Techgage logo

Intel’s Sandy Bridge Revealed: Core i5-2500K & i7-2600K Reviewed

Date: January 3, 2011
Author(s): Rob Williams

The long-awaited launch of Intel’s Sandy Bridge is here, and we have all of the details of what to expect, what you need to “upgrade”, what models will be available at launch, and of course, their prices. We’re taking a look at two of the higest-end offerings, the Core i5-2500K and i7-2600K – both quad-cores and both fully unlocked.



Introduction

Exactly one year ago today, we posted our biggest (in terms of wordage) launch article ever. The product at hand? Intel’s Clarkdale processors, which at the time, were the most sophisticated offerings on the market. They were also the first to bring a GPU onto the same PCB as the CPU, which not only improved efficiency, but made things more convenient for the consumer.

The Clarkdale/Arrandale launch for Intel was quite successful, as the microarchitecture proved to be the fastest and most capable to date. In some cases, it even outperformed the company’s much larger Bloomfield offerings – especially where encryption is concerned. So overall, the value was quite high.

Where can things be improved? Sandy Bridge is Intel’s answer to that question, and it brings a lot to the table. There’s not only lots to talk about architecture-wise, but rather the platform as a whole. That includes motherboards, chipsets and even some of Intel’s major marketing focuses.

Unlike our Clarkdale launch article, this one is not going to be a 10,000 word mega-article, but rather it’s going to be a little more simple, tackling the basics and also covering what’s new. In the weeks ahead, we’ll be following-up with other content that delves a bit deeper into certain aspects of Sandy Bridge, including the new integrated graphics chip, the overclockability, and of course, UEFI (also known as the “new BIOS”).

To help put the new launch into perspective, I’m thrown together our largest processor table list to date. This beast includes almost all of Intel’s current offerings (I’ve removed impossible-to-purchase models), and nothing here has yet been made redundant. There are sure to be removals made in the weeks ahead, but until it happens, the models should be listed.

Intel CPU
Cores
Threads
Clock
Turbo
Cache
GPU
TDP
1Ku Price
Core i7-980X (4)
6
12
3.33GHz
3.60GHz
12MB
N/A
130W
$999
Core i7-970 (4)
6
12
3.20GHz
3.46GHz
12MB
N/A
130W
$885
Core i7-2600K (5)
4
8
3.40GHz
3.80GHz
8MB
~1350MHz
95W
$317
Core i7-2600S (5)
4
8
2.80GHz
3.80GHz
8MB
~1350MHz
65W
$???
Core i5-2600 (5)
4
8
3.40GHz
3.80GHz
8MB
~1350MHz
95W
$294
Core i7-960 (1)
4
8
3.20GHz
3.46GHz
8MB
N/A
130W
$562
Core i7-950 (1)
4
8
3.06GHz
3.33GHz
8MB
N/A
130W
$294
Core i7-870 (2)
4
8
2.93GHz
3.60GHz
8MB
N/A
95W
$294
Core i7-870S (2)
4
8
2.66GHz
3.60GHz
8MB
N/A
82W
$351
Core i7-875K (2)
4
8
2.93GHz
3.60GHz
8MB
N/A
95W
$342
Core i7-860S (2)
4
8
2.53GHz
3.46GHz
8MB
N/A
82W
$337
Core i7-930 (1)
4
8
2.80GHz
3.06GHz
8MB
N/A
130W
$294
Core i7-860 (2)
4
8
2.80GHz
3.46GHz
8MB
N/A
95W
$284
Core i5-2500K (5)
4
4
3.30GHz
3.70GHz
6MB
~1100MHz
95W
$216
Core i5-2500T (5)
4
4
2.30GHz
3.30GHz
6MB
~1250MHz
65W
$???
Core i5-2500S (5)
4
4
2.70GHz
3.70GHz
6MB
~1100MHz
65W
$???
Core i5-2500 (5)
4
4
3.30GHz
3.70GHz
6MB
~1100MHz
95W
$205
Core i5-2400 (5)
4
4
3.10GHz
3.40GHz
6MB
~1100MHz
95W
$184
Core i5-2400S (5)
4
4
2.50GHz
3.30GHz
6MB
~1100MHz
65W
$???
Core i5-2390T (5)
2
4
2.70GHz
3.50GHz
3MB
~1100MHz
35W
$???
Core i5-2300 (5)
4
4
2.80GHz
3.10GHz
6MB
~1100MHz
95W
$177
Core i5-760 (2)
4
4
2.80GHz
3.33GHz
8MB
N/A
95W
$205
Core i5-680 (3)
2
4
3.60GHz
3.86GHz
4MB
733MHz
73W
$294
Core i5-750S (2)
4
4
2.40GHz
3.20GHz
8MB
N/A
82W
$259
Core i5-655K (3)
2
4
3.20GHz
3.46GHz
4MB
733MHz
73W
$216
Core i5-750 (2)
4
4
2.66GHz
3.20GHz
8MB
N/A
95W
$196
Core i5-661 (3)
2
4
3.33GHz
3.60GHz
4MB
900MHz
87W
$196
Core i5-660 (3)
2
4
3.33GHz
3.60GHz
4MB
733MHz
73W
$196
Core i5-650 (3)
2
4
3.20GHz
3.46GHz
4MB
733MHz
73W
$176
Core i3-2120 (5)
2
4
3.30GHz
N/A
3MB
~1100MHz
65W
$138
Core i3-2100 (5)
2
4
3.10GHz
N/A
3MB
~1100MHz
65W
$117
Core i3-2100T (5)
2
4
2.50GHz
N/A
3MB
~1100MHz
35W
$???
Core i3-560 (3)
2
4
3.33GHz
N/A
4MB
733MHz
73W
$138
Core i3-550 (3)
2
4
3.20GHz
N/A
4MB
733MHz
73W
$117
Core i3-540 (3)
2
4
3.06GHz
N/A
4MB
733MHz
73W
$117
Pentium G6950 (3)
2
2
2.80GHz
N/A
3MB
533MHz
73W
$87
Microarchitecture: (1) Bloomfield, (2) Lynnfield, (3) Clarkdale, (4) Gulftown, (5) Sandy Bridge

The most expensive part we’ll see from this launch is the Core i7-2600K, at $317. It’s a quad-core offering that features eight threads and is clocked at 3.40GHz. Like the Clarkdale models released last January, Sandy Bridge brings to the table a very effective Turbo mode, and in the case of this particular CPU, it can top out at 3.80GHz when under stress.

At the very bottom we have the Core i3-2100, at $117. It’s a dual-core model that lacks a Turbo mode but does include HyperThreading, which could be argued to be even more important. Unlike the bigger models, the Core i3s (and also the Core i5-2390T) feature just 3MB of L3 Cache. The bigger models, including the Core i7s and most of the Core i5s, include either 6MB or 8MB.

One of the most ambitious goals of Sandy Bridge was to meld the GPU and CPU together, and going forward, that’s something we’ll be seeing from most released models (the exceptions will be future high-end parts). For launch models, the GPU core will be clocked at up to 1350MHz, and will vary depending on the model. Because Intel introduced a mode similar to Turbo for the GPU, the clocks will be a little more modest when idle.

Along with this launch comes some new box art, and also revised “Core i” logos. The previous box art looked quite nice as it was, but the revised art is even more modern, and a bit more colorful.

If there’s a downside to Sandy Bridge that’s worth mentioning up front, it’s that it supports the LGA1155 socket, not LGA1156. While AMD has managed to keep its sockets the same from launch to launch, Intel has been releasing a different design for each major one, which complicates the upgrade process for many. With the sockets being near-identical in size here, Intel no doubt wanted to use LGA1156, but due to important architecture differences, it couldn’t.

There is an upside, though. Because the physical size of the CPU hasn’t changed, neither has the mounting holes for the CPU coolers. That means that all LGA1156 coolers will fit an LGA1155 socket no problem.

We’re going to spend the next couple of pages tackling more of the specifics of Sandy Bridge and also its related components, so let’s get right into it!

Trends, Desktop Roadmap, Chip Design

To help fill us all in on what makes Sandy Bridge special, Intel provided a multitude of documents to peruse, including a mammoth 95-page “master” deck, taking care of all that’s important. We’re going to be using this deck to make sure we don’t miss anything noteworthy, so prepare for lots of screenshots!

Right off the bat, I have to mention a trend that I noticed with this launch that’s not immediately noticeable. Whenever Intel hands us PDFs explaining a new architecture, it normally tackles the desktop components before the mobile. Not here. Mobile comes first, further proving that mobile computing has reached a point where, quite frankly, it’s more important than desktop computing. This of course depends on how you look at things, though.

For the past few launches, Intel has touched base on the trends of content creation and also online video consumption, and it again talked about it for this launch. According to eMarketer.com, the number of people producing their own content will rise from ~80 million in 2008 to ~115 million by 2013, putting companies like Intel in a great spot to sell its fast processors.

There is one scary statistic to take away from this slide. In a single day, Facebook processes 415,000 video uploads. On a similar token, it’s estimated that 90% of the bandwidth people use in 2013 will be dedicated to video traffic. This isn’t much of a surprise, given the rising popularity of downloadable movies, and also services like Netflix.

Sandy Bridge, if not already obvious, is a rather major launch by Intel. It falls into the “Tock” part of Intel’s release schedule as it’s a new microarchitecture. The next “Tick” will be the 22nm Ivy Bridge, which if all goes well, could be released in one year’s time. Just think about it… 22nm. These transistors sure are getting small!

According to Intel’s roadmap, Sandy Bridge desktop processors are going to replace Clarkdale/LGA1156 offerings, and should remain through the entire year. As it usually goes, we’re sure to see speed bump products released at certain points, but nothing major will change.

Higher-end Lynnfield/LGA1156 and Bloomfield/LGA1366 parts will remain throughout the year as well, and we might even see follow-up products there even though some portions of the architecture have been superseded by Sandy Bridge. Though not shown on the roadmap, Intel is likely to release LGA1356 six and eight-core models in the fall, which would replace current Bloomfield parts. These CPUs would differ from the rest of the Sandy Bridge line-up as they wouldn’t include an integrated GPU.

Something that’s immediately noticeable about Sandy Bridge chips is that they’re a lot longer than what we’re used to seeing, thanks entirely to the GPU that’s fused onto the left (bottom) side. Unfortunately, that portion of the die isn’t nearly as pretty as the rest, but like most things in hardware, it’s not the looks that count, but the functionality.

For a much larger and untouched version of the die shot, click here.

The slide above looks simple, but it’s in fact showing off some major changes introduced with Sandy Bridge. Whereas the GPU with Clarkdale had to use the cache off of the CPU in a less-than-ideal fashion, the fusion design improves cache sharing tremendously. Both the CPU and GPU are literally touching the cache in this design, so there’s not much of a bottleneck.

The other perk is that because the memory controller is based inside of this chip, rather than on a different one, the memory latencies that seemed to be sky-high with Clarkdale should be much improved with Sandy Bridge. All in all, this entire design as a whole is amazingly efficient.

Architecture Improvements, Graphics

On the previous page we had a basic overview of the Sandy Bridge chip, but here we’ll take a more specific look, and see what’s been brought to the table. As the slide below shows, the chip is dominated in the center with the cores and cache, and at the bottom is the graphics. At the top is the System Agent, PCIe and DMI buses, display and also the integrated memory controller. As mentioned before, the L3 cache is accessible to both the cores and GPU.

Like Clarkdale, a dual-channel memory controller is included here, so there’s still no sign of a triple-channel controller for the mainstream anytime soon. This is likely due to the fact that it’s simply not needed, because those who do require major bandwidth tend to be running large workstations and more beefier components, such as a six-core processor. The other reason could be related to costs, as the extra DIMM slots and circuitry required for triple-channel motherboards could result in needless extra costs for the average consumer.

Sandy Bridge brings a couple of improvements over the Clarkdale architecture, including a more effective Turbo and also AVX instructions, which are media-focused similar to SSE4. HyperThreading has again been brought back, as has been the embedded DisplayPort functionality. The L3 cache has been renamed to “Last Level Cache”, as it now supplies more than just the CPU.

I’ve mentioned Turbo a couple of times already, so let’s finally talk about it. Like previous implementations, Turbo works on the basis of providing faster clock speeds to cores that could benefit from it. If just one core is being stressed, then it can reach huge heights, but as more cores are brought into the picture, each one gets a little bit of a boost, rather than a huge one.

This slide might look like some of the other Turbo ones we’ve posted before, but this one brings “Dynamic” Range into things, which is an additional bin that can be executed if the IGP is not being used to a great extent. This would apply to regular desktop work, so if you are actually utilizing the GPU in any real way, that top bin wouldn’t likely spring into action.

As you might have guessed, the IGP also partakes in the Turbo functionality, running at lower frequencies while at idle, and ramping up with the need arises.

Clarkdale brought quite a nice IGP to the table, one that could handle HD tasks of all sorts, but Sandy Bridge improves things a bit further. In what the company calls “Quick Sync”, Intel has brought the preprocessing and encoding tasks to the hardware, rather than leave it up to the software. This acceleration should result in far improved performance when converting video, and because it can handle up to 1080p, its uses are huge.

But what can we expect from gaming? Unfortunately, due to time and other factors, I was unable to “sync” a lot of time into graphics testing, but it’s a priority of mine to tackle after CES is all over and done with. So far, experiences have been good, but I don’t have enough information to produce results. Instead, I’ll pass along Intel’s results:

This launch introduces two different GPU models, HD Graphics 2000 and HD Graphics 3000. The former is meant to be a nice improvement over the GPU seen in the Clarkdale-built Core i5-661, while the latter aims to deliver the best possible integrated gaming experience.

Specifics aren’t given, but it’s clear that both GPUs to be bundled with Sandy Bridge processors are going to deliver a much better experience than what we’ve been used to. Of course, integrated graphics have never been nearly as impressive as even lower-end discrete solutions, but we are seeing some major strides being made here, with the HD Graphics 3000 being about 2x as fast as the GPU found in the Core i5-661.

Intel made sure that with its Clarkdale launch, many media technologies were supported, and Sandy Bridge not surprisingly improves things there as well. Introduced is encode acceleration for H.264 and MPEG 2, three post-processing enhancements and also support for HDMI 1.4.

Looking deeper into the GPUs included here, the Unified Shader Architecture is of the 4th generation, and 6 or 12 execution units can be employed, depending on the model. The latest offerings deliver a dedicated math box and media processing, and also DX10.1 support. No, there’s no DX11 here, but being a lower-performance part, there’s little reason for it to exist.

The GPU core with any Clarkdale model topped out at 900MHz, but Intel increased that quite a bit, with models now ranging up to 1350MHz. Once again, the 2560×1600 resolution is supported, and with HDMI 1.4 added in, so is 3D.

On the next page, we’ll quickly talk about some interesting desktop trends Intel has raised, UEFI, and also talk about the models that feature S, T and K suffixes.

CPU Model Suffixes, UEFI, Desktop Trends

Well into the Core 2’s life cycle, and at the height of the “green” computing boom, Intel introduced low-power variants and denoted them with the help of an “S” at the end of the model name. That scheme has been brought back once again with this Sandy Bridge launch, with three different “S” models to be made available at launch, or shortly after.

As expected, the Sandy Bridge “S” models feature lower stock frequencies, but identical Turbo frequencies. In addition, the TDPs have been lowered, and prices raised (although pricing is being kept a secret until the parts hit retail). For the most part, the “S” models haven’t been too confusing, as there were either non-S or “S” models… normal or low-power. Nothing too hard to grasp.

In the spring of last year, though, Intel decided to spice things up by releasing “K” models. They were unlocked chips, designed for overclocking, and also have been brought to the Sandy Bridge launch, with two models immediately available.

To spice this launch up a bit more, a brand-new suffix has been added, “T”. This is a bit different, because like the “S” models, these don’t feature lower TDPs, but do feature lower clock speeds… and that includes the Turbo. At the same time, the GPU is cranked up, which in the end, helps round out the TDP numbers.

It can be a little confusing, but to sum up…

Pricing for both the “S” and “T” models is being hidden at the moment, but once we become aware of them, we’ll post about it in our news section.

To talk further about the “K” model, we can refer to the following slide. Non “K” models will be restricted to an increase of the Turbo multiplier up to 4x, while “K” models can see their Turbo multipliers cranked to about 20x past the stock. On a 3.30GHz CPU, that would bring things to 5.7GHz, which is more than enough for most people. For the hardcore overclockers out there, going higher than 5.7GHz would require a boost of the base clock, which for Sandy Bridge is 100MHz.

Last week I posted a simple question to our news section, “What’s UEFI, and Should You Care?”. I recommend reading it because it’ll tackle a bit more information than I will here, but the listos is that the UEFI is a BIOS replacement. It’s more capable, looks better, and is faster.

Here’s ASUS’ implementation of UEFI on its P67 line-up:

UEFI can feature a full graphical interface, and defaults to the “safe” resolution of the monitor, which will differ from model to model. In my case, my 30″ 2560×1600 display defaults to 1440×900, so that’s the resolution UEFI ran at for me. As you can see in the shot above, the UEFI BIOS looks almost like a full-screen Windows application. It features full color support and is very modular.

Note the boot priority at the bottom. This is one of my favorite features, because rather than use the + and – buttons to change priority, here you simply click on any drive and move it around. It’s amazingly simple, and intuitive. If you want low-power computing, or a minor overclock, you can even click one of those respective options, then reboot. Of course, it’s hard to ignore the easy-to-read info bits, such as the CPU temperature and voltage levels.

This is just the start screen for the BIOS, and I haven’t taken a screenshot of the rest of it. But once into the “Advanced Mode”, the layout is similar to the BIOS of old, except it has the same graphical detail as the main screen. Overall, UEFI has major potential, and so far, implementations I’ve used have been an unbelievable step up from the bland blue screen we’re all used to.

Alright, trends aren’t all that interesting in a CPU article, I realize, but this one is different. Intel included a slide that showed off desktop market forecasts, and as we tend to focus a lot on the desktop, it’s an important slide to pay attention to.

Some have been declaring the death of the desktop for a while, but according to these predictions, there will be no such thing. At least soon. According to IDC Worldwide, desktop usage will peak in 2012, then decrease in 2013 and again increase in 2014. Strange, but I’ve learned it’s better to not ask questions.

Split down into form factors, the rise of “small” designs is going to be rising for quite some time, but that’s of no real surprise. The “Mini-Tower PC” (mid-tower) has had its usage decline for the past couple of years, but it appears that after squinting at the “Desktop Workstation” segment, it might increase in adoption as the years go on, just slightly.

The “Tower” PC, which I assume means full-tower, is also going to see an increase, like the Desktop Workstation. That’s quite interesting, because while mid-towers are on the decline, towers and workstations are on the incline? That’s not what I would have expected. And of course, “Ultrasmall PCs” and “All-in-Ones” are on a path to some nice adoption in the years ahead.

Test System & Methodology

At Techgage, we strive to make sure our results are as accurate as possible. Our testing is rigorous and time-consuming, but we feel the effort is worth it. In an attempt to leave no question unanswered, this page contains not only our testbed specifications, but also a fully-detailed look at how we conduct our testing.

If there is a bit of information that we’ve omitted, or you wish to offer thoughts or suggest changes, please feel free to shoot us an e-mail or post in our forums.

Test System

The table below lists the hardware for our two current machines, which remains unchanged throughout all testing, with the exception of the processor. Each CPU used for the sake of comparison is also listed here, along with the BIOS version of the motherboard used. In addition, each one of the URLs in this table can be clicked to view the respective review of that product, or if a review doesn’t exist, you will be led to the product on the manufacturer’s website.

Please note that for benchmarking the Core i7-980X, we deviated from the Rampage II Extreme that we tested all other Bloomfield’s with in favor of taking Gigabyte’s X58A-UD5 for a spin. This motherboard change won’t effect the performance, but it will change the power consumption just a wee bit. As mentioned before, we’re in the process of upgrading our entire CPU test suite, and the X58A-UD5 will become the base of our new LGA1366 test platform.

Component
AMD AM2+/AM3 Test System
Processors

AMD Phenom II X4 965 Black Edition – Quad-Core, 3.40GHz, 1.325v
AMD Phenom II X4 955 Black Edition – Quad-Core, 3.20GHz, 1.325v
AMD Phenom II X3 720 Black Edition – Tri-Core, 2.80GHz, 1.325v
AMD Phenom II X2 555 Black Edition – Dual-Core, 3.20GHz, 1.325v

AMD Athlon II X4 635 – Quad-Core, 2.90GHz, 1.325v
AMD Athlon II X4 620 – Quad-Core, 2.60GHz, 1.375v
AMD Athlon II X3 435 – Tri-Core, 2.90GHz, 1.325v
AMD Athlon II X2 240e – Dual-Core, 2.80GHz, 1.325v
Motherboard
Gigabyte MA790GP-DS4H – 790GX-based, F3 BIOS (01/13/09)
Memory

Corsair XMS3 DHX 2x2GB – DDR3-1333 7-7-7-20-2T, 1.65v
Graphics
Audio
On-Board Audio
Storage

Intel X-25M 80GB SSD
Seagate Barracuda 500GB 7200.11
Power Supply
Corsair HX1000W
Chassis
SilverStone TJ10 Full-Tower
Display
Gateway XHD3000 30″
Cooling
Thermaltake V1
Et cetera
Windows Vista Ultimate 64-bit

Component
Intel LGA1155 Test System
ProcessorsIntel Core i7-2600K – Quad-Core, 3.40GHz, ~1.05v
Intel Core i5-2500K – Quad-Core, 3.30GHz, ~1.05v
Motherboard
ASUS P8P67 Deluxe (0907 BIOS, 12/23/2010)
Memory
Corsair XMS3 DHX 2x2GB – DDR3-1333 7-7-7-20-2T, 1.65v
Graphics
ATI Radeon HD 4870 512MB (Catalyst 8.11)
Audio
On-Board Audio
Storage

Intel X-25M 80GB SSD
Seagate Barracuda 500GB 7200.11
Power Supply
Corsair HX1000W
Chassis
SilverStone TJ10 Full-Tower
Display
Gateway XHD3000 30″
Cooling
Intel BXXTS100H CPU Cooler
Et cetera
Windows Vista Ultimate 64-bit

Component
Intel LGA1156 Test System
ProcessorsIntel Core i7-870K – Quad-Core, 2.93GHz, ~1.25v
Intel Core i7-870 – Quad-Core, 2.93GHz, ~1.25v
Intel Core i5-750 – Quad-Core, 2.66GHz, ~1.25v
Intel Core i5-665K – Dual-Core, 3.20GHz, ~1.25v
Intel Core i5-661 – Dual-Core, 3.33GHz, ~1.10v
Intel Core i3-530 – Dual-Core, 2.93GHz, ~1.00v
Motherboard
Lynnfield: Gigabyte P55-UD5 – P55-based, F3 BIOS (08/01/09)
Westmere: ASUS P7H55D-M EVO – H55-based, 0503 BIOS (12/02/09)
Memory

Corsair XMS3 DHX 2x2GB – DDR3-1333 7-7-7-20-2T, 1.65v
Graphics
ATI Radeon HD 4870 512MB (Catalyst 8.11)
Audio
On-Board Audio
Storage

Intel X-25M 80GB SSD
Seagate Barracuda 500GB 7200.11
Power Supply
Corsair HX1000W
Chassis
SilverStone TJ10 Full-Tower
Display
Gateway XHD3000 30″
Cooling
Thermalright MUX-120
Et cetera
Windows Vista Ultimate 64-bit

Component
Intel LGA1366 Test System
Processors

Intel Core i7-980X Extreme Edition
Six-Core, 3.33GHz, 1.30v
Intel Core i7-975 Extreme Edition – Quad-Core, 3.33GHz, 1.30v
Intel Core i7-920 Quad-Core, 2.66GHz, 1.30v
Motherboard
ASUS Rampage II Extreme – X58-based, 0705 BIOS (11/21/08)
Gigabyte X58A-UD5 – X58-based, F4 BIOS (02/12/10)
(Gigabyte’s board used only for Core i7-980X)
Memory

OCZ Gold 3x2GB – DDR3-1066 7-7-7-20-1T, 1.56v (920/940)
OCZ Gold 3x2GB – DDR3-1600 7-7-7-20-1T, 1.56v (965/975/980X)
Graphics
Audio
On-Board Audio
Storage
Intel X-25M 80GB SSD

Seagate Barracuda 500GB 7200.11
Power Supply
SilverStone DA1200
Chassis
SilverStone TJ10 Full-Tower
Display
Gateway XHD3000 30″
Cooling
Thermalright TRUE Black 120
Et cetera
Windows Vista Ultimate 64-bit

Component
Intel LGA775
Processors

Intel Core 2 Quad Q9650 – Quad-Core, 3.00GHz, 1.30v (Sim)
Intel Core 2 Duo E8400 – Dual-Core, 3.00GHz, 1.30v
Motherboard
ASUS Rampage Extreme – X48-based, 0501 BIOS (08/28/08)
Memory

Corsair XMS3 DHX 2x2GB – DDR3-1333 7-7-7-15-1T, 1.91v (1333FSB)
Corsair XMS3 DHX 2x2GB – DDR3-1066 6-6-6-15-1T, 1.91v (1066FSB)
Corsair XMS3 DHX 2x2GB – DDR3-800 6-6-6-15-1T, 1.91v (800FSB)

Graphics
Audio
On-Board Audio
Storage
Intel X-25M 80GB SSD

Seagate Barracuda 500GB 7200.11
Power Supply
Corsair HX1000W
Chassis
SilverStone TJ10 Full-Tower
Display
Gateway XHD3000 30″
Cooling
Thermalright TRUE Black 120
Et cetera
Windows Vista Ultimate 64-bit

(Sim) represents models that were tested using a faster, but under-clocked processor. For example, for the Q9550, we used the QX9770, since the specs are identical all-around, except for the clock speeds. Those were adjusted appropriately, effectively giving us a Q9550 to test with.

When preparing our testbeds for any type of performance testing, we follow these guidelines:

To aide with the goal of keeping accurate and repeatable results, we alter certain services in Windows Vista from starting up at boot. This is due to the fact that these services have the tendency to start up in the background without notice, potentially causing slightly inaccurate results. Disabling “Windows Search” turns off the OS’ indexing which can at times utilize the hard drive and memory more than we’d like.

Application Benchmarks

To help test out the real performance benefits of a given processor, we run a large collection of both real-world and synthetic benchmarks, including 3ds Max, Adobe Lightroom, TMPGEnc Xpress, Sandra 2009 and many more.

Our ultimate goal is always to find out which processor excels in a given scenario and why. Running all of the applications in our carefully-chosen suite can help better give us answers to those questions. Aside from application data, we also run two common games to see how performance scales there, including Call of Duty 4 and Half-Life 2: Episode Two.

Game Benchmarks

In an attempt to offer “real-world” results, we do not utilize timedemos in any of our reviews. Each game in our test suite is benchmarked manually, with the minimum and average frames-per-second (FPS) captured with the help of FRAPS 2.9.5.

To deliver the best overall results, each title we use is exhaustively explored in order to find the best possible level in terms of intensiveness and replayability. Once a level is chosen, we play through repeatedly to find the best possible route and then in our official benchmarking, we stick to that route as close as possible. Since we are not robots and the game can throw in minor twists with each run, no run can be identical to the pixel.

Each game and setting combination is tested twice, and if there is a discrepancy between the initial results, the testing is repeated until we see results we are confident with.

The two games we currently use for our motherboard reviews are listed below, with direct screenshots of the game’s setting screens and explanations of why we chose what we did.

Call of Duty: World at War

1680×1050
2560×1600




Crysis Warhead

1680×1050
2560×1600




Half-Life 2: Episode Two

1680×1050
2560×1600

Workstation: Autodesk 3ds Max, Cinebench, POV-Ray

Autodesk’s 3ds Max is without question an industry standard when it comes to 3D modeling and animation, with DreamWorks, BioWare and Blizzard Entertainment being a few of its notable users. It’s a multi-threaded application that’s designed to be right at home on multi-core and multi-processor workstations or render farms, so it easily tasks even the biggest system we can currently throw at it.

For our testing, we use two project files that are designed to last long enough to find any weakness in our setup and also allows us to find a result that’s easily comparable between both motherboards and processors. The first project is a dog model included on recent 3ds Max DVD’s, which we infused with some Techgage flavor.

Our second project is a Bathroom scene that makes heavy use of ray tracing. Like the dog model, this one is also included with the application’s sample files DVD. The dog is rendered at an 1100×825 resolution, while the Bathroom is rendered as 1080p (1920×1080).

So far, things are off to a great start! The Core i7-2600K, a $317 offering, slots right behind the $999 Core i7-980X Extreme Edition six-core… that’s impressive. It also proves faster than the Core i7-975 Extreme Edition quad-core – another good sign.

The $216 Core i5-2500K, which lacks HyperThreading, is noticeably slower than the Core i7-2600K, but it still manages to surpass most of the models tested here.

Cinebench R10

Like 3DS Max, Cinema 4D is another popular cross-platform 3D graphics application that’s used by new users and experts alike. Its creators, Maxon, are well aware that their users are interested in huge computers to speed up rendering times, which is one reason why they released Cinebench to the public.

Cinebench R10 is based on the Cinema 4D engine and the test consists of rendering a high-resolution model of a motorcycle and gives a score at the end. Like most other 3D applications on the market, Cinebench will take advantage of as many cores as you can throw at it.

The Sandy Bridge models continue to impress here, out-pacing everything except for the high-end Extreme Edition six-core. Even the single-threaded performance of the Core i5-2500K manages to beat the faster single-thread clock speed of the Core i7-980X. Need proof that Sandy Bridge is an efficient architecture? There it is.

Multi-Media: Adobe Lightroom, TMPGEnc Xpress

Photo manipulation benchmarks are more relevant than ever, given the proliferation of high-end digital photography hardware. For this benchmark, we test the system’s handling of RAW photo data using Adobe Lightroom, an excellent RAW photo editor and organizer that’s easy to use and looks fantastic.

For our testing, we take 100 RAW files (in Nikon’s .NEF file format) which have a 10-megapixel resolution, and export them as JPEG files in 1000×669 resolution, similar to most of the photos we use here on the website. Such a result could also be easily distributed online or saved as a low-resolution backup. This test involves not only scaling of the image itself, but encoding in a different image format. The test is timed indirectly using a stopwatch, and times are accurate to within +/- 0.25 seconds.

Unlike our 3D rendering tests on the previous page, Adobe Lightroom doesn’t take full advantage of all the cores that a CPU might offer, but the single-threaded performance can still be improved, and in this case, it has been. Both of the Sandy Bridge models performed exceptionally, placing themselves ahead of most of the competition.

TMPGEnc Xpress 4.5

When it comes to video transcoding, one of the best offerings on the market is TMPGEnc Xpress. Although a bit pricey, the software offers an incredible amount of flexibility and customization, not to mention superb format support. From the get go, you can output to DivX, DVD, Video-CD, Super Video-CD, HDV, QuickTime, MPEG, and more. It even goes as far as to include support for Blu-ray video!

There are a few reasons why we choose to use TMPGEnc for our tests. The first relates to the reasons laid out above. The sheer ease of use and flexibility is appreciated. Beyond that, the application does us a huge favor by tracking the encoding time, so that we can actually look away while an encode is taking place and not be afraid that we’ll miss the final encoding time. Believe it or not, not all transcoding applications work like this.

For our test, we take a 0.99GB high-quality DivX H.264 AVI video of Half-Life 2: Episode Two gameplay with stereo audio and transcode it to the same resolution of 720p (1280×720), but lower the bit rate in order to attain a modest file size. This test also utilizes the SSE instruction sets, either SSE2 or SSE4, depending on what the chip supports.

Both CPUs continue to impress here, although for some reason the 2600K proved to be a tad slower in our mobile encoding test. Due to time, I wasn’t able to investigate this, but it could be a fluke. The other tests did line up as expected, though.

Multi-Media: ProShow Gold, Sandra Multi-Media

While TMPGEnc XPress’ purpose is to convert video formats, ProShow from Photodex helps turn your collection of photos into a fantastic-looking slide show. I can’t call myself a slide show buff, but this tool is unquestionably definitive. It offers many editing abilities and the ability to export in a variety of formats, including a standard video file, DVD video and even HD video.

Like TMPGEnc and many other video encoders, ProShow can take full advantage of a multi-core processor. It doesn’t support SSE4 however, but hopefully will in the future as it would improve encoding times considerably. Still, when a slide show application handles a multi-core processor effectively, it has to make you wonder why there is such a delay in seeing a wider-range of such applications on the marketplace.

Our version of ProShow Gold is a bit aged at this point, but it still has what it takes to stress our processors, and once again Sandy Bridge delivers. The Core i7-2600K surpassed the performance of the Core i7-975 Extreme Edition in the DVD encode test, and lined up with the same result in the HD encode.

Sandra 2009 Multi-Media

This test here stresses the CPU’s ability to handle multi-media instructions and data, using both MMX and SSE2/3/4 as the instruction sets of choice. The results are divided by integer, floating point and double precision, three specific numbering formats used commonly in multi-media work.

Just as we’d expect a synthetic benchmark to do, we’re given rather predictable results here. Interestingly, AMD’s Phenom II X6 1090T out-paces the Core i7-2600K in the integer test, but in the float and double, Intel’s chip soars ahead.

Mathematics: Sandra Arithmetic, Crypto, Microsoft Excel

With each new processor launch, one thing that’s bound to prove faster are mathematical equations, which when all said and done, plays a massive role in a lot of our computing today. The faster an equation can be completed, the faster a math-heavy process can finish.

Sandra includes applications designed to specifically test the mathematical performance of processors, with the main one being the arithmetic test.

Isn’t it a bit sad to see a brand-new “mainstream” part beat out Intel’s highest-end six-core offering? That’s what happened with our Dhrystone test, but the Core i7-980X still proved its superiority in the Whetstone test. I guess you could say the results aren’t so cut and dhry! Alright, that was bad.

Sandra 2009 Cryptography

Crypto is a major part of computing, whether you know it or not, and certain processes can prove slower than others, depending on their algorithms. User passwords on your home PC are encrypted, as are user passwords on web servers (like in our forums). Past that, crypto is used in other areas as well, such as with creating of unbreakable locks on files or assigning a hash to a particular file (like MD5).

In Sandra’s Cryptography test, the results are outputted as MB/s, higher being better. Although this is somewhat of an odd metric to go by, generally speaking, the higher the number, the faster the CPU tears through the respective algorithm, which comes down to how fast a password is either encrypted, decrypted, signed, et cetera.

In one of our more humorous graphs, our Sandy Bridge models deliver some explosive performance as expected, although both the 2500K and 2600K flip-flop in their strengths. The overall results are far more impressive than most of our line-up, though, only to be surpassed by the Core i7-980X.

Microsoft Excel 2007

Most, if not all, businesses in existence have to crack open a spreadsheet at some point. Though simple in concept, spreadsheets are an ideal way to either track information or compute large calculations all in real-time. This is important when you run a business that deals with a large amount of expenses.

Although the importance of how fast a calculation takes in an Excel file is, we include results here since they heavily test the mathematical capabilities of each processor. Because Excel 2007 is completely multi-threaded (it can even take advantage of an 8-Core Skulltrail), it makes for a great benchmark to show the scaling between all of our CPUs.

I’ll let Intel explain the two files we use:

Monte CarloThis workload calculates the European Put and Call option valuation for Black-Scholes option pricing using Monte Carlo simulation. It simulates the calculations performed when a spreadsheet with input parameters is updated and must recalculate the option valuation. In this scenario we execute approximately 300,000 iterations of Monte Carlo simulation. In addition, the workload uses Excel lookup functions to compare the put price from the model with the historical market price for 50,000 rows to understand the convergence. The input file is a 70.1 MB spreadsheet.

CalculationsThis workload executes approximately 28,000 sets of calculations using the most common calculations and functions found in Excel*. These include common arithmetic operations like addition, subtraction, division, rounding and square root. It also includes common statistical analysis functions such as Max, Min, Median and Average. The calculations are performed after a spreadsheet with a large dataset is updated with new values and must re-calculate many data points. The input file is a 6.2 MB spreadsheet.

I love this test not only because it’s so quick to run, but because with each new architecture we test, we continue to see lower figures. And once again, Sandy Bridge wipes the floor here, delivering truly impressive results.

System: Sandra Memory, Multi-Core Efficiency

Generally speaking, the faster the processor, the higher the system-wide bandwidth and the lower the latency. As is always the case, faster is better when it comes to processors, as we’ll see below. But with Core i7, the game changes up a bit.

Whereas previous memory controllers utilized a dual-channel operation, Intel threw that out the window to introduce triple-channel, which we talked a lot about at August’s IDF. Further, since Intel integrates the IMC onto the die of the new CPUs, benefits are going to be seen all-around.

Before jumping into the results, we already had an idea of what to expect, and just as we did, the results seen are nothing short of staggering.

Mentioned earlier, the improved design of Sandy Bridge should give us far better memory latencies compared to Clarkdale, and sure enough, that’s true. In fact, the latencies we see are some of the best we’ve come across, and in terms of bandwidth, that’s also been amped up.

Sandra 2009 Multi-Core Efficiency

How fast can one core swap data with another? It might not seem that important, but it definitely is if you are dealing with a true multi-threaded application. The faster data can be swapped around, the faster it’s going to be finished, so overall, inter-core speeds are important in every regard.

Even without looking at the data, we know that Core i7 is going to excel here, for a few different reasons. The main is the fact that this is Intel’s first native Quad-Core. Rather than have two Dual-Core dies placed beside each other, i7 was built to place four cores together, so that in itself improves things. Past that, the ultra-fast QPI bus likely also has something to do with speed increases.

While memory latencies are impressive, the results here are a strange turn of events. The inter-core bandwidth and latencies are actually worsened compared to Clarkdale and other Intel offerings. Fortunately this hasn’t been reflected in our other results.

Power Consumption

It goes without saying that power efficiency is at the forefront of many consumers’ minds today, and for good reason. Whether you are trying to save money or the environment – or both – it’s good to know just how much effort certain vendors are putting into their products to help them excel in this area. Both AMD and Intel have worked hard to develop efficient chips, and that’s evident with each new launch. The CPUs are getting faster, and use less power, and hopefully things will stay that way.

To help see what kind of wattage a given processor draws on average, we use a Kill-A-Watt that’s plugged into a power bar that’s in turn plugged into one of the wall sockets, with the test system plugged directly into that. The monitor and other components are plugged into the other socket and is not connected to the Kill-A-Watt. For our system specifications, please refer to our methodology page.

To test, the computer is first boot up and left to sit at idle for five minutes, at which point the current wattage is recorded if stable. To test for full CPU load, LinX is run with 2560MB memory usage for a total of five minutes. During that run, the highest point the wattage reaches on the meter is captured and becomes our “Max Load”.

As is the way with improved architectures, power efficiency is a definite priority of both AMD and Intel, and on Sandy Bridge, the improvements are evident. How so? Well, the quad-core/eight-thread Core i7-2600K uses less power than the dual-core/four-thread Core i5-530 Clarkdale processor. Given the major performance advantage that the 2600K has, that’s without question one of the more impressive features of Sandy Bridge.

Final Thoughts

One of the nice things about reviewing a processor based on a new microarchitecture is that the chances of it being a bland release is slim. After all, the likes of AMD and Intel are constantly striving to release faster, more efficient, and feature robust offerings, so any downsides are likely to be minor. In the case of Sandy Bridge, there are no immediate downsides to speak of.

The most interesting aspect of Sandy Bridge to tackle is of course performance, which as a whole is rather incredible, yet still expected. With a new microarchitecture, we always expect to see performance improvements, but that doesn’t stop us from getting excited about them.

We saw the quad-core Core i7-2600K stick right behind the six-core Core i7-980X Extreme Edition in most tests. The former costs $317 and the latter $999… how could that not be exciting?

The value gets even better when overclocking is brought into things, although that’s a topic I’m going to leave to later this week or next to talk more about. What I can say is that hitting at least 4.50GHz on these tested CPUs isn’t much of a chore with an appropriate motherboard, and in our case we didn’t even have to manually adjust the CPU voltage.

Clock for clock, Sandy Bridge is faster than previous incarnations, and that in itself is worthy of note. Intel’s been leading the pack for a while, so when a faster and better architecture comes out, it always brings good things to the table. This launch is a little more interesting than the rest though, not only because of the feature-set, but because of things that come around it.

UEFI, for example, is a major step forward for the BIOS, and it’s a bit surprising that Intel itself isn’t talking about it at all. The reason is likely due to the fact that UEFI has actually been available for a while, and the ones that are doing really cool things with it are the other board vendors, like ASUS, GIGABYTE, MSI and so forth.

UEFI in particular is a subject I’m going to talk more about later this month, with comparisons between initial implementations. A BIOS might still be a BIOS, but UEFI is so robust that it deserves some attention. Current implementations look good, react well, perform fast and in most cases, are simply more enjoyable to use.

One thing we didn’t touch on too much is the AVX instruction sets, but that’s something we’ll no doubt be exploring more throughout the year, as current support is a little slim, as are most instruction sets after release. AVX does have major potential for boosting multimedia performance, though, so I’m looking forward to seeing solutions hitting the market.

The GPU is another thing worthy of talking about, although I do regret not getting in enough testing to talk about it here. Both GPUs Intel is deploying here is faster than previous solutions though, so that in itself is good. But on top of it, post-processing enhancements have been added, along with HDMI 1.4, so all-in-all, Intel’s HD Graphics is being constantly improved.

Most of the other improvements have been discussed in the earlier pages of the review, so rather than be redundant, I’ll sum up here.

It’s of no surprise, but Intel once again has delivered the most impressive CPU microarchitecture to date; one that improves on all aspects of previous ones. It’s faster, improved with useful features, includes faster integrated graphics, is much more efficient, and of course, is priced right. Even the lowest-end models are rather attractive, and though they cost more than AMD’s competition at the moment, the performance is likely to favor Intel given Sandy Bridge’s hard-to-beat architecture.

Quad-core Sandy Bridge models should be hitting e-tail this week, and dual-core models should be hitting them later this month, or next, depending. Likewise, motherboards from the multitude of vendors should also become immediately available, and as for CPU coolers, and LGA1156 model should work just fine.

This is far from being the last bit of content regarding Sandy Bridge, as we’ll be posting more throughout the month. The first follow-up article will be featuring overclocking information of the two CPU models we tested, and after that, we’ll dive into UEFI a bit deeper. Of course, you can expect a lot of motherboard reviews as well. Stay tuned!


Intel’s Sandy Bridge Processors

Discuss this article in our forums!

Have a comment you wish to make on this article? Recommendations? Criticism? Feel free to head over to our related thread and put your words to our virtual paper! There is no requirement to register in order to respond to these threads, but it sure doesn’t hurt!

Copyright © 2005-2018 Techgage Networks Inc. - All Rights Reserved.