying motherboards will be widely available from literally the moment you read this. There are also some pretty strong indications that the platform itself is already quite mature in the eyes of motherboard manufacturers and as such, you will see not only price-conscious products but also high-end, enthusiast-centric units as well.

With all these leaks over the past few months, people are naturally excited about Lynnfield and let us tell you right now: there is plenty of reason to be excited. As a mainstream processor family, everything points to this being a strong product but can the new i5 and i7 products make themselves a viable alternative for consumers looking for a budget-conscious entry into the Nehalem family? Let’s find out.

 

[MAC]

Lynnfield - Core i5 & Core i7

Lynnfield - Core i5 & Core i7

Now that the media embargo has ended - and it was a particularly painful embargo for us since the retail processors and motherboards were being sold left, right, and center - we can finally reveal (or least confirm) everything you have wanted to know about the immediately available new Core i5-700 series and Core i7-800 series processors, codenamed Lynnfield.

First and foremost, let's talk about the naming scheme, arguably the most controversial aspect of this new processor family. Generally speaking, most people agree that the inclusion of Lynnfield models into the Core i7 series will confuse a great number of consumers, and it has...if our forums are anything to go by. However, it is what we're stuck with at the moment, so let's try to break down the product segmentation as best as we can. Intel have chosen this naming scheme to help highlight the specific technologies that each processors series supports. Simply put, Core i7 models are eight-thread processors which feature both Hyper-Threading and Turbo Boost technology. The Core i5 models are four-thread processors which support Turbo Boost, but not Hyper-Threading. Lastly, we eventually expect some additional Core i3 and/or i5 models with two-cores, Hyper-Threading, Turbo Boost, and even an IGP. One thing that all these chips share in common though is the socket: LGA1156. Even though the "i7" denominator may make them look compatible, none of these processors will fit into an LGA1366 X58 motherboard.

Clear enough? Good, let's take a closer look at what these new chips have to offer.

Specifications

As you can see, the three initial launch models are somewhat similar, but there are notable differences based on the aforementioned i5/i7 segmentation. Fundamentally speaking, all three processors are based on the exact same core, but the i5-750 model has the Hyper-Threading (HT) feature disabled. The i5 model is further limited by the fact that it can only Turbo Boost up by four multipliers, while i7-860/870 have access to five additional multipliers, which equates to 666Mhz of core clock headroom. The i7-800 series models are not only HT-enabled, and have more aggressive Turbo Boost profiles, but they also benefit from a higher Uncore frequency, which means that their integrated memory controller (IMC) and L3 cache are operating faster than that of the i5-750 model. Having said that, while limited compared to its peers, the i5-750 does have a very attractive $199 price tag, especially compared to the ultra-pricey i7-870, which encroaches into i7-950 territory and defies its supposedly 'mainstream' roots. It is pretty clear to us that Intel is trying to channel consumers towards the i7-860 model, but we will have to see whether its performance warrants that 43% premium.


Packaging & Chips:

​

A new product launch requires a new packaging design, and both Core i5-700 series and Core i7-800 series processors will ship in this attractive new box (with appropriate i5/i7 designation, of course).

Bloomfield Cooler vs. Lynnfield Cooler - Click on image to enlarge​

Intel has designed a new stock cooler for Lynnfield and it is tiny! It is a typical Intel design, featuring an aluminium body with an integrated copper core, and push-pins as the mounting system. Evidently, they are quite confident in their 95W TDP rating, since this cooler is significantly smaller than the one that ships with the 130W Bloomfield Core i7 chips. This potentially bodes well for overclockers though, since these must be really cool running chips, and we will definitely testing to see how well this cooler manages to cool Lynnfield at full load.

The Stars: i5-750 on the left, i7-870 on the right - Click on image to enlarge​

To the untrained eye, Lynnfield processors look identical to Bloomfield chips, however the placement of the gold contact dots and the two 'wings' that protrude from the integrated heatspreader (IHS) are dead giveaways. Needless to say that these are both engineering samples, and thus the stepping is different from what you will find in the retail channel. However, the cores themselves are identical.

Click on image to enlarge​

As per the LGA1156 socket name, those are 1156 contact points, a decrease from the LGA1366 Core i7 900 series, but still a huge increase from the 775 that are found on all Core 2 models. The layout of the micro SMD resistors is very interesting because it mimics the layout of the actual core, which you can see on the next page.

Bloomfield / Lynnfield / Core 2 Quad / Phenom II - Click on images to enlarge​

When placed side-by-side, you can clearly see how much smaller the Lynnfield chip is compared to the Bloomfield, and this despite the fact that Lynnfield actually has a slightly bigger core. The reason? Well, the LGA1366 package was designed with future headroom in mind, specifically larger six-core and maybe even eight-core dies. It is also worth noting that the Lynnfield chip is actually exactly the same size as the venerable Core 2 Quad.

Don't expect to see these stock values often with Turbo Boost enabled - Click on image to enlarge​

As usual, the CPU core speed is derived by a multiplier times bus speed formula. Since the FSB is no longer present, the bus speed in question is the base clock (BCLK), which has a stock frequency of 133MHz. As mentioned above, although our chips are engineering samples, they are manufactured with the final retail stepping, so they perform the same as the chips you will be able to buy in the retail channel.

i5-750 on the right, i7-870 on the left - Click on image to enlarge​

The integrated memory controller and the L3 cache operate on a seperate frequency called the Uncore clock (NB frequency in CPU-Z), which is derived by the uncore multiplier time the base clock (BLCK). The i5-750 model has a 16X uncore multiplier and a 2133Mhz Uncore frequency. The i7-860 and i7-870 chips have an 18X uncore multiplier and a 2400Mhz Uncore frequency. One very exciting change with Lynnfield is that the Uncore has now been isolated from the DDR3 memory frequency, which means that Lynnfield can potentially achieve much higher memory speeds since it is no longer artifically limited by the Uncore ratio.

Now let’s take an in-depth look at the Nehalem microarchitecture upon which these Lynnfield processors are based.

 

[MAC]

Microarchitecture Dissected

Microarchitecture Dissected

In 2007, Intel unveiled the Tick-Tock Model as a demonstration of the company's dedication towards continued rapid technological innovation. The "tick" is a shrinking of the previous architecture manufacturing process (65nm --> 45nm --> 32nm) and the "tock" is a new architecture. Since Penryn was a shrink and slight improvement of the preceding Core architecture, it was time for a brand new architecture and that is where Nehalem came in last November. Lynnfield remains a Nehalem part, since it is neither a new architecture, nor does it feature a smaller manufacturing process. Lynnfield is simply a new core variant based on the Nehalem architecture.

Bloomfield die on the left, new Lynnfield die on the right.​

On the left we have the current "Bloomfield" die, which was Intel's first native quad-core design, unlike Core 2 Quad models which are effectively two dual-core dies mounted in one CPU package. The Bloomfield and Lynnfield dies are obviously quite similar. Same number of cores, same size L1/L2/L3 cache size, system request queue (used for inter-core communication), similar integrated memory controller (IMC), etc. However, Lynnfield does away with Bloomfield's two QuickPath Interconnects (QPI), has one less 64-bit DDR3 memory channel, and most importantly features an industry-first integrated PCIe controller, which support 16 PCIe 2.0 lanes. Can you say ultra-low latency CPU-GPU communication?

In quantifiable terms, the die size of the Lynnfield processors is 296mm². By comparison, the Bloomfield measures 263mm², while the Core 2 Quad "Yorkfield" die is 214mm². Lynnfield clocks in at 774 million transistors, which is 43 million more than Bloomfield (731M.), but still less than Yorkfield, which had 820 million transistors due to its large and extremely transistor dense 12MB L2 cache. All three cores are manufactured using the same 45nm High K + metal gate transistor technology.

Now many of you are probably looking at the die pictures and saying "It is pretty but what am I looking at exactly?". A valid question, so let's take a look at the Lynnfield core layout:

​

As you can see, the integrated PCIe controller takes up a big chunk of the die, and it definitely explains why Lynnfield has 43 million more transistors than the Bloomfield core.

Part of the reason that Intel was able to design and manufacture Lynnfield so quickly is because the Nehalem architecture is dynamically scalable, and it was designed with modularity in mind. What this means is that Intel can custom create processors based on the needs of the market without having to go design a brand new chip from scratch. They can add or remove cores, L3 cache, number of QPI links, number of memory channels, type of memory supported, power management, and even integrated graphics. Therefore, Intel have the ability to add new blocks to the core without having to go to the drawing board and redesigning the whole layout. Basically, they are only limited by how much stuff they can actually fit on one CPU package. Think of it as a multi-million dollar Lego set.


In the following page, we will examine some of the more functional features and technologies that Intel have built into the Core i5-750 and i7-860/870 processors.

 

[MAC]

Microarchitecture Dissected Continued

Microarchitecture Dissected Continued

Despite sharing significant roots with the original P6 microarchitecture that was debuted in the Pentium Pro in 1995, and being fundamentally derived from 'Penryn', the Nehalem microarchitecture represents one of the most significant overhauls ever. Intel's engineers added significant performance-oriented features, like an integrated memory controller, a completely new system interconnect (QPI), and a multi-level shared cache, while still focusing a great deal on the chip's power efficiency capabilities.

With Lynnfield, Intel have kept what makes the current 'Bloomfield' Core i7 900 series processors great, while removing the aspects that really only catered to the Server/Workstation segment, such as the triple-channel memory interface and the QuickPath Interconnect (QPI). Compared to the current mainstream Core 2 Quad offerings, Intel have worked extensively to improve the power efficiency of this new processors family, while significantly increasing performance in most consumer-oriented applications, in part thanks to the very aggressive Turbo Boost technology. Furthermore, by integrating the PCIe controller onto the processor itself, Intel have been able to do away with the northbridge and create a 2-chip platform, which should help reduce motherboard prices and overall power consumption.

Let's examine some of these advancements:

• Integrated Memory Controller (IMC)

i5-750 on the right, i7-870 on the left.​

As with the current Intel Core i7 900 series processors, the Lynnfield family features an integrated memory controller. The benefits of this design are that the memory is directly connected to the processor, which not only means significantly lower latency, but much higher bandwidth as well. Unlike the Bloomfield chips though -which have a triple-channel memory interface- the Lynnfield chips feature a more standard dual-channel interface. While this may seem like a significant step back, the good news is that the supported DDR3 memory speed has increased from DDR3-1066 to DDR3-1333. When all is said and done, Lynnfield processors have a very respectable 21.2GB/s of memory bandwidth, compared to 25.6GB/s for Bloomfield. Furthermore, extensive testing of dual-channel versus triple-channel on the Core i7 900 series has shown that a dual-channel interface has noticeably lower latency. If this holds true with Lynnfield, we expect it's memory subsystem performance to be quite good.

• Integrated PCIe Controller

An industry first, Intel have moved the PCIe controller from the northbridge onto the processor itself, continuing the push towards a true System on Chip (SoC) design. This integrated memory controller supports 16 PCI-E 2.0 lanes, which can directed towards a single PCI-E x16 slot or two mechanical PCI-E x16 slots in x8/x8 configuration. While is this only half as many as the 32 PCI-E 2.0 lanes available on the Bloomfield/X58 platform, there is low latency advantage attributable to having the PCIe controller built into the CPU die.

• SSE4.2

Building upon Penryn's implementation of SSE4.1, which was focused on improving video encoding, image/video editing, faster 3D game physics, etc...the Nehalem architecture adds 7 new instrutions, namely Accelerated String and Text New Instructions (STTNI) and Application Targeted Acceleration (ATA), which focus on faster XML parsing, faster search and pattern matching, and other cryptic processor functions.

Keep in mind that with Penryn, the SSE4 instructions were responsible for the most significant performance increases, so we definitely look forward to seeing what Intel can accomplish with these latest instructions.

• Hyper-Threading

Although the i5-750 doesn't feature this technology, the Core i7 800 series chips support Hyper-Threading (HT). With HT enabled, a processor with four physical cores is viewed by the operating system as having eight logical cores. A core usually processes the pieces of the different threads one after another, however an HT-enabled core can process two threads in a simultaneous manner. While Hyper-Threading did not perform particularly well on the Pentium 4, Nehalem's architecture was designed to remove many of the processing bottlenecks that had previously crippled feature. Depending on the workload, and how effectively multi-threaded an application is, the performance increases can be 20% or higher.

• Power Control Unit (PCU)

Nehalem’s Power Control Unit (PCU) is an extremely innovative power management feature that uses an on-chip micro-controller to actively manage the power and performance of the entire processor with the help of numerous integrated power sensors. The PCU can dynamically alter the voltage and frequency of the CPU cores to lower power consumption or provide performance boost in conjunction with the Turbo Mode feature. Also, thanks to a development know as Power Gates, idle cores can be completely shut down and placed in a C6 sleep mode while other cores continue working. This is noteworthy because C6 mode had previously only been featured on mobile processors. On Lynnfield, the PCU has been tweaked to further improve power efficiency, and Intel is claiming that the i5-750's idle power consumption is up to 50% lower than that of the Core 2 Quad series.

• Turbo Mode

Lynnfield's more aggressive Turbo Boost technology has been highly advertised for months, and it is arguably one of this product's best selling points. Much like on the Bloomfield chips, Turbo Boost automatically overclocks the processor based on the workload demand. All Core i5 processors come with four additional speed bins, which is to say that they have four higher multipliers that they can use under certain scenarios. The Core i7 800 series have five extra additional speed bins, which equates to a roughly 666Mhz speed boost. For example, if you are using a single-threaded application, the PCU will down-clock or shut down three cores, thereby freeing up power and lowering heat output while "overclocking" that one core that is in use. If an application is multi-threaded and the processor is not running too hot, the PCU will overclock all the loaded cores up by 2-to-4 speed bins. The only limit to Turbo Mode is the power and thermal headroom, so keeping your processor cool is an important greater priority with Lynnfield chips.

If you are more visually-inclined, the following illustration should help explain the new Turbo Boost implementation:

While the Core i7 900 series can only provide a 133Mhz (single-thread) or 266Mhz (multi-thread) speed boost, Lynnfield can Turbo Boost up by 532Mhz (i5-750) or 666Mhz (i7-860/870). This is a pretty strong selling point, and as you will see shortly, the performance gains are impressive.


As we have stated in the past, the new performance and energy-saving features are what truly distinguish Nehalem as a veritable next-generation microarchitecture. There are several little technologies at work that some users may never know exist, but which ultimately deliver a superior computing experience. With regard to Lynnfield, the downgrade from a triple-channel to dual-channel memory interface is really nothing to fret about, especially given the higher supported memory speeds and the much more impressive Turbo Boost implementation. The integration of the PCIe controller into the chip is novel idea, and we look forward to seeing whether there are any performance advantages/disadvantages to this approach. Overall, for it being marketed as 'mainstream' product, Lynnfield brings a surprisingly robust spec sheet to the table. While some have suggested that Lynnfield is 'Nehalem Lite', we suggest that on paper it appears to be Nehalem 2.0.

 

[MAC]

P55 Express Chipset Examined

P55 Express Chipset Examined

Intel's new P55 Express 'Ibex Peak' chipset is a true break from their traditional chipset design. Unlike all previous Intel chipsets which featured both a northbridge and a southbridge (eg. X58 Express + ICH10R), the P55 is a one-chip solution. As such, it has been given the new designation of Platform Controller Hub (PCH). When it comes to PCI-Express 2.0 connectivity things get a little complicated with this chipset since in the past, the northbridge supplied the graphics-related PCI-E lanes. However, Lynnfield processors feature an industry-first: an integrated PCI-E controller that supports 16 PCI-E 2.0 lanes supplying two mechanical PCI-E x16 slots. If only one graphics card is installed, it will operate at the full electrical x16 speed and if two graphics cards are installed, the PCI-E lanes are divided between both PCI-E x16 slots and they will operate at x8 each. On motherboards with three mechanical PCI-E x16 slots, the first two slots will each operate at x8 while the third slot will operate at x4. How is this possible if we have already established that the integrated PCI-E controller only supports 16 PCI-E lanes? The additional 4 PCI-E lanes come from the P55 PCH itself, which can supply up to 8 PCI-E 1.0 lanes in total.

On the connectivity front, the P55 supports 14 USB 2.0 ports and 6 SATA II ports with Matrix Storage Technology. It also features one Gigabit LAN port and HD Audio Technology. It does not feature support for Intel's Trusted Execution Technology (TXT), formerly known as LaGrande, which provides hardware-level protection against malicious software.

The P55 PCH communicates to the processor via the Direct Media Interface (DMI), which is a 2 GB/s point-to-point connection, which is roughly equivalent to a PCI-E x4 1.0 link. By the way, the DMI is by no means new, it has long been used as the link between the northbridge and southbridge.

Much like the P45 Express and X58 Express chipsets, the P55 PCH is manufactured on the venerable 65nm process, and it has a low default voltage of 1.0V. As a result of this low voltage, and the simple fact that the P55 does not actually do much, it does run quite cool. Did we mention that it is also relatively tiny? The P55 package size of just 27mm x 27mm, and the actual die is a minuscule 8mm x 8mm as revealed below:

Next let's see what kind of motherboards Intel themselves have built around this new chipset.

 

[MAC]

The Platform in Pictures

The Platform in Pictures

By now, most of you have seen the P55 motherboard previews that have been plastered all over the internet, however Intel's own P55 motherboard have been kept a secret...until now, of course.

On a side note, if you have not yet had the opportunity to check them out, you may be interested in our P55 motherboard previews. We have taken a look at several very interesting models, namely the ASUS P7P55D Deluxe, ASUS RoG Maximus III Formula, EVGA P55 FTW, and the MSI P55-GD80.

Intel DP55KG (Kingsberg)

​

The DP55KG is Intel's enthusiast-oriented SKU, as you can see by the omnipresent Skull that has been the logo for Intel's performance motherboards for the last few years. From a specifications point-of-view, this model features two PCI-E x16 slots, one PCI-E x4 slot, two PCI-E x1 slots, and two legacy PCI slots. This is one of Intel's most connectivity-rich motherboards, with 8 angled SATA-II ports, 2 eSATA ports, 2 FireWire ports and up to 13 USB 2.0 ports (via headers). It also features one Intel Gigabit LAN port and a 7.1 Intel HD Audio codec with Dolby Digital Home Theatre technology. This particular model features a six-phase power design with robust DrMOS and CHiL components. Compared to the DDR3-2000+ memory frequency support that other manufacturers have given their models, Intel have gone the conservative route by topping out at DDR3-1600. Those who overclock will be glad to see MOSFET heatsinks and the onboard LED post code display. Aside from the Skull on the PCB, which actually has LED backlighting and can serve as the HDD activity light, one of this board's most unique features is built-in bluetooth connectivity, which is something that we have yet to see on any other P55 motherboard.

Intel DP55SB (Sharpsburg)

​

The MicroATX form factor has been increasing in popularity during the last 12-18 months, and you can expect to see many mATX P55 motherboards on the market very soon. The DP55SB is Intel's first compact enthusiast offering, and it is part of the Extreme Series motherboard family. It is similar to to DP55KG, but minus the PCI-E x4 slot, the two PCI slots, and two SATA-II ports.

Intel DP55WG (Warrensburg)

​

Next up we have the DP55WG, which is a Media Series product but "with a Touch of Extreme". This model features two PCI-E x16 slots, one PCI-E x4 slot, two PCI-E x1 slots, and two legacy PCI slots. It comes with 6 angled SATA-II ports, 2 FireWire ports, up to 14 USB 2.0 ports (via headers). It features the same six-phase DrMOS + CHiL power design as the DP55KG, and the onboard LED post code display too. The only really notable omission is the MOSFET heatsinks.

Intel DP55WB (Whitesburg)

​

The DP55WB is the barebones model of the four. It features a cut-down 3-phase power design and only one PCI-E x16 slot. We don't know the specifics of this model, but we wouldn't be surprised to see it retailing for under $80.



For this CPU launch, we were provided with the enthusiast Kingsberg model, which is what we used in this review.

Click on images to enlarge​

As with all Intel's enthusiast-oriented Extreme Series motherboards, the Kingsberg features a simple black and blue theme. As previously mentioned, you can see that the skull logo once again has a predominant place on board. However, what you can't tell is just awesome that skull is since it is backlit and the eyes glow red, serving as hard drive activity lights.

The full platform in all its glory. The cooler is the new Thermalright MUX-120, which features a tweaked and lightened Ultra 120 Extreme design. In addition Patriot was kind enough to send us an early sample of their upcoming Sector 5 Viper II DDR3 memory which has been specifically designed for the capabilities of the P55 / Lynnfield platform.


The specifics regarding all of our test platforms, as well as our methodology, can be found on the next page.

 

[MAC]

Test Setups & Methodology

Test Setups & Methodology

For this review, we have prepared four different test setups, representing all the popular platforms at the moment, as well as most of the best-selling processors. As much as possible, the four 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.

Intel Core i5 & Core i7 "Lynnfield" Test Setup​

​

Intel Core i7 "Bloomfield" Test Setup​

​

Intel Core 2 Test Setup​

​

AMD Phenom II AM3 Test Setup​

​

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 followed by a defragment and a reboot.

C)To ensure consistent results, a few tweaks were applied to Windows Vista and the NVIDIA control panel:

• Sidebar – Disabled
• UAC – 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
• NVIDIA PhysX – Disabled
• V-Sync – Off

D) Programs and games are then installed & updated followed by another defragment.

E) Windows updates are then completed installing all available updates followed by a defragment.

F) Benchmarks are each ran three times after a clean reboot for every iteration of the benchmark unless otherwise stated, the results are then averaged. If they were any clearly anomalous results, the 3-loop run was repeated. If they remained, we will mention it in the indvidual benchmark write-up.

Here is a full list of the applications that we utilized in our benchmarking suite:

• 3DMark06 Professional v1.1.0
• 3DMark Vantage Professional Edition v1.0.1
• Cinebench R10 64-bit
• Crysis v1.21
• Far Cry 1.02
• HyperPi 0.99b
• High Definition Experience and Performance Ratings Test 2009 (HDxPRT 2009) (Adobe Elements 7.0/QuickTime Player/iTunes 8.0.1/Sorenson Squeeze/PowerDVD 8/DiVX engine)
• Lame Front-End 1.0
• Lavalys Everest Ultimate v5.02.1834 Beta
• Left 4 Dead (Latest update)
• PCMark Vantage Advanced 64-Bit Edition (1.0.0.0)
• Photoshop CS4 Extended (64-bit)
• ScienceMark 2.0 Build 21MAR05
• SiSoft Sandra Professional 2009.9.15.124
• Street Fighter 4 Demo
• Supreme Commander v1.1.3280
• Valve Particle Simulation Benchmark
• WinRAR 3.8.0
• World in Conflict v1.010
• x264 HD Benchmark v1.0

That is about all you need to know methodology wise, so let's get to the good stuff!

 

[MAC]

Feature Test: Hyper-Threading (HT)

Feature Test: Hyper-Threading (HT)

Eight-threaded goodness​

The Nehalem microarchitecture marks the return of Hyper-Threading (HT), which is a feature that was first implemented on the Pentium 4 "Northwood" but with little success. Thankfully, as we demonstrated in our Core i7 review, this new microarchitecture has really been designed to take advantage of HT's multi-threading performance benefits.

Intel is so confident in the value of Hyper-Threading that with Lynnfield they have segregated the models based on support for this feature, with the cheaper Core i5-750 lacking HT. So does Hyper-Threading's increased multi-threading performance really warrant a price premium? Let's find out with a small selection of multi-threaded applications:

<table align="center" table border="0" bgcolor="#666666" cellpadding="5" cellspacing="1" width="735px"><tr><td align="center" bgcolor="#cc9999" width="130"><b></b></td><td align="center" bgcolor="#cc9999" width="180"><b>Intel Core i7-870 - HT Off</b></td><td align="center" bgcolor="#cc9999" width="180"><b>Intel Core i7-870 - HT On</b></td><td align="center" bgcolor="#cc9999" width="180"><b>Performance Difference</b></td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>Cinebench R10 64-bit: xCPU</b></td><td align="center" bgcolor="#ececec" width="100">16170</td><td align="center" bgcolor="#ececec" width="100">18115</td><td align="center" bgcolor="#ececec" width="100">+12%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>WinRAR 3.8.0 Compression</b></td><td align="center" bgcolor="#ececec" width="100">185 secs.</td><td align="center" bgcolor="#ececec" width="100">162 secs.</td><td align="center" bgcolor="#ececec" width="100">+14%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>x264 HD Benchmark</b></td><td align="center" bgcolor="#ececec" width="100">21.31 FPS</td><td align="center" bgcolor="#ececec" width="100">26.33 FPS</td><td align="center" bgcolor="#ececec" width="100">+24%</td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>3DMark Vantage: CPU Score</b></td><td align="center" bgcolor="#ececec" width="100">14639</td><td align="center" bgcolor="#ececec" width="100">19543</td><td align="center" bgcolor="#ececec" width="100">+33%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>Far Cry 2</b></td><td align="center" bgcolor="#ececec" width="100">72.18 FPS</td><td align="center" bgcolor="#ececec" width="100">71.40 FPS</td><td align="center" bgcolor="#ececec" width="100">0%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>Left 4 dead</b></td><td align="center" bgcolor="#ececec" width="100">126.11 FPS</td><td align="center" bgcolor="#ececec" width="100">128.97 FPS</td><td align="center" bgcolor="#ececec" width="100">+2%</td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>Street Fighter 4</b></td><td align="center" bgcolor="#ececec" width="100">192.76 FPS</td><td align="center" bgcolor="#ececec" width="100">192.69 FPS</td><td align="center" bgcolor="#ececec" width="100">0%</td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>Valve Particle Simulation Benchmark</b></td><td align="center" bgcolor="#ececec" width="100">138 Score</td><td align="center" bgcolor="#ececec" width="100">152 Score</td><td align="center" bgcolor="#ececec" width="100">+10%</td></tr></table>

The results speak for themselves. In highly multi-threaded applications, HT can make a significant difference, speeding up a real-life workload by 12% to 24%.

As you can see in Far Cry 2, Left 4 Dead, Street Fighter 4, even games with solid multi-threaded engines don't usually recognize more than four-threads, which negates any possible HT-related performance gains. We are quite confident that HT-friendly engines are coming down the pipeline though, so Hyper-Threading may eventually play a role in gaming scenarios.

 

[MAC]

Feature Test: Intel Turbo Boost - Turbocharged

Feature Test: Intel Turbo Boost - Turbocharged

2.93Ghz Core i7-870 Turbo'ing up to 3.59Ghz​

For those of you who skipped the microarchitecture section, let's recap what Turbo Boost is, what it does, and how it is better than ever on Lynnfield.

Turbo Mode is an exciting feature that automatically unlocks additional speed bins (multipliers) and allows the processor to self-overclock based on thermal conditions and workload. For example, if the Power Control Unit (PCU) senses that only one core is active and the other three are in an idle state, it will use the unused power and thermal headroom to overclock that single active core to ensure superior single-threaded performance. Conversely, if you are running a multi-threaded application, the PCU will measure the thermal headroom and if the processor is running cool enough it will overclock all four cores. On the Core i7 Bloomfield chips, Turbo Mode could provide a 266Mhz speed boost in single-threaded workloads and 133Mhz in multi-threaded applications. However, for Lynnfield Intel have really kicked things up a notch. As we have all seen these last few months, Core i7 processors have a lot of overclocking headroom, which is testament to Intel's top-notch 45nm High K manufacturing process. With this in mind, Intel's engineers have put this substantial headroom to good use by giving Core i5-700 series processors 4 additional speed bins (up to 532Mhz faster) and Core i7-800 series model 5 additional speeds (up to 666Mhz faster).

Although the results will be fairly self-evident, let's check out the performance gains that Turbo Mode provides on the top-end Core i7-870 model. As per the above, thermal conditions permitting, it will run one core at 3.59GHz for single-threaded workloads, and anywhere between 3.06Ghz to 3.46Ghz in multi-threaded applications.

<table align="center" table border="0" bgcolor="#666666" cellpadding="5" cellspacing="1" width="735px"><tr><td align="center" bgcolor="#cc9999" width="130"><b></b></td><td align="center" bgcolor="#cc9999" width="180"><b>Intel Core i7-870<br>Turbo Boost Off</b></td><td align="center" bgcolor="#cc9999" width="180"><b>Intel Core i7-870<br>Turbo Boost On</b></td><td align="center" bgcolor="#cc9999" width="180"><b>Performance Difference</b></td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>Cinebench R10 64-bit: Single Thread</b></td><td align="center" bgcolor="#ececec" width="100">3853</td><td align="center" bgcolor="#ececec" width="100">4580</td><td align="center" bgcolor="#ececec" width="100">+19%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>Cinebench R10 64-bit: Multi-Thread</b></td><td align="center" bgcolor="#ececec" width="100">16856</td><td align="center" bgcolor="#ececec" width="100">18115</td><td align="center" bgcolor="#ececec" width="100">+7%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>HDxPRT</b></td><td align="center" bgcolor="#ececec" width="100">203</td><td align="center" bgcolor="#ececec" width="100">222</td><td align="center" bgcolor="#ececec" width="100">+9%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>Lame Front-End</b></td><td align="center" bgcolor="#ececec" width="100">172s</td><td align="center" bgcolor="#ececec" width="100">144s</td><td align="center" bgcolor="#ececec" width="100">+19%</td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>Photoshop CS4 64-bit</b></td><td align="center" bgcolor="#ececec" width="100">248.1s</td><td align="center" bgcolor="#ececec" width="100">206.1s</td><td align="center" bgcolor="#ececec" width="100">+20%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>SuperPi 32M</b></td><td align="center" bgcolor="#ececec" width="100">781.592s</td><td align="center" bgcolor="#ececec" width="100">648.211s</td><td align="center" bgcolor="#ececec" width="100">+20%</td><tr><td align="center" bgcolor="#ececec" width="100"><b>WinRAR 3.8.0 Compression</b></td><td align="center" bgcolor="#ececec" width="100">200s</td><td align="center" bgcolor="#ececec" width="100">162s</td><td align="center" bgcolor="#ececec" width="100">+23%</td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>x264 HD Benchmark</b></td><td align="center" bgcolor="#ececec" width="100">24.84 FPS</td><td align="center" bgcolor="#ececec" width="100">26.33 FPS</td><td align="center" bgcolor="#ececec" width="100">+6%</td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>3DMark Vantage: CPU Score</b></td><td align="center" bgcolor="#ececec" width="100">18005</td><td align="center" bgcolor="#ececec" width="100">19543</td><td align="center" bgcolor="#ececec" width="100">+8%</td></tr><tr><td align="center" bgcolor="#ececec" width="100"><b>Valve Particle Simulation Benchmark</b></td><td align="center" bgcolor="#ececec" width="100">142 Score</td><td align="center" bgcolor="#ececec" width="100">152 Score</td><td align="center" bgcolor="#ececec" width="100">+7%</td></tr></table>

As promised, the performance gains are there and they are substantial, ranging from 6-7% in highly multi-threaded applications to 20% in single-thread workloads. The performance gains that Turbo Boost provides are definitely nothing to scoff at, especially since they free and automatic which is ideal for less computer savvy users.

Now when you combine Turbo Boost with Hyper-Threading, you have two technologies that can work together to create some very noticeable performance improvements. Enough talking, time for some real benchmarks.

 

[MAC]

Synthetic Benchmarks: CPU & Cache

Synthetic Benchmarks: CPU & Cache

Lavalys Everest Ultimate v5.02

Everest Ultimate is the most useful tool for any and all benchmarkers or overclockers. With the ability to pick up most voltage, temperature, and fan sensors on almost every motherboard available, Everest provides the ability to customize the outputs in a number of forms on your desktop. We selected two of Everest's seven CPU benchmarks: CPU Queen and FPU Mandel. According to Lavalys, CPU Queen simple integer benchmark focuses on the branch prediction capabilities and the misprediction penalties of the CPU. It finds the solutions for the classic "Queens problem" on a 10 by 10 sized chessboard. At the same clock speed theoretically the processor with the shorter pipeline and smaller misprediction penalties will attain higher benchmark scores. The FPU Mandel benchmark measures the double precision (also known as 64-bit) floating-point performance through the computation of several frames of the popular "Mandelbrot" fractal. Both tests consume less than 1 MB system memory, and are HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.

When it comes to computational power it is pretty clear that the Core i7 processors are in a class of their own. Both the i7 800 & 900 series put up impressive numbers largely due to the fact that they are Hyper-Threading enabled. The $562 i7-870 is secondly only to a $999 i7-975 Extreme Edition in the CPU Queen test, and third behind the i7-950 in the FPU Mandel test, not a bad first showing. The non-HT i5-750 is quite competitive in CPU Queen, slightly surpassing the AMD Phenom II X4 945, but falling behind the venerable Core 2 Quad Q9550, a chip that is currently priced $25 more than the i5 model. In FPU Mandel though, the tables are turned with the $170 X4 945 surpassing both the i5-750 and Q9550.


Lavalys Everest Ultimate v5.02

As part of its enthusiast favourite Cache & Memory Benchmark, Everest provides very useful and in-depth cache performance figures. For this chart, we have combined the read, write, and copy bandwidth figures to achieve an aggregate bandwidth figure for each cache stage.

On the L1 front, the i7-975 and i7-860 chips once again lead the way, but AMD's Phenom II AM3 processors put up very impressive numbers as well. The overall field is really not stretched out by much though, except when you get to the lower-clocked Core 2 models and 3-year-old Q6600. When it comes to L2 cache bandwidth, the Nehalem-based chips leave all other architectures in the dust, posting aggregate bandwidth numbers that are 60% to 100% higher than the rest. We were in for a surprise when it came to the L3 cache bandwidth, since the new Lynnfield models blew away their Bloomfield counterparts. We do know that the Uncore has been tweaked on these new chips, but seeing the i7-870 (2400Mhz Uncore) post 45% higher bandwidth numbers than the i7-975 (2666Mhz Uncore) is unusual. The bulk of the performance improvement was in the read bandwidth, while the i7-900 series appeared to still have an edge with regard to write and copy speeds. Further testing (and perhaps a new version of Everest) will be needed to settle this issue.


Now we also ran an assortment of SiSoft Sandra Pro 2009.SP4 (15.124) benchmarks, however we encountered too many anomalies to be able to firmly stand by the results, especially with regard to the Core i5-750. Nevertheless, here are the charts for the curious among you.

• Arithmetic Benchmark (Pay attention to the i5-750's low GFLOPS)
• Multi-Media Benchmark
• Multi-Core Efficiency (Pay attention to the i5-750's low bandwidth and high latency)
• Cache/Memory Bandwidth
• Cache Benchmark (Pay attention to the i5-750's abnormally high L2 & L3 cache results)

In each case, there is no rhyme or reason to explain the clearly anomalous results, and they were repeatable 3 to 10 times in a row. As a result, we chose to disregard them just to be safe.