I’ve been writing about the “new” Intel for nearly three years now. It’s been so long that I almost forgot what the old Intel was like. It’s not that the old Intel wasn’t competitive performance-wise, it’s that the old Intel wasn’t pleasant to work with. The old Intel was the one that always thought the Pentium 4 was the fastest thing on the planet, even when it wasn’t. The old Intel wasn’t forthcoming with information and acted like it worked in a world where it had no competition. The old Intel wasn’t a very good Intel.

The new one is nothing like that. We get open sharing of information, real discussion about AMD’s strengths and weaknesses and it also helps that we also get the world’s fastest microprocessors with it.

We’ve seen that the new Intel can stand the test of time, at least over the past three years. But can the new Intel last when it’s not always winning reviews? Sure, Intel’s Core i7 remains untouched but what about at cheaper price points? Last month we found out that Intel is quite competitive at the $70 with its Pentium E5300. But between the $70 E5300 and the $280 Core i7-920 there are a few price points where AMD is recommendable.

The question then becomes how does the new Intel deal when it isn’t the fastest on the market?

Surprisingly well it turns out.

This is the Core 2 Quad Q8400:

It’s a quad-core chip running at 2.66GHz with a 2MB L2 per pair of cores (4MB total). It’s like two Pentium Dual Core processors on a single package.

These chips actually have a 6MB L2 but with 2MB disabled either because they have irrecoverable defects in the cache or simply to hit the right price point. In other words, the Q8400 is literally a Q9400 but with 2MB of its L2 disabled.

The Q8400 is Intel’s most recent response to the Phenom II X4 940. Initially AMD priced the 940 similarly to the Q9400. Then, Intel cut prices so that the 940 had to compete with the much faster Q9550. AMD responded, with a price cut that put the 940 on par with the Q9400 again.

Once AMD began shipping Socket-AM3 Phenom IIs, it dropped the prices on its Socket-AM2+ parts once more. This wasn’t so much a price cut but rather a gradual phasing out of the AM2+ CPUs, eventually I expect an AM3-only market since those chips can also work in AM2+ boards.

Rather than take the bait and drop the Q9400 prices once more, Intel instead responded with the introduction of a similarly priced Q8400 at $183.

AMD’s Phenom II X4 940 was generally the same speed if not faster than Intel’s Core 2 Quad Q9400. With less cache, the Q8400 shouldn’t perform any better than the Q9400, so the question is - does it perform any worse?

Then there’s power consumption to worry about and overclocking, but we’ll get to those in due time. Let’s just say that the situation is far more complex than it seemed at first sight.

Core 2 Quad Q9400 vs. Q8400: An Extra $30 buys you 6%

I hope you guys like these charts, but what I’m showing here is at a glance how much faster the Q9400 is than the Q8400. The results are fairly telling; in 3D rendering and video encoding, the extra cache doesn’t really matter. In games, it’s responsible for a 2 - 7% performance increase. The Excel Monte Carlo simulation and our WinRAR test however both show significant gains. If you average out all of the applications where there is a performance gain, the Q9400 manages a 6% advantage over the Q8400. The SYSMark 2007 results pretty much echo that.

The price difference between the Q9400 and the Q8400 is $30. That $30 will buy you an extra 6%. With the Q9400 already faster than the Q8400, the Phenom II X4 940 shouldn’t have a problem competing against the Q8400.

Phenom II Earns a Financially Troubled AMD Less per Chip than Core 2 Quad

The global economy isn’t exactly strong right now. People are still buying, just not nearly as much as before when it felt like money grew on trees. Financially, all companies have been hurt, but AMD has much bigger issues. The table below shows net income in millions of US dollars before taxes for AMD and Intel over the past four quarters:

Net Income Before Taxes	Q1 2009	Q4 2008	Q3 2008	Q2 2008
AMD	-$298 Million	-$1,358 Million	$22 Million	-$682 Million
Intel	$629 Million	$369 Million	$2,833 Million	$2,313 Million

 

Yeah. Ouch. Granted Intel going from ~$2.8B of income in a quarter down to under $400M must’ve hurt, but AMD has lost over $2.3B in the past four quarters. The company isn’t profitable and unfortunately is stuck between a rock and a hard place when it comes to fixing that.

The financial issues extend beyond simple CPU sales but I must point out the obvious issue with AMD’s current strategy. We know that the Phenom II is competitive, but look at what it’s competing against:

Processor	L1 Cache	L2 Cache	L3 Cache	Total Cache (4-core)	Transistor Count	Die Size
AMD Phenom II	128K per core	512KB per core	6MB	8.5MB	758M	258mm2
Intel Core 2 Quad Q9x50	64KB per core	12MB	N/A	12.25MB	820M	214mm2
Intel Core 2 Quad Q9x00	64KB per core	6MB	N/A	6.25MB	456M	164mm2
Intel Core 2 Quad Q8xxx	64KB per core	4MB	N/A	4.25MB	456M	164mm2

 

Every single Phenom II uses a single 258 mm2 45nm die, that’s nearly Nehalem-sized. The problem is that the Phenom II parts generally compete against Intel’s Core 2 Quad Q9x00 and Q8xxx series, both of which have a total die size of 164mm2. AMD’s Phenom II die is 57% larger.

AMD and Intel both manufacture on 300mm wafers, but Intel can get nearly 60% more CPUs for each wafer than AMD can thanks to its die size advantage. That translates into more revenue per wafer and a significant profit advantage for Intel.

AMD’s Phenom II is very competitive, but the strategy does not have much long term staying power. AMD needs to introduce smaller die versions of its CPUs soon.

The deeper ramifications of AMD’s current situation are troubling. I’m not sure what impact all of this is having on the development of AMD’s next-generation architectures, but I suspect that it can’t be good.

The Test

Motherboard:	Intel DX48BT2 (Intel X48) MSI DKA790GX Platinum (AMD 790GX)
Chipset:	Intel X48 AMD 790GX
Chipset Drivers:	Intel 9.1.1.1010 (Intel) AMD Catalyst 8.12
Hard Disk:	Intel X25-M SSD (80GB)
Memory:	G.Skill DDR2-800 2 x 2GB (4-4-4-12) G.Skill DDR2-1066 2 x 2GB (5-5-5-15) Qimonda DDR3-1066 4 x 1GB (7-7-7-20)
Video Card:	eVGA GeForce GTX 280
Video Drivers:	NVIDIA ForceWare 180.43 (Vista64) NVIDIA ForceWare 178.24 (Vista32)
Desktop Resolution:	1920 x 1200
OS:	Windows Vista Ultimate 32-bit (for SYSMark) Windows Vista Ultimate 64-bit

SYSMark 2007 Performance

Our journey starts with SYSMark 2007, the only all-encompassing performance suite in our review today. The idea here is simple: one benchmark to indicate the overall performance of your machine.

The Phenom II X4 940 and the Core 2 Quad Q9400 offered identical performance under SYSMark 2007. As we saw earlier, the Q9400 ends up being around 6% faster than the Q8400 - which we see represented here in the SYSMark results. Overall, the Phenom II 940 kicks things off with a 6.4% overall advantage over the Q8400. It's near the border of being a noticeable performance difference, but it is a difference nonetheless.

To put performance in perspective the Q8400 with only 4MB of total L2 cache performs similar to a Q6600, our first quad-core pick, with its 8MB L2. A slightly higher clock speed and half the L2 cache are able to give us the same performance in this case. What's even cooler is this, look at the die size comparison between the Q8400 and the Q6600:

Processor	Manufacturing Process	Total L2 Cache	Transistor Count	Die Size
Intel Core 2 Quad Q6600	65nm	8MB	582M	286mm2
Intel Core 2 Quad Q8400	45nm	4MB	456M	164mm2

 

It's the beauty of Moore's Law: with fewer transistors crammed into a much smaller area, we're able to see the same performance. Let's also keep in mind that the Q6600 cost $851 at its release while the Q8400 debuted at $183. There is a benefit to jumping on the quad-core bandwagon late.

Adobe Photoshop CS4 Performance

To measure performance under Photoshop CS4 we turn to the Retouch Artists’ Speed Test. The test does basic photo editing; there are a couple of color space conversions, many layer creations, color curve adjustment, image and canvas size adjustment, unsharp mask, and finally a gaussian blur performed on the entire image.

The whole process is timed and thanks to the use of Intel's X25-M SSD as our test bed hard drive, performance is far more predictable than back when we used to test on mechanical disks.

Time is reported in seconds and the lower numbers mean better performance. The test is multithreaded and can hit all four cores in a quad-core machine.

Intel continues to dominate our Photoshop CS4 test. The Q8400 and Q9400 are equals in this case and both are nearly 15% faster than the Core 2 Quad Q6600. There's a definite benefit to going with a quad-core CPU in our CS4 results, it's not the same sort of boost you get in video encoding but it's appreciable.

DivX 8.5.3 with Xmpeg 5.0.3

Our DivX test is the same DivX / XMpeg 5.03 test we've run for the past few years now, the 1080p source file is encoded using the unconstrained DivX profile, quality/performance is set balanced at 5 and enhanced multithreading is enabled:

Video encoding performance is a tie between our two competitors. Both perform among the fastest quad-core CPUs on the market today, but clearly under the devine i7. Looking at the Pentium E5300 performance we see that even early Core 2 Duo systems stand to see a 30% boost in performance by moving to one of these affordable quad-core alternatives.

If you want to extrapolate however, here's a fun game. It took Intel roughly three years to offer the performance of a $500 CPU in something that costs less than $80. By 2011 you'll be able to get this sort of performance at under $80 as well. Mmm, ubiquitous multi-core.

x264 HD Video Encoding Performance

Graysky's x264 HD test uses the publicly available x264 codec (open source alternative to H.264) to encode a 4Mbps 720p MPEG-2 source. The focus here is on quality rather than speed, thus the benchmark uses a 2-pass encode and reports the average frame rate in each pass.

AMD continues to have a large advantage in two pass x264 encoding thanks to a lot of unaligned loads that penalize Intel's pre-Nehalem architectures. The second, more strenuous pass shows the two contenders in a tie just as we saw in our DivX results:

Windows Media Encoder 9 x64 Advanced Profile

In order to be codec agnostic we've got a Windows Media Encoder benchmark looking at the same sort of thing we've been doing in the DivX and x264 tests, but using WME instead.

While DivX and x264 both showed our contenders as equals, our WME test has the Phenom II 940 in the lead by nearly 6%.

3dsmax 9 - SPECapc 3dsmax CPU Rendering Test

Today's desktop processors are more than fast enough to do professional level 3D rendering at home. To look at performance under 3dsmax we ran the SPECapc 3dsmax 8 benchmark (only the CPU rendering tests) under 3dsmax 9 SP1. The results reported are the rendering composite scores:

Once more we have a situation where the Q8400 and Phenom II 940 are similar in performance, with AMD holding onto a negligible 2% advantage.

Cinebench R10

Created by the Cinema 4D folks we have Cinebench, a popular 3D rendering benchmark that gives us both single and multi-threaded 3D rendering results.

There's no questioning the Cinebench results, this is one of those areas where the Phenom II is just plain faster than the Q8400. The lead here is around 7%.

POV-Ray 3.73 beta 23 Ray Tracing Performance

POV-Ray is a popular, open-source raytracing application that also doubles as a great tool to measure CPU floating point performance.

I ran the SMP benchmark in beta 23 of POV-Ray 3.73. The numbers reported are the final score in pixels per second.

The trend continues in POV-ray, the Phenom II 940 holds onto a 5% performance advantage over the Q8400 here.

Blender 2.48a

Blender is an open source 3D modeling application. Our benchmark here simply times how long it takes to render a character that comes with the application.

The more your application is optimized for Intel's architecture however, the faster the tables can turn. Performance in Blender is dominated by the Q8400.

Sony Vegas Pro 8: Blu-ray Disc Creation

Although technically a test simulating the creation of a Blu-ray disc, the majority of the time in our Sony Vegas Pro benchmark is spent encoding the 25Mbps MPEG-2 video stream and not actually creating the Blu-ray disc itself.

As we saw in previous reviews, even the Q9550 is unable to outperform AMD's Phenom II 940 in our Sony Vegas BD creation test. The Q8400 doesn't stand a chance.

Sorenson Squeeze: FLV Creation

Another video related benchmark, we're using Sorenson Squeeze to convert regular videos into Flash videos for use on websites.

Microsoft Excel 2007

Excel can be a very powerful mathematical tool. In this benchmark we're running a Monte Carlo simulation on a very large spreadsheet of stock pricing data.

PAR2 Multithreaded Archive Recovery Performance

Par2 is an application used for reconstructing downloaded archives. It can generate parity data from a given archive and later use it to recover the archive.

Chuchusoft took the source code of par2cmdline 0.4 and parallelized it using Intel’s Threading Building Blocks 2.1. The result is a version of par2cmdline that can spawn multiple threads to repair par2 archives. For this test we took a 708MB archive, corrupted nearly 60MB of it, and used the multithreaded par2cmdline to recover it. The scores reported are the repair and recover time in seconds.

Both our PAR2 and WinRAR tests have AMD in the lead - real applications and a real performance advantage.

WinRAR - Archive Creation

Our WinRAR test simply takes 300MB of files and compresses them into a single RAR archive using the application's default settings. We're not doing anything exotic here, just looking at the impact of CPU performance on creating an archive:

WinRAR is a clear win for the 940. The Q8400 is hurt by its lack of cache and lower clock speed in this case.

Fallout 3 Performance

Bethesda’s latest game uses an updated version of the Gamebryo engine (Oblivion). This benchmark takes place immediately outside Vault 101. The character walks away from the vault through the Springvale ruins. The benchmark is measured manually using FRAPS.

Gaming performance is also a strength of the Phenom II X4 940, it's anywhere from 4 - 10% faster than the Q8400 with the exception of Far Cry 2 where the Q8400 has a 6% advantage.

Left 4 Dead Performance

 

FarCry 2 Multithreaded Game Performance

FarCry 2 ships with the most impressive benchmark tool we’ve ever seen in a PC game. Part of this is due to the fact that Ubisoft actually tapped a number of hardware sites (AnandTech included) from around the world to aid in the planning for the benchmark.

For our purposes we ran the CPU benchmark included in the latest patch:

Crysis Warhead Performance

Power Consumption

AMD did a lot of work on reducing idle power consumption with the move to Phenom II and thus we see the 758M transistor X4 940 draw less power than even the Q8400 at idle.

Under load it's a different story entirely, the Phenom II X4 940 needs another 32.1W. Granted it's a faster chip, but with all of those transistors switching at 3GHz the power consumption is simply higher.

Overclocking at Stock Voltages

The Phenom II X4 940 was AMD's first Phenom II to hit the market, the downside being that it wasn't the best example of AMD's 45nm manufacturing process. While the chips could overclock, their headroom without increasing their default core voltage just wasn't that high. The 940 we used for this review could only make it to 3.2GHz with a 2.2GHz NB frequency without increasing core voltage beyond the default 1.35V (as reported by the BIOS, 1.336V as reported by CPU-Z):

Phenom II X4 940 max overclock at stock voltage

The Core 2 Quad Q8400 on the other hand is based on a core that's been shipping for quite a while, at stock voltage using the stock cooler I was able to hit 3.08GHz:

Now remember that Phenom II isn't faster than Intel's Core 2 Quad clock for clock. AMD gets an advantage only because it sells higher clocked Phenom IIs at the same price as lower clocked Core 2 Quads. With only a 4% clock speed advantage, the performance is distinctly in Intel's favor:

Processor	Adobe Photoshop CS4	x264 HD - 2nd pass	POV-Ray	Far Cry 2	Idle Power	Load Power
AMD Phenom II X4 940 @ 3.2GHz	22.3s	19.0 fps	2539	51.5 fps	120.2W	209W
Intel Core 2 Quad Q8400 @ 3.08GHz	19.7s	20.4 fps	2597	53.5 fps	131.9W	181.5W

 

Photoshop and Far Cry 2 were both Q8400's strengths at stock clock speeds, so it makes sense that the Q8400 would widen the gap after we overclocked both chips. The x264 and POV-Ray tests were equal and in AMD's favor, respectively, at stock clocks. When overclocked without touching anything but the clock multiplier in the case of the Phenom II or FSB frequency and memory ratios in the case of the Core 2 Quad.

There is merit to looking at this sort of overclocking performance since it's the simplest way to overclock, but if you're willing to give your chip a little more voltage you can get a lot more clock speed...

Overclocking with a 10% Increase in Core Voltage

While there's something nice about not having to do anything to overclock your CPU other than adjusting a single parameter, tweaking your core voltage has the potential to significantly increase your overclocking potential.

Let's start with the Phenom II X4 940. Using the stock AMD cooler and pushing a max 10% increase in core voltage we ended up with 3.6GHz and a 2.2GHz NB frequency at 1.475V:

I was able to get the chip up to 3.9GHz but not stable enough to complete all of my tests over a significant period of time. Still, a 20% overclock with a moderate increase in core voltage isn't bad at all.

The Core 2 Quad Q8400 saw a 15% increase in clock speed without touching the core voltage. Bumping the core voltage up to 1.45V bought us another 11% taking the chip up to 3.36GHz:

At 3.6GHz vs. 3.36GHz, AMD's clock speed advantage grows to ~7%. Not as big as the 12.7% clock speed advantage at stock frequencies but bigger than the 4% difference in our stock voltage OC attempt.

Processor	Adobe Photoshop CS4	x264 HD - 2nd pass	POV-Ray	Far Cry 2	Idle Power	Load Power
AMD Phenom II X4 940 @ 3.6GHz	20.3s	21.3 fps	2839	51.5 fps	120.8W	256W
Intel Core 2 Quad Q8400 @ 3.36GHz	18.0s	22.3 fps	2771	59.5 fps	139.9W	213W

 

At max overclock something interesting happens. Obviously the Q8400 holds on to the lead in Photoshop and Far Cry 2, but POV-Ray goes back to AMD. The x264 encode ends up being a mild victory by the Q8400. Far Cry 2 is bizzare as the Phenom II saw no performance improvement from 3.2GHz up to 3.6GHz, regardless of how many times I ran the test I could never get a higher frame rate out of the system - even with CnQ disabled.There must be some other bottleneck operating here, but it's not a GPU one since the Q8400 was able to scale all the way up to 59.5 fps with the same graphics card.

The situation seems to boil down to this - at max frequency the Phenom II will probably retain its performance lead in 3D rendering and ray-tracing applications, while the video encoding lead goes to Intel. Power consumption also favors the Q8400 significantly.

Final Words

It took dual-core chips falling below $200 to start increasing their prevalence in the market, and today only one of our standard CPU tests won't see a performance increase from a dual core chip. I believe we're at the beginning of that same transition for quad-core CPUs. Many of our tests show a benefit from having four cores over two but in the next two years that should change significantly. The advent of GPU computing and the impending release of Larrabee will both bring about more focus on multi-threaded development. In the coming years a new group of applications that can run on both GPUs and multi-core CPUs will cement the transistion from applications that struggle to stress more than two cores to applications that scale to a virutally infinite number of cores.

The sheer affordability of quad-core processors today is impressive; $180 - $190 will buy you a Core 2 Quad Q8400 (2.66GHz/4MB L2), a Core 2 Duo E8500 (3.16GHz/6MB L2) or a Phenom II X4 940 (3.0GHz). Whether you go dual or quad is really a personal choice depending on the types of apps you run. If you look at our SYSMark 2007 results you’ll see that the E8500 is a better choice overall. Personally I’d opt for the quad core but that’s because when I’m most performance constrained it’s in applications that scale well to four cores, but if you don't do any 3D rendering or video encoding (or heavy multitasking between two multithreaded apps) then a fast dual-core may make the most sense for you today. If you're buying for a system that you plan on keeping for 3 - 5 years however, I suspect that quad-core is the way to go.

Between the Q8400 and the Phenom II X4 940, at stock clock speeds, the 940 is the way to go unless you're very concerned about power consumption or happen to be running applications that are very well optimized for Intel's Core architecture. Update #2: Intel has just confirmed that the Core 2 Quad Q8400 does support Intel's VT-x from the start, so the update below is incorrect. The Q8300, E5400, E5300, E7500 and E7400 will also end up transitioning to versions with VT-x support as well but only the Q8400 supports it from launch. Update: As many readers have pointed out, the Q8400 does not support Intel's VT for hardware accelerated virtualization. Honestly it's silly that Intel is attempting to use VT as a profit driver at this point. Not supporting VT on any quad-core CPU just doesn't make sense. The Phenom does support AMD's hardware virtualization AMD-V, and thus gives it a tremendous leg up if you care about the feature.

If you plan on doing some light overclocking, the Q8400 has more inherent potential. Start bumping up core voltages and the Phenom II X4 940 regains strength as it's able to increase the clock speed advantage once more. Throw overclocking into the mix and the comparison isn't quite as clean cut, both AMD and Intel trade blows in their advantages. I'd say AMD would probably have more wins in our applications but at the expense of much greater power consumption.

It's good to see that there's competition here, but Intel's profit margin advantage on the Q8400 is ridiculous. AMD has to sell something Nehalem sized for under $200 to remain relevant today. I'm far less concerned about who pulls ahead while overclocked and far more concerned about AMD's health at the end of all of this. Maybe the right way of looking at this isn't by talking about a 6% performance advantage, but instead talking about whether or not you want there to be a real competitor to Intel in the future. Maybe the Phenom II X4 940 should get the win here just to ensure we have an AMD to talk about in a couple of years...