Sign in with
Sign up | Sign in
Intel Core i7-3770K Review: A Small Step Up For Ivy Bridge
By ,
1. Ivy Bridge: Was It Worth The Wait?

Editor’s Note: Eager to show off what is has done with Intel’s Ivy Bridge architecture, system builder CyberPower PC is offering Tom’s Hardware's audience the opportunity to win a new system based on Intel’s Core i5-3570K processor. Read through our review, and then check out the last page for more information on the system, plus a link to enter our sweepstakes!

Both AMD and Intel know that great products create lofty expectations. When you follow up an Athlon with an Athlon 64, or a Core 2 with a Core i7, customers start looking for big progress from subsequent generations, too.

Well, we all know that Sandy Bridge was a home run on the desktop. Can all of the enthusiasts who posted to forums about waiting for Ivy Bridge really be faulted for hoping to see another round of impressive performance gains?

Ivy Bridge was never projected to be as impactful as its predecessor, though. The company’s “tick-tock” cadence defines alternating steps forward in processor architecture and manufacturing technology. When Intel pulls off a successful new design based on mature lithography, the improvements tend to be big, bold, and beautiful. Nehalem and Sandy Bridge, both “tocks,” left us satisfied and smiling. A process shrink typically introduces other benefits, such as smaller dies and power savings. Benchmark results, however, typically don't change as drastically. 

Westmere was Intel’s most recent “tick,” heralding the arrival of 32 nm transistors. The company used more diminutive geometry to create room for additional cores on its fastest desktop-oriented CPUs, giving us chips like the Core i7-990X.

Today’s Ivy Bridge launch represents the next “tick”—a 22 nm die shrink of the same fundamental architecture we already know as Sandy Bridge. Intel is calling it a “tick-plus,” though, because there actually are a few under-the-hood improvements.

Unfortunately for desktop enthusiasts, the most significant changes center on the design’s integrated graphics engine, which most of us simply don’t utilize.  

Naturally, the story is different in the mobile space, where lower power consumption and “fast enough” 3D capabilities combine to enable big battery life numbers and surprisingly acceptable performance. But today’s Core i7-3770K review doesn’t cover a mobile processor. Rather, we’re looking at Intel’s fastest multiplier-unlocked model, positioned to succeed the existing Core i7-2700K and -2600K.

Meet Ivy Bridge

Intel built Sandy Bridge-based chips in three different configurations: one quad-core and two dual-core designs. The most complex implementation included 995 million transistors in a 216 mm² piece of silicon. In comparison, the biggest Ivy Bridge die incorporates 1.4 billion transistors on a 160 mm² die.

Ivy Bridge: 1.4 billion transistors; 160 square millimetersIvy Bridge: 1.4 billion transistors; 160 square millimeters

Sandy Bridge: 995 million transistors; 216 square millimetersSandy Bridge: 995 million transistors; 216 square millimeters

If you were to show your grandmother die shots of Sandy Bridge and Ivy Bridge, she should be able to point out where most of those 400 million new transistors were added. Clearly, the built-in graphics engine is more prominent.

Much of the expansion is attributable to an increase in execution units—the programmable shaders responsible for graphics processing—which Intel claims boost 3D performance by as much as 2x. Sandy Bridge’s HD Graphics 3000 employed 12 EUs; Ivy Bridge’s HD Graphics 4000 pushes that number to 16. HD Graphics 4000 also uniquely supports DirectX 11, up to three display outputs, OpenCL and DirectCompute, and better Quick Sync performance, all of which we’ll test for you.

The rest of the layout should look fairly familiar. Given a mainstream focus, the die driving Core i7-3770K is a quad-core, Hyper-Threaded part with 8 MB of shared L3 cache divided up into four 2 MB slices, same as the Core i7-2600K we reviewed more than a year ago. There are a handful of small tweaks to what the cores themselves can do. Intel says that those adjustments, plus tweaks in the cache and memory controller, help improve the number of instructions per clock cycle this architecture executes. We’ll be running per-clock comparisons between Ivy and Sandy Bridge to help quantify those claims as well.


Cores / Threads
Base Freq.
Max. Turbo
L3 Cache
HD Graphics
Graphics Base Freq.
Graphics Max. Freq.
TDP (W)
Price
Third-Gen Core i7 Family
3770K
4/8
3.5 GHz
3.9 GHz
8 MB
4000650 MHz
1.15 GHz77
$313
3770
4/83.4 GHz
3.9 GHz
8 MB
4000650 MHz1.15 GHz77
$278
3770T
4/82.5 GHz
3.7 GHz
8 MB
4000650 MHz1.15 GHz45
$278
3770S
4/83.1 GHz
3.9 GHz
8 MB
4000650 MHz1.15 GHz65
$278
Third-Gen Core i5 Family
3570K
4/4
3.4 GHz
3.8 GHz
6 MB4000
650 MHz1.15 GHz77
$212
3570T
4/42.3 GHz
3.3 GHz
6 MB2500
650 MHz1.15 GHz45
$194
35704/43.4 GHz
3.8 GHz
6 MB2500650 MHz1.15 GHz77
$194
3550
4/43.3 GHz
3.7 GHz
6 MB2500650 MHz1.15 GHz77
$194
3550S
4/43.0 GHz
3.7 GHz
6 MB2500650 MHz1.15 GHz
65
$194
3470
4/43.2 GHz
3.6 GHz
6 MB
2500650 MHz1.1 GHz
77
$174
3470T
2/4
2.9 GHz
3.5 GHz
3 MB
2500650 MHz1.05 GHz
35
$174
3470S
4/42.9 GHz
3.6 GHz
6 MB2500650 MHz1.1 GHz
65
$174
3450
4/43.1 GHz
3.5 GHz
6 MB2500650 MHz1.1 GHz
77
$174
3450S
4/42.8 GHz
3.5 GHz
6 MB2500650 MHz1.1 GHz
65
$174


We’re still looking at a dual-channel memory controller, though it’s now rated for DDR3-1600 data rates. Given the right memory, enthusiasts have the option to overclock as high as 2667 MT/s (up from 2133 MT/s) in more granular 200 MHz increments.

And although Ivy Bridge carries over Sandy Bridge’s 16 lanes of on-die PCI Express connectivity, we now have official PCIe 3.0 support for cards like AMD’s Radeon HD 7000s and Nvidia’s GeForce GTX 680.

All told, Ivy Bridge is yet another highly integrated processor design from Intel. Its pieces were constructed by independent teams throughout the world—engineers in Israel are responsible for the IA cores, a team in Folsom, CA built the graphics engine, and a second team in Folsom implemented the interconnects, cache, and system agent. Of course, a process development group up in Oregon made sure it’d all come together on the new 22 nm node.

What is the final product capable of? Let’s walk through Ivy Bridge step by step, exploring where it excels and digging into where it falls short of the community’s expectations.

2. The Ivy Bridge Core: I Think I Know You

Intel purposely carries over a lot of the technology it introduced with Sandy Bridge, allowing the company to focus more intently on a smoother transition from 32 to 22 nm manufacturing. Thus, the capabilities of an Ivy Bridge-based IA core are very much similar to the prior generation.

Each core still hosts 32 KB of L1 data and L1 instruction cache, along with a 256 KB L2 cache. Moreover, quad-core models like the Core i7-3770K share up to 8 MB of last-level cache. Latencies appear very similar, indicating comparable cache bandwidth as Sandy Bridge.

Intel claims it made subtle adjustments to the IA cores, however, that improve performance in certain situations. Company representatives didn’t go into much detail about core architecture improvements at last year’s IDF, mentioning only that there are about half a dozen features in the core and another six or so more in the memory controller/cache that accelerate IA workloads. Fortunately, it’s easy enough for us to run a handful of single-threaded tests with Turbo Boost disabled to see how Core i7-3770K compares to Core i7-2700K, both operating at 3.5 GHz.

Ivy Bridge takes about three seconds off of our Lame, iTunes, and PDF creation metrics. That’s decidedly less impressive than what Sandy Bridge did compared to Nehalem—but again, it’s a result we expected.

The bottom line for enthusiasts is that Ivy Bridge’s IPC-oriented improvements alone are not compelling enough to warrant an upgrade from Sandy Bridge chips running at similar frequencies. 

Intel does incorporate a pair of security-oriented features that software developers will be able to exploit moving forward: a Digital Random Number Generator instruction and Supervisor Mode Execution Protection.

Designed to be standards-compliant, the DRNG’s purpose is to provide a high-quality and high-performance source of entropy—the measure of a cryptographic key’s unpredictability. As a result, an application can exploit the DRNG, and get reliably good random numbers at up to 2-3 Gb/s. Intel makes the instruction available to operating system- and user-level code at all privilege levels.

The other new feature, abbreviated SMEP, attempts to thwart escalation of privilege attacks that seek access to resources normally protected from a less privileged ring. Simply, it prevents the execution of supervisor mode code in user-mode memory pages.

3. HD Graphics 4000: The Plus In Intel’s Tick+

Tom Piazza, the Intel fellow who unveiled Ivy Bridge’s graphics subsystem at last year’s IDF, made the case that even though we’re looking at a die shrink—a tick in the company’s cadence, by definition—the integrated GPU is more accurately characterized as a tock.

I already mentioned that Intel improved the performance of its integrated GPU by adding four more execution units to its highest-end implementation, that it finally folded in DirectX 11 support, that Quick Sync is faster now than it was last generation, and that up to three displays are supported natively.

Getting there required a reorganization of how Intel approached processor-based graphics, allowing the company to not only set a more aggressive roadmap for scaling graphics in the future, but also to fix some of the inefficiencies that held Sandy Bridge back. The result is an architecture partitioned into five domains.

  1. The first domain includes global assets like the geometry pipeline. Incorporating programmable hull and domain shader stages complement a fixed-function tessellation unit as requisites for DirectX 11 support.
  2. Intel refers to the second domain as Slice Common, which hosts the rasterizer, the pixel back-ends, and L3 cache. Sandy Bridge didn’t have L3 dedicated to graphics because Intel wasn’t able to derive meaningful performance from it. The processor’s ring bus provided sufficient bandwidth that the shared L3 worked well enough. But because Ivy Bridge pushes graphics harder, the dedicated L3 supplements its bandwidth requirements, simultaneously reducing power consumption when the engine can go to its own repository rather than spinning up the ring.
  3. Domain three, dubbed Slice, includes the shaders, texture samplers, L1 instruction cache, and the Media Sampler used by Quick Sync. In future generations, this is one collection of resources Intel plans to use to scale performance up. It’s also able to lay down additional Slice Commons to scale back-end throughput accordingly.
  4. The fourth domain is made up of fixed-function media features. This can also be scaled up or down, depending on how granular Intel chooses to get with media-oriented performance.
  5. Display outputs constitute the last domain. You can do three digital outputs on a desktop platform (provided a motherboard vendor enables them), but two have to be DisplayPort connections—one at up to 2560x1600 and one at up to 1920x1200. The third screen can be HDMI (up to 1080p), DVI, VGA, or DisplayPort at up to 1920x1200.

Within each domain, Intel says it tweaked and tuned for additional performance, increasing geometry throughput, optimizing buffer-clearing, improving anisotropic sampling quality, maximizing sustained compute performance, and bolstering performance per watt by leveraging that dedicated graphics L3 cache.

From Theory To HD Graphics 4000

All of those features materialize on Core i7-3770K as the HD Graphics 4000 engine, armed with 16 EUs, a base frequency of 650 MHz, and a maximum dynamic frequency of 1.15 GHz. At idle, the graphics logic spins down to 350 MHz, creating more thermal headroom for the IA cores.

Last year, I lamented the fact that Intel armed most of its mobile processors and its K-series desktop SKUs with HD Graphics 3000, the fastest implementation available. Meanwhile, 12 other desktop-oriented models got stuck with HD Graphics 2000, hobbling their performance. Over the course of 2011, Intel slowly rectified that situation by launching additional SKUs with HD Graphics 3000.

This time around, Intel divides up 3D alacrity a little differently. All launched mobile and desktop Core i7s get HD Graphics 4000, and all but one (Core i5-3570K) desktop Core i5s get HD Graphics 2500.

Instead of 16 EUs, HD Graphics 2500 only offers six. Intel says to expect somewhere between 10-20%-better performance from HD Graphics 2500 compared to 2000. We have some i5s in the lab and will have a closer look at HD Graphics 2500 in the days to come.

What about HD Graphics 4000, though?

4. HD Graphics 4000: Performance In 3DMark 11 And Batman

Because 3DMark 11 requires DirectX 11 support, Sandy Bridge-based CPUs can’t run it. That leaves us with the Core i7-3770K, AMD’s A8-3850, and a discrete Radeon HD 6570.

Despite Intel’s efforts to double graphics performance, the Core i7-3770K cannot outmaneuver AMD’s Llano-based A8-3850.

Here’s the funny thing: every time Intel advances its on-board graphics technology, I talk to industry insiders who worry that the processor company’s progress will completely kill off the entry-level discrete market. But a Radeon HD 6570, available for under $60, still manages to double Ivy Bridge’s best effort. At least on the desktop, even inexpensive add-in boards still have their place.

The Physics suite, which measures processor performance, is particularly interesting. Tons of testing over the past year tells us that Intel’s Sandy Bridge cores get a lot more work done per clock cycle than AMD’s. Llano’s mediocre performance consequently isn’t news. However, the fact that Core i7-3770K picks up more than 2000 points with a discrete GPU installed is a good indication that the integrated core’s thermal requirements limit what the IA cores can do.  

In a real-world game like Batman: Arkham City, HD Graphics 4000 nearly does manage to catch AMD’s A8-3850. But generating playable average frame rates requires the title’s lowest possible detail settings and fairly unattractive resolutions. In contrast, even a cheap Radeon HD 6570 can handle entry-level detail at up to 1920x1080.

5. HD Graphics 4000: Performance In Skyrim And WoW

AMD’s A8-3850 holds onto its lead in Skyrim. Again, though, the real story isn’t a few frames per second separating integrated graphics processors. More disheartening is the fact that you have to dial down to 1280x720 and use detail settings that make five-year-old consoles look good.

While we’re anxious to see how AMD augments graphics performance with its Trinity-based APUs, there’s no reason to shy away from a discrete card on the desktop. If your budget isn’t flexible, shave off $60 from somewhere else and grab an add-in board for gaming.

World of Warcraft: Cataclysm tends to be very processor-bound. But an emphasis on graphics performance has an adverse effect on host processing, similar to what we saw in 3DMark 11’s Physics test. The result is that HD Graphics 4000 offers very little over HD Graphics 3000—despite the fact that the former enjoys additional performance from DirectX 11 mode.

A8-3850 looks comparatively strong, yielding modest performance all the way through 1920x1080 using the game’s middle-of-the-road Good quality preset.

Without question, though, an entry-level discrete card is still superior.

6. HD Graphics 4000: Native Compute Support

Ivy Bridge also includes the groundwork necessary to support OpenCL and DirectCompute 5.0.

But wait, didn’t Intel already release a driver for Sandy Bridge that enabled compute functionality? It did. HD Graphics 3000/2000 doesn’t actually support these APIs, though. They’re emulated and executed on the host processor, which is why compute workloads peg Sandy Bridge-based chips at 100% utilization.

HD Graphics 4000, on the other hand, supports FP32/FP64 under DirectCompute and FP32 in OpenCL. Intel currently lacks Khronos certification for ARB_gpu_shader_fp64, so it’s not enabled.

It’s immediately apparent how much faster native FP32 runs on HD Graphics 4000 compared to Sandy Bridge’s IA cores emulating OpenCL support. Because Sandra has to emulate FP64 through FP32, native double-precision performance looks a lot lower. However, that’s still GPU-only.

With Nvidia’s Kepler architecture capping FP64 at 1/24 single-precision performance, it’ll be interesting to see how HD Graphics 4000 compares to derivative discrete GPUs from Nvidia (particularly if Intel adds its own OpenCL FP64 extension).

Again, HD Graphics 3000 lacks native OpenCL support, so it’s missing from this chart. However, we see that Intel’s quad-core processors emulate OpenCL very well (albeit at full processor load and higher power consumption).

The discrete Radeon HD 6570 trails both Ivy and Sandy Bridge architectures, and is turn followed by HD Graphics 4000 with native OpenCL support. But instead of the IA cores completely tied up, CPU utilization sits at 0%, while power hovers 50 W lower. As far as performance per watt goes, that’s pretty darned impressive.

7. Quick Sync: A Secret Weapon, Refined

Back when Intel launched its Sandy Bridge architecture, I identified Quick Sync as the design’s secret weapon. Developed quietly for five years, it caught both AMD and Nvidia completely off guard. I projected that it’d take a year for both competitors to respond. And they have—AMD with its Video Codec Engine and Nvidia with NVEnc.

Unfortunately, AMD’s solution is still missing in action four months after it was first promised. Encoding on a Radeon HD 7000-series card has to be achieved through programmable shaders, rather than more energy-efficient fixed-function logic sitting idle on the die.

NVEnc is up and running, and a GeForce GTX 680 manages to outperform Intel’s first-gen Quick Sync implementation.

Nvidia’s victory is short-lived, though. HD Graphics blows everything else out of the water—and that’s even after biasing MediaEspresso toward quality rather than performance.

Arcsoft’s MediaConverter supports Quick Sync just fine, but the latest build doesn’t behave as well under APP or CUDA/NVEnc. Although scaling isn’t as aggressive, we still see how HD Graphics 4000 slices into the time it takes to transcode a large video file into something better suited to a portable device.

How’d They Do It?

I had the pleasure of sitting down with Dr. Hong Jiang, Intel’s chief media architect, before last year’s Sandy Bridge introduction to get an in-depth look at how the company implemented Quick Sync. This year, he led a session at IDF discussing the improvements included with Ivy Bridge. The focus, he said, was squarely on performance. Faster processing gives developers more flexibility in implementing higher-quality filters. It also punches through workloads more quickly, returning the processor to idle and saving power.

An increase in EU count helps Intel’s performance story, as does the inclusion of dedicated graphics L3 cache and greater Media Sampler throughput. Because the Media Sampler is part of that scalable third domain referred to as Slice, Intel can add resources in future generations to ratchet 3D and media performance up even more.

The Multi-Format Codec Engine (MFX) carries over from Sandy Bridge, enabling hardware-based H.264, VC-1, and MPEG-2 decoding, along with H.264 encoding. Intel apparently reworked its context-adaptive variable-length encoding and context-based adaptive binary arithmetic coding engines, though, which are both big mouthfuls referring to lossless encoding techniques that the MFX can decode faster.

Anticipating increasing demand for resolutions beyond 1080p in Ivy Bridge’s life cycle, Intel adds support to the MFX for 4096x4096 video decoding. In fact, Intel’s Jiang even claims the MFX can decode multiple 4K streams simultaneously.

Moving beyond Ivy Bridge’s decode capabilities, the media team also sought to improve the performance and quality of encoding tasks. A couple of paragraphs back I mentioned that the faster Media Sampler plays a part in Quick Sync’s speed-up. Specifically, it does the Motion Estimation stage’s heavy lifting, so greater throughput helps accelerate that step.

Now, Intel claims that its hardware-based encode solution achieves similar quality as a software solution. Last year, we wrote Video Transcoding Examined: AMD, Intel, And Nvidia In-Depth and found that the quality of every hardware-based encode engine sacrificed some degree of quality compared to a pure software solution. Faced with three new accelerated transcode technologies, we really need to spend some time putting each under a microscope to analyze how that story may have changed.

8. Platform Compatibility: Are Motherboard Vendors Ready?

All of Intel’s Ivy Bridge-based CPUs employ the existing LGA 1155 interface, setting up warranted questions about compatibility.

Naturally, the third-gen Core processors work right out of the box on 7-series platforms with Management Engine firmware version 8.x. Second-gen Core CPUs based on Sandy Bridge also drop right into motherboards with 7-series chipsets.

Support gets more conditional when you start talking about dropping third-gen Core chips on older boards with 6-series core logic. Each platform vendor is responsible for updating its H61-, H67-, P67-, and Z68-based motherboards with new Management Engine firmware, a BIOS, and graphics driver updates. Ivy Bridge CPUs are not supported on Q65-, Q67-, and B65-based boards.

Just days before the launch, the number of vendors with 22 nm-ready firmware is smaller than those without. Asus, Gigabyte, and Intel were able to get us updates for boards we have in the lab. MSI, EVGA, and Foxconn are purportedly working on updates. Biostar has firmware posted with 22 nm support, though the files are from late 2011 and we’re not sure if they accommodate retail boxed processors or not. ECShas firmware posted for a handful of its H61-based boards, but P67 and H67 support will come later. ASRock says it'll go live with its 22 nm-ready updates on launch day.

Here’s the thing to remember, though. If you plan on using an Ivy Bridge-based processor in an existing 6-series platform, be sure to update to a 22 nm-ready firmware with Sandy Bridge installed before ditching your old chip. To be clear, if you buy a 6-series board that doesn't have an Ivy Bridge-compatible firmware and then try to drop in a third-gen Core processor, it won't boot up.

PCI Express 3.0 On 6-Series Boards

Ivy Bridge-based CPUs include 16 lanes of PCI Express 3.0 connectivity. You can get 8 GT/s from one PCIe 3-ready add-in card on 6- and 7-series motherboards without any intervention at all. A couple of caveats apply, though.

First, if you’re using a 6-series board with a bridge chip like Nvidia’s NF200 to enable three- and four-way SLI, the switch limits you to second-gen PCIe signaling. Such is the case with Gigabyte’s Z68X-UD7-B3 motherboard. Nothing will ever get that board working at PCIe 3.0.

Other 6-series boards employ switches that automatically reconfigure PCI Express lanes in multi-card arrays, turning one 16-lane link into a pair of eight-lane connections, for example. In order to connect Ivy Bridge to dual Radeon HD 7000 or GeForce GTX 680 graphics cards at third-gen speeds, those switches have to support PCI Express 3.0 too. Otherwise you drop back to PCI Express 2.0 signaling.

Plug a card like AMD’s Radeon HD 7850 into something like Gigabyte’s Z68-powered G1.Sniper 2 with a Core i7-3770K, though, and you won’t have any problems at all.

9. Overclocking Ivy Bridge: Core i7-3770K Is A Mixed Bag

It’s easy to assume that shifting down to 22 nm lithography makes the Ivy Bridge design a more receptive platform for overclocking. Although it packs more transistors into less space, increasing density, the fact that the highest-end parts sport 77 W TDPs suggest lower thermal output.

At the same time, the 22 nm node is brand new, and Intel proudly admits that it changed more of its design than it normally would heading into a die shrink. That could spell trouble.

So, while it’s unclear how much truth is behind early reports that Ivy Bridge-based processors aren’t overclocking well, our own experiences in the lab certainly indicate a lack of consistency. Between three editors currently working with Core i7-3770K CPUs, two are able to hit 4.7 GHz and one saw frequencies as high as 4.9 GHz on air and using BIOS-specified voltages between 1.3 and 1.35 V. Curiously, though they seem to run Prime95 and Linpack without protest for hours on end, certain applications in our benchmark suite (mainly, 3ds Max) take these chips down within seconds. Achieving stability required dropping to 4.5, 4.5, and 4.77 GHz on our three samples.

In comparison, my boxed Core i7-2700K loads to Windows at 4.8 GHz and runs solidly at 4.7 GHz at 1.35 V.

The main difference between overclocking these two architectures, aside from Ivy Bridge appearing comfortable at just slightly slower frequencies, is that Intel’s 22 nm-based chips also run significantly warmer (somewhere around 10°C higher at each CPU’s maximum stable overclock). Thus, overclockers using big air will likely have the most difficult time nailing settings that are suitably aggressive, but still modest enough to promote longevity.


Ivy Bridge
Sandy Bridge-E
Sandy Bridge
Core Ratio Multipliers
Up to 63x
Up to 57x
Up to 59 x
Real-Time Core Ratio Changes
Yes
All Cores Simultaneously
No
Integrated Graphics Ratio Multipliers
Up to 60x
N/A
Up to 57x
Real-Time Graphics Ratio Changes
Yes
N/A
No
DDR Ratio/Frequency
Up to DDR3-2667
Up to DDR3-2400
Up to DDR3-2133
DDR Granularity
200/266 MT/s
266 MT/s
266 MT/s
XMP Reference Code
v.1.3
v.1.2 or 1.3
v.1.2
BCLK Overclocking
Limited
1.0/1.25/1.67x Straps
Limited


The phase/dry ice/LN2 crowd is in a different position. Enthusiasts able to keep Ivy Bridge cold have already exceeded 6.9 GHz with four cores active, partly because Intel increased Sandy Bridge’s 59x maximum core ratio to 63x. Because Z77 suffers the same limited BCLK headroom as Z68 and P67, most reports put 107 or 108 MHz as the practical limit. But boards able to hit 110 or 111 MHz couple well with higher multipliers and extreme cooling.

What can you expect from a Core i7-3770K at 4.5 GHz compared to a Core i7-2700K at 4.7 GHz? Nearly exactly the same performance, it turns out. Though I’m not entirely certain you’d want to leave the -3770K running with one core at 88 degrees for very long, it is stable.

In 3ds Max, which is very well threaded, overclocked Ivy Bridge finishes our workload just two seconds faster, narrowing a five-second gap at stock clock rates. iTunes puts the two chips just one second apart. At 3.5 GHz, they’re separated by three seconds.

Consider this a cursory look at overclocked performance in single- and multi-threaded tests. Our German team is working on a more in-depth overclocking analysis that should go live sometime this week.

10. Ivy Bridge Memory Scaling

A deliberate effort went into extending Ivy Bridge’s memory overclocking ceiling and adding more granular settings. In theory, scaling memory bandwidth up could have a big impact on integrated graphics performance (more throughput certainly helped AMD’s A8-3850). So, is it worth spending extra on modules rated for higher data rates?

A synthetic like Sandra 2012 demonstrates sizable gains. Bandwidth literally doubles as we move from two channels of DDR3-1066 to DDR3-2133.

Real-world performance improvements trail off a lot faster though, likely because memory isn’t the most debilitating bottleneck.

Assuming you use a high-end Ivy Bridge chip with a discrete GPU, does faster memory affect our other benchmarks? WinRAR is notoriously sensitive to bandwidth changes, and it easily shows where more throughput helps…and where it ceases to make a difference.

It definitely makes sense to buy a DDR3-1600 kit, and even DDR3-1866 nudges performance forward a little. Stepping up to DDR3-2133 really doesn’t do anything though.

On the other end of the spectrum, well-threaded compute-intensive titles like 3ds Max give you nothing back in return for installing faster memory.

11. Test Setup And Benchmarks
Test Hardware
Processors
Intel Core i7-3770K (Ivy Bridge) 3.5 GHz (35 * 100 MHz), LGA 1155, 8 MB Shared L3, Hyper-Threading enabled, Turbo Boost enabled, Power-savings enabled

Intel Core i7-3930K (Sandy Bridge-E) 3.2 GHz (32 * 100 MHz), LGA 2011, 12 MB Shared L3, Hyper-Threading enabled, Turbo Boost enabled, Power-savings enabled

Intel Core i7-3960X (Sandy Bridge-E) 3.3 GHz (33 * 100 MHz), LGA 2011, 15 MB Shared L3, Hyper-Threading enabled, Turbo Boost enabled, Power-savings enabled

AMD FX-8150 (Zambezi) 3.6 GHz (18 * 200 MHz), Socket AM3+, 8 MB Shared L3, Turbo Core enabled, Power-savings enabled

AMD Phenom II X6 1100T (Thuban) 3.3 GHz (16.5 * 200 MHz), Socket AM3, 6 MB Shared L3, Turbo Core enabled, Power-savings enabled

Intel Core i7-2700K (Sandy Bridge) 3.5 GHz (35 * 100 MHz), LGA 1155, 8 MB Shared L3, Hyper-Threading enabled, Turbo Boost enabled, Power-savings enabled

Intel Core i5-2550K (Sandy Bridge) 3.5 GHz (35 * 100 MHz), LGA 1155, 6 MB Shared L3, Turbo Boost enabled, Power-savings enabled
Thermal Paste
Zalman ZM-STG1
Motherboard
Intel DX77GA-70K (LGA 1155) Intel Z77 Express Chipset, BIOS GA.3254

Gigabyte X79-UD5 (LGA 2011) Intel X79 Express Chipset, BIOS F10

Gigabyte 990FXA-UD5 (Socket AM3+) AMD 990FX/SB950 Chipset, BIOS F7

Gigabyte A75-UD4H (Socket FM1) AMD A75 Chipset, BIOS F7

Gigabyte X68X-UD7-B3 (LGA 1155) Intel Z68 Express Chipset, BIOS F10 (for compatibility testing)

Gigabyte G1. Sniper 2 (LGA 1155) Intel Z68 Express Chipset, BIOS F5 (for compatibility testing)
Memory
G.Skill 16 GB (4 x 4 GB) DDR3-1600, F3-12800CL9Q2-32GBZL @ 9-9-9-24 and 1.5 V

Kingston 4 GB (2 x 2 GB) DDR3-2133, KHX2133C9AD3T1K2/4GX @ up to DDR3-2133 1.65 V for memory overclocking
Hard Drive
Intel SSD 510 250 GB, SATA 6 Gb/s
Graphics
Nvidia GeForce GTX 680 2 GB
Power Supply
Cooler Master UCP-1000 W
System Software And Drivers
Operating System
Windows 7 Ultimate 64-bit
DirectX
DirectX 11
Graphics DriverNvidia GeForce Release 301.10


We used Intel's new DZ77GA-70K for our LGA 1155 testing, but pulled out a number of boards from Gigabyte for testing compatibility of older Z68-based platforms.

Game Benchmarks And Settings
Batman: Arkham City
Game Settings: High Quality Settings, Anti-Aliasing: Disabled/8x MSAA, V-sync: Disabled, DirectX 11 Mode, 1680x1050, 1920x1200, 2560x1600, Built-in Benchmark
The Elder Scrolls V: Skyrim
Game Settings: High Quality (8x AA / 8x AF) / Ultra Quality (8x AA, 16x AF) Settings, FXAA disabled, V-sync: Disabled, 1680x1050 / 1920x1080 / 2560x1600, 25-second playback, Fraps
World of Warcraft: Cataclysm
Game Settings: Ultra Quality Settings, Anti-Aliasing: 1x AA and 8x AA, Anisotropic Filtering: 16x, Vertical Sync: Disabled, 1680x1050, 1920x1080, 2560x1600, Demo: Crushblow to The Krazzworks, DirectX 11
Audio Benchmarks and Settings
iTunesVersion: 10.4.10, 64-bit
Audio CD ("Terminator II" SE), 53 min., Convert to AAC audio format
Lame MP3Version 3.98.3
Audio CD "Terminator II SE", 53 min, convert WAV to MP3 audio format, Command: -b 160 --nores (160 Kb/s)
Video Benchmarks and Settings
HandBrake CLIVersion: 0.9.5
Video: Big Buck Bunny (720x480, 23.972 frames) 5 Minutes, Audio: Dolby Digital, 48 000 Hz, Six-Channel, English, to Video: AVC Audio: AC3 Audio2: AAC (High Profile)
MainConcept Reference v2.2
Version: 2.2.0.5440
MPEG-2 to H.264, MainConcept H.264/AVC Codec, 28 sec HDTV 1920x1080 (MPEG-2), Audio:
MPEG-2 (44.1 kHz, 2 Channel, 16-Bit, 224 Kb/s), Codec: H.264 Pro, Mode: PAL 50i (25 FPS), Profile: H.264 BD HDMV
Application Benchmarks and Settings
WinRARVersion: 4.11
RAR, Syntax "winrar a -r -m3", Benchmark: 2010-THG-Workload
WinZip 16Version: 16.0 Pro
WinZip CLI, Benchmark: 2010-THG-Workload
7-Zip
Version 9.22 beta
LZMA2, Syntax "a -t7z -r -m0=LZMA2 -mx=5", Benchmark: 2010-THG-Workload
Adobe Premiere Pro CS 5.5
Paladin Sequence to H.264 Blu-ray
Output 1920x1080, Maximum Quality, Mercury Playback Engine: Software Mode
Adobe After Effects CS 5.5
Version: CS5.5
Tom's Hardware Workload, SD project with three picture-in-picture frames, source video at 720p, Render Multiple Frames Simultaneously
BlenderVersion: 2.62
Syntax blender -b thg.blend -f 1, Resolution: 1920x1080, Anti-Aliasing: 8x, Render: THG.blend frame 1, Cycles renderer and internal tile renderer (9x9)
Adobe Photoshop CS 5.1 (64-Bit)Version: 11
Filtering a 16 MB TIF (15 000x7266), Filters:, Radial Blur (Amount: 10, Method: zoom, Quality: good) Shape Blur (Radius: 46 px; custom shape: Trademark sysmbol) Median (Radius: 1px) Polar Coordinates (Rectangular to Polar)
ABBYY FineReaderVersion: 10 Professional Build (10.0.102.82)
Read PDF save to Doc, Source: Political Economy (J. Broadhurst 1842) 111 Pages
3ds Max 2012
Version: 10 x64
Rendering Space Flyby Mentalray (SPECapc_3dsmax9), Frame: 248, Resolution: 1440 x 1080
Adobe Acrobat X Professional
PDF Document Creation (Print) from Microsoft PowerPoint 2010
SolidWorks 2010
PhotoView 360
Render 01-Lighter Explode.SLDASM (SolidMuse.com)
Image Output Resolution: 1920x1080, Render: Preview Quality “Good”, Final Render Quality “Best”
Visual Studio 2010
Compile Chrome project (1/31/2012) with devenv.com /build Release
Synthetic Benchmarks and Settings
PCMark 7Version: 1.0.4
3DMark 11
Version 1.0.3
SiSoftware Sandra 2012 SP3
CPU Test=CPU Arithmetic/Multimedia, Memory Test=Bandwidth Benchmark, Cryptography, Cache Latency
12. Benchmark Results: PCMark 7

Right out of the gate, Intel’s fastest Ivy Bridge-based processor, Core i7-3770K, appears to be a stunner, coasting right past Intel’s thousand-dollar Core i7-3960X.

But the Core i7-2700K also appears faster than the Sandy Bridge-E-based flagship. What gives?

Well, certain parts of PCMark 7 are optimized for at least 16 threads, we’re told by Futuremark. That’s evident in the Computation and Creativity suites below, where the -3960X does particularly well. Otherwise, the components of Windows 7 used to create Futuremark’s metric appear predominantly optimized for quad-core parts running at higher clock rates.

Of course, if our real-world benchmark suite tracks with PCMark, it’d be a lot harder to recommend Core i7-3930K. But I have the feeling that some of our other tests won’t be as friendly to the quad-core contenders.

13. Benchmark Results: 3DMark 11

Six-core CPUs help Nvidia’s GeForce GTX 680 stretch its legs in the 3DMark 11 Performance suite, and both Sandy Bridge-E-based chips take first and second place, barely edging out the Core i7-3770K and -2700K CPUs at higher clock rates.

The Core i5-2550K and AMD’s two entries trail by a more substantial margin.

Neither the Graphics nor the Combined suites are particularly conclusive, which isn’t surprising in light of the GPU that remains constant throughout. However, the Physics subtest and Physics frame rate metric are both intended to differentiate processors. A heavy emphasis on threading does just that, too.

14. Benchmark Results: Sandra 2012 SP3

More so than PCMark or 3DMark, Sandra is good for extracting the potential performance from each of these CPUs.

Clearly, the six-core Sandy Bridge-E parts serve up the highest numbers in Sandra’s Arithmetic test. Ivy Bridge’s slight IPC improvements translate into a comparably slim advantage over Core i7-2700K, and both of those parts are substantially faster than the remaining three processors.

I broke out the multimedia results into two charts, since Phenom II doesn’t support AVX. Integer performance is a strong suit for AMD’s FX processor, which makes sense given its eight integer cores. Floating-point throughput isn’t as compelling. Again, that’s expected given the shared nature of the architecture’s resources devoted to floating-point math.

Leveraging AVX instructions, the eight-core FX is even able to outperform Intel’s Core i7-3770K in the integer component of this test. AMD is humbled by the floating-point portion, though, succumbing even to Intel’s Core i5-2550K. Its four modules (with four corresponding floating-point units) cannot keep up.

Ivy Bridge, meanwhile, is only marginally faster than Sandy Bridge at the same clock rate.

The clear winners here are the Core i7-3960X and 3930K, which brute-force the test using six cores.

Hardware support for AES encryption and decryption means that the Ivy Bridge, Sandy Bridge, and Bulldozer architectures can operate on data as fast as it’s delivered. Thus, the AES256 results scale according to memory bandwidth—the bottleneck in this equation. SHA hashing is a much slower process, which is dominated by the Sandy Bridge-E-based parts.

And here are the bandwidth numbers to drive our Cryptography test results home. Operating at the same DDR3-1600 data rate, Ivy Bridge doesn’t demonstrate any advantage over Sandy Bridge, and in fact realizes slightly less overall throughput.

During the course of my review, SiSoftware issued Sandra 2012 SP4 based on discussions between Adrian Silasi and I about how the software addresses the clock rates of CPUs with dynamic frequency adjustment like Turbo Boost and Turbo Core.

Because I had all of my data compiled using SP3, though, I decided to use the older build, which reports results based on rated clocks, rather than actual clocks. And really, that decision was made based on the fact that Ivy Bridge and Sandy Bridge demonstrate nearly identical behavior, suggesting very similar cache latencies. Moving forward, I’ll swap over to SP4. But for now, the desired effect is achieved.

15. Benchmark Results: Adobe CS 5.5

Here, the -3770K ties its predecessor, making for a wholly unimpressive finish.

Not surprisingly, both six-core Intel CPUs turn in the fastest times. But it actually is quite a shock to see AMD’s FX-8150 slotting in ahead of the Bridges, particularly after finishing so far behind in the previous tests. It just goes to show that Bulldozer isn’t beyond the point of redemption.

Officially, Nvidia’s GeForce GTX 680 is too new to be supported by Adobe’s Mercury Playback Engine. Off the record, it’d still be pretty easy to add the card to Premiere Pro’s text document-based list of compatible cards.

At any rate, the current lack of support makes it really easy for us to stack CPUs up against each other without interference from a graphics card. And from that experiment, we again see the six-core Sandy Bridge-E parts tearing through our Paladin workload. Ivy Bridge and Sandy Bridge take third and fourth place, separated by 36 seconds.

The gap between the rest of the field is far more substantial. Curiously, AMD’s FX falls to the bottom of the stack, unable to overcome its six-core predecessor, Phenom II X6 1100T.

Separation between the Intel-based contenders is narrow. But the gap between the slowest Intel chip, Core i7-3930K, and the fastest AMD processor, Phenom II X6 1100T, is larger.

16. Benchmark Results: Content Creation

Well-optimized for threading, the two six-core Sandy Bridge-E-based chips run away with first and second place in 3ds Max 2012.

Core i7-3770K takes third place, narrowly outperforming the existing Core i7-2700K. The implication here is that Ivy Bridge is only getting a marginal speed-up due to the IPC-oriented improvements that Intel folded in. A difference of five seconds is hardly noteworthy.

The new Cycles engine wouldn’t cooperate with our AMD-based platforms, so we have the results from five different machines with Intel CPUs in them. Both six-core models land in the lead, followed by the Core i7-3770K and -2700K nearly tied for third place. The Core i5-2550K, lacking Hyper-Threading and some shared L3 cache, trails a ways back.

We didn’t have any issues using the older rendering engine on any of our platforms, though the finishing order is exactly the same amongst the Intel-based CPUs. Factoring in AMD’s Phenom II X6 1100T and FX-8150 only tacks those two parts onto the end of the chart.

Same story in SolidWorks. The six-core Intel chips dominate, followed by the Core i7-3770K and -2700K, seemingly joined at the hip. Intel’s Core i5-2550K lands in fifth place, trailed by both AMD processors. Despite the fact that we saw FX-8150 do so stinking well in Photoshop, it simply gets outclassed in all of our content creation titles.

17. Benchmark Results: Productivity

The ongoing pattern seems consistent enough that I could probably get away without any analysis here at all. In a threaded title like FineReader, the six-core Sandy Bridge-E chips walk away with an advantage.

Behind them, quad-core Ivy and Sandy Bridge-based chips running at the same clock rate (Core i7-3770K and -2700K, more specifically) land a few seconds apart, but always with Ivy Bridge in the lead. 

Meanwhile, AMD’s Phenom II and FX flagships seem content to do battle with Core i5-2550K. Though, in this case, the FX springs up into the lead ahead of Intel’s mainstream processor.

I typically don’t use Fritz in my processor launch stories, but I’ve received enough requests to reincorporate it that I’m including results in today’s piece.

Sandy Bridge-E clearly rules, while the IPC optimizations inherent to Ivy Bridge give it the slightest edge over Sandy Bridge at a given clock rate.

Aggressive threading allows FX-8150 to outpace Phenom II X6, as both AMD chips outrun Intel’s Core i5-2550K.

A demanding Google Chrome compile workload takes more than 30 minutes on three of today’s contenders. The two six-core Sandy Bridge-E-based parts get the job done in well under 20 minutes.

Core i7-3770K and -2770K fall somewhere in between, separated by just 22 seconds.

Our PDF document creation is decidedly single-threaded. Naturally, then, Intel’s Ivy and Sandy Bridge architectures do particularly well, while the lower IPC of AMD’s designs cannot be overcome by clock rate.

The slight improvements in Ivy Bridge are enough to give Core i7-3770K a first-place finish. Intel’s two Sandy Bridge implementations follow closely behind.

18. Benchmark Results: File Compression

Recently adopting WinZip 16 doesn’t seem to have helped with the software’s inability to utilize multi-core processors effectively. As such, Ivy Bridge’s slight IPC advantage garners it another top finish, just ahead of the Core i7-3960X and -2700K. Fortunately, we have inside information that indicates WinZip’s next incarnation will be more competitive.

More aggressive optimizations for parallelism give Intel’s Core i7-3960X an edge in WinRAR, allowing it to reclaim the top spot, though the quad-core Ivy Bridge chip does manage to tie the $600 Core i7-3930K.

Let’s not mince words, though. Core i7-2700K is just one second behind the new chip, and the $220 Core i5-2550K is just two seconds back. You certainly won’t have much to complain about if saving some money is more important than waiting a couple of extra seconds for a folder full of files to compress.

Long known as the file compression/decompression utility in our suite best optimized for multi-core processors, 7-Zip puts both Sandy Bridge-E based CPUs in the lead, followed closely by Core i7-3770K.

The real loser here is Core i5, which sports four cores, but lacks Hyper-Threading, landing it in last place. AMD’s Bulldozer-based FX-8150 fares better in the face of more aggressive threading, though it can’t quite catch the Core i7-2700K.

19. Benchmark Results: Media Encoding

Lame is another single-threaded title that we know will reward the most efficient architecture running at the highest clock rate.

Ivy Bridge, enjoying subtle IPC-oriented improvements, scores a narrow victory when Turbo Boost kicks in.

Because Core i7-3960X and Core i7-2700K hit similar single-threaded clocks, they show similarly in our Lame test.

Without overclocking, the Core i7-3930K can’t quite keep up to Intel’s Core i5-2550K (after all, six cores isn’t an advantage in a test only capable of taxing one).

We see a very similar situation play out in iTunes. Some of the Sandy Bridge-based chips swap places, but they’re all very, very close together. Most notable is that the Core i7-3770K takes first yet again.

The performance you can expect from Ivy Bridge is fairly easily characterized in one of two ways: lightly-threaded apps that favor an efficient architecture tend to show the design in a positive light against competing architectures at the same clock rate, while more parallelized workloads favor Sandy Bridge-E, so long as it wields more cores.

You’ll notice I left the -3820 out of this review altogether. I cannot come up with any situation where the -3820 is a product I’d recommend, even in a world without Ivy Bridge. If you’re going to spend big on Sandy Bridge-E, go for a six-core model, at least. If not, Sandy Bridge, Z68, and dual-channel memory kits are a better buy.

MainConcept illustrates the reason why nicely. At stock clocks, there’s a nice, gradual progression from -3960X, -3930K, -3770K, and -2700K. The larger drop-off happens when you shift down to the -2550K and AMD’s offerings.

The exact same conclusion applies to HandBrake, a front-end for the x264 encoder. Intel’s six-core chips rock, though it’s easy to get good performance from the quad-core, Hyper-Threaded models when cash is more of a concern.

20. Benchmark Results: Batman: Arkham City

Given the decidedly graphics-bound nature of Crysis 2, I decided to give Batman: Arkham City a shot. As it turns out, the title actually is somewhat sensitive to processor performance at 1680x1050 and 1920x1080, so long as anti-aliasing is disabled.

At those resolutions, higher clock rates seem more important than the difference between four or six cores. It just so happens that the first-place Core i7-3960X enjoys the benefit of a 3.9 GHz top Turbo Boost bin, plus a large 15 MB shared L3 cache.

You’ll notice that both AMD chips lag behind at our two most differentiated resolutions. Turn on anti-aliasing or bump up to 2560x1600, though, and the delta narrows substantially. Bottom line: FX and Phenom II both emerge as bottlenecks long before Intel’s Core processors. But it takes a $500 graphics card to uncover the difference most dramatically.

21. Benchmark Results: The Elder Scrolls V: Skyrim

Given its mainstream appeal, Skyrim was developed to scale across a broad range of platforms. Although its most demanding graphics settings easily slow down a single GeForce GTX 680, there is still plenty of host processor influence to extrapolate through 2560x1600.

At 1680x1050 and 1920x1080, lofty Turbo Boosted clock rates and a big shared L3 cache help propel the Core i7-3960X to a symbolic victory in a second consecutive game. The Core i7-3770K’s second-place finish is consistent across both resolutions, though it’s just as insignificant considering the narrow performance spread.

More troubling is the significant drop-off experienced by the two AMD CPUs. Even at 2560x1600 using the Ultra quality preset, an FX-8150 holds our GeForce GTX 680 back from achieving its average frame rate potential.

22. Benchmark Results: World Of Warcraft: Cataclysm

The same phenomenon plagues the Phenom II and FX in World of Warcraft. Only now, the FX manages to deliver marginally-better numbers.

Meanwhile, Core i7-3960X and -3770K trade blows at 1680x1050 and 1920x1080, the former dominating with anti-aliasing turned off as the latter claims a small victory under higher graphics loads.

Once we hit 2560x1600, Intel’s Ivy Bridge-based part snags first place in both situations, though the discrepancy in average frame rate between the -3770K and -2550K is nearly negligible with anti-aliasing enabled (and plenty playable at 80+ FPS).

23. Power Consumption And Efficiency

We now know that, as a host processor, Core i7-3770K really isn’t that much faster than Core i7-2700K. But we also know that Ivy Bridge should be capable of delivering better (or at least similar) performance in a significantly lower thermal envelope. In this case, the -3770K is a 77 W part, whereas Core i7-2700K is a 95 W processor and the Sandy Bridge-E-based chips are rated for 130 W.

Rather than simply letting each chip sit on the Windows desktop and then spin them all up, reporting idle and load consumption figures, we instead created a script using all of the tests in our benchmark suite (aside from the games and the Google Chrome compilation) with short rests in between. The resulting metric is more than an hour long on even the fastest processor in our collection, and it’s truly a real-world mix of idle and load.

It’s frankly pretty difficult to make sense of the raw, logged data, even with our normal chart expanded out to make the hour-plus run more readable. Certain segments make it quite clear, however, that the Core i7-3770K is the lowest-power contender tested, followed by the Core i7-2700K, the Core i7-3930K, the Phenom II X6 1100T, and finally AMD’s FX-8150. What about the -3960X? We have the data for that one as well, but the additional line makes this chart even more of a mess, so I left it off. I also have data for the -2550K, but I’m working on a surprise with that information.

There we go. It’s much easier to average out power consumption during each run and compare the result. As hypothesized using the line graph, we see each CPU land exactly where we expected it to, with the -3960X falling in just behind the -3930K.

It’s pretty darned easy to measure how long each sequence runs, since our logger takes a sample once every two seconds. Not surprisingly, the Core i7-3960X finishes first. It’s more interesting that the Core i7-2700K is just slightly quicker through the benchmarks than Core i7-3770K. Both AMD trail quite a ways behind.

Multiplying average power by the time required to complete the benchmark gives us total energy used in watt-hours, reflecting efficiency.

The Core i7-3770K’s solid performance, coupled with a reduction in power consumption, results in the lowest power use in the course of our suite. The Core i7-2700K follows closely behind, though.

Both of the Sandy Bridge-E chips facilitate impressive performance numbers throughout our testing, excelling particularly in threaded workloads. However, 130 W TDPs penalize the LGA 2011-based CPUs when it comes to power consumption, so they fall into third and fourth place.

The Phenom II X6 and FX bring up the rear, as they both use more power and underperform the Intel competition. It’s an unfortunate state of affairs for AMD, but the Piledriver core is being prepped for operation in the company’s next APU design. Hopefully, changes made to Bulldozer ameliorate some of what we’ve come to dislike about today’s FX.

24. How Much Faster Is Core i7-3770K Than -2700K And i5-2550K?

As a host processor, Core i7-3770K is only marginally faster than the former flagship of Intel’s Sandy Bridge family, Core i7-2700K.

Sure, there are gaming numbers from HD Graphics 4000 and improved Quick Sync results we could talk about here, but this $317 chip’s job is as a CPU, first and foremost.

If a less expensive Core i5-2550K was good enough for you in a world where Core i7-2700K represented the fastest LGA 1155 processor you could buy, than it should be good enough for you in a world suddenly populated by Ivy Bridge-based chips, too.

Clearly there are workloads where buying an i7 warrants spending $100 more, though. Premiere Pro render jobs and Visual Studio projects are some of the most taxing in our suite—both get big boosts from the additional L3 cache and Hyper-Threading support offered by the highest-end mainstream CPUs.

25. An Evolution That Makes Sense, But Doesn't Impress

A group of journalists recently went to visit AMD in Austin (including one of our writers) and came back talking about the value of an “experience,” and how benchmarks can’t tell you if a given piece of hardware is “good enough.”

I believe benchmarks are important and will remain the lowest-level tools for quantifying one component’s value over another. They’re the most precise measure of “good enough.” You can look at performance numbers and generalize for a broad audience using hard data. It’s not as easy to tell how long you’ll spend compiling code based on one person’s opinion that a workstation is fine and dandy, though.

But that’s a conversation for another day. Regardless of which side of the fence you find yourself, Core i7-2700K is subjectively “good enough” compared to Intel’s Ivy Bridge-based Core i7-3770K. No question. If you’re die-hard about data, the numbers also make it objectively clear that there is no reason to upgrade a high-end desktop Sandy Bridge CPU to a high-end Ivy Bridge CPU.

Intel succeeds in bolstering the performance of its integrated graphics solution, but insofar as HD Graphics 4000 applies to gamers from any walk of life, you’d really be selling yourself short by not complementing a ~$300 CPU with an add-in card. Although AMD’s A8-3850 is nowhere near as fast as Core i7-3770K in processing workloads, the $130 APU does deliver better frame rates, if entry-level gaming is all you need.

Can Core i7-3770K catch a break with power users eager to overclock? Unless you’re using an extreme form of cooling, I’m afraid not. Our boxed Core i7-2700K hit a more aggressive frequency, nearly matching the -3770K’s performance in the process.

What if you saw the award that Core i5-2500K won last year in Intel’s Second-Gen Core CPUs: The Sandy Bridge Review but didn’t upgrade? What if you’re still stuck on an old Core 2- or Phenom-based platform and need something new? In that case, of course a desktop Ivy Bridge-based chip makes more sense than buying what is now last-generation hardware. The Core i7-3770K is one option, but we’d also be fairly confident in a Core i5-3570K for $100 less, too. Intel is actually being really reasonable on pricing, so you’ll pay less for the i7-3770K than you would have for a -2700K yesterday, and less for an i5-3570K than the -2550K. Not bad at all. 

A Little Perspective

Although Core i7-3770K, as one model in Intel’s line-up, is fairly easy for enthusiasts with modern machines to dismiss, don’t take our judgment as a cloud over the Ivy Bridge architecture.

An emphasis on integrated graphics performance and lower thermal design power points makes it clear that Intel is out to conquer smaller form factors like all-in-one desktops and thin/light notebooks.

Soon, the first wave of Ultrabooks based on Ivy Bridge, code named Chief River, will wash over the mobility-obsessed masses, more accurately representing the purpose of Intel’s newest family of processors.

But before that happens, we have more Ivy Bridge-based coverage planned, including our first round-up of Z77 Express-based motherboards driven by a Core i7-3770K, a look at how four different Ivy Bridge-based Core i5s compare at as many thermal ceilings, more depth on overclocking performance, and a review of mobile Ivy Bridge in a brand new notebook. Stay tuned!

For a chance at winning a Core i5-3570K-based PC, please click this link to enter our CyberPower PC/Tom's Hardware sweepstakes. The system's specs are as follows:

  • Intel Core i5-3570K 3.40 GHz (Overclocked 4.5GHz)
  • NZXT Switch 810 Hybrid Full-Tower Gaming Case
  • Asus P8Z77-V LX LGA 1155 ATX Mainboard
  • AMD Radeon HD 7770 1 GB PCI Express 3.0 Video Card
  • Asetek 570LX Liquid Cooling System w/ 240 mm Radiator and Dual Fans
  • Corsair Vengeance 8 GB DDR3-1600 Dual-Channel Memory Kit
  • Corsair Force Series 60 GB SATA 6Gb/s SSD
  • 1 TB SATA 6Gb/s 32 MB Cache 7200 RPM HDD
  • NXZT 750 W Power Supply 
  • Microsoft Windows® 7 Home Premium  
  • 24x Dual-Layer Dual-Format Drive

The sweepstakes opens on April 23, 2012 9 AM PDT and closes May 7, 2012 9:00 AM PDT.