Today’s processors: the value angle

Almost a year has passed since our first CPU value article. Since then, we’ve taken a look at the value propositions of current graphics processors, but we haven’t taken another shot at CPUs—until today, that is. Fresh from our review of AMD’s triple-core Phenom X3 processors, we’ve gathered all of our benchmark results and tossed current official pricing into the equation to see how much bang Intel’s and AMD’s latest processors deliver for your buck.

In this look at the CPU market, we explore how value scales across two, three, and four cores, whether bargain prices make AMD’s Phenoms attractive from a value perspective, and how Intel’s blazing-fast dual-core CPUs like the Core 2 Duo E8400 fit into the picture. Our results span 20 processors, 18 different benchmarks, and some custom tests that take power efficiency and more complete system pricing into account. Keep reading to see where the best values lie in the CPU market.

Quantifying CPU value

We noted in our GPU value article that it’s much easier to calculate value with microprocessors than with graphics processors. That’s true, but CPU value isn’t a completely straightforward affair either. We have two main variables to work with—performance and price—so working out a “performance per dollar” figure is easy enough. However, that equation doesn’t take into account a wealth of other performance factors, including memory bandwidth, storage bottlenecks, and graphics processing power in games. Power consumption also plays an important role, especially when it comes to the cost of a cooling solution. We’ll focus on CPU performance-per-dollar almost exclusively in this article, but while reading, keep in mind that there are other factors to take into account.

With that out of the way, let’s look at our test subjects. We have 20 processor configurations, 11 of which bear the Intel logo:

Model Clock
speed
Cores L2
cache (total)
Fab
process
TDP Price
Core
2 Duo E7200
2.53GHz 2 3MB 45nm 65W $133
Core
2 Duo E6750
2.66GHz 2 4MB 65nm 65W $183
Core
2 Duo E8400
3GHz 2 6MB 45nm 65W $183
Core
2 Quad Q6600
2.4GHz 4 8MB 65nm 95W $224
Core
2 Duo E8500
3.16GHz 2 6MB 45nm 65W $266
Core
2 Quad Q9300
2.5GHz 4 6MB 45nm 95W $266
Core
2 Quad Q9450
2.66GHz 4 12MB 45nm 95W $316
Core
2 Extreme QX6850
3GHz 4 8MB 65nm 130W $999
Core
2 Extreme QX9650
3GHz 4 12MB 45nm 130W $999
Core
2 Extreme QX9770
3.2GHz 4 12MB 45nm 136W $1399
Dual
Core 2 Extreme QX9775
3.2GHz 8 24MB 45nm 300W $2998

The other nine chips come from AMD:

Model Clock
speed
Cores L2 Cache/L3 cache Fab
process
TDP Price
Athlon
X2 5600+
2.8GHz 2 1MB 90nm 80W $122
Phenom X3 8450 2.1GHz 3 1.5MB/2MB 65nm 95W $145
Athlon
X2 6000+
3GHz 2 1MB 90nm 125W $153
Athlon
X2 6400+
3.2GHz 2 1MB 90nm 125W $163
Phenom
X3 8750
2.4GHz 3 1.5MB/2MB 65nm 95W $195
Phenom
X4 9500
2.2GHz 4 2MB/2MB 65nm 95W $195
Phenom
X4 9600
2.3GHz 4 2MB/2MB 65nm 95W $215
Phenom
X4 9750
2.4GHz 4 2MB/2MB 65nm 125W $215
Phenom
X4 9850
2.5GHz 4 2MB/2MB 65nm 125W $235

As in our last CPU value piece, we’ve taken bulk prices straight from Intel and AMD. Volume prices aren’t always representative of what consumers pay, but the two are close enough right now, and selecting “official” figures saves us the trouble of working out averages from fickle online retailers. Again, there are other factors to take into account here, such as platform and power costs. We’ll look at how platform pricing fits into all this soon, but let’s focus on just the processors for now.

Oh, also, we’ve included AMD’s Phenom 9500 and 9600 processors in our tests. These two chips are plagued by the infamous TLB erratum, the fix for which (enabled in most BIOSes by default, we might add), hurts performance significantly. We’ve disabled the fix for our tests, so our results should roughly represent the performance of the new, erratum-free 9550 and 9650 CPUs. We nevertheless resisted the temptation to simply rename our test chips, because the 9550 and 9650 are based on a new stepping of silicon and may perform slightly differently.

To paint a reasonably thorough picture of how all these processors handle common tasks, we’ve selected 18 benchmarks spanning everything from games, general productivity applications, and media encoding to folding, 3D rendering, and game development. The vast majority of these tests cover real-world applications, which we think is fitting considering the practical aim of this article.

Organizing all this data in a readable fashion isn’t an easy enterprise, so we’ve relied on a mix of charts and scatter plots once again. For each benchmark, we’ll start with a look at raw performance, then move on to a bar chart that ranks chips based on how much performance they deliver per dollar. When we’re dealing with tests that generate higher scores with faster CPUs, we’ll look at “score point per dollar.” However, when the performance metric shifts to time in seconds, where lower scores represent better performance, we’ll rank chips based on test run rate per dollar. The test run rate will be the reciprocal of the test score, expressed in megahertz.

The third tool we use quantify value is our trusty scatter plot:

Those who’ve checked out our previous two value articles will know the drill. The best possible chip would be at the top left, costing nothing while delivering maximum performance, while the worst would be at the bottom right, offering the lowest performance at the highest cost.

There’s a twist here, though. Since we have lots of chips, and our two-way Core 2 Extreme QX9775 would prolong the X axis substantially due to its high price, we’ve made two compromises in order to keep our scatter plots readable. For one, you won’t find the QX9775 setup listed at all in our scatter plots. This isn’t much of a sacrifice, since the priciest configuration almost always offers the worst value. The second compromise involves cropping the bottom of our Y axis. In the plot above, for example, the Y scale starts at 2000 instead of zero. We prefer to maintain a proper scale for our graphs here at TR, but in this case, we’ve elected to err on the side of readability. The scatter plots are a little tricky to decipher as it is, and without cropping, all our results would often be squeezed into roughly two square inches of graph area.

Let’s now have a brief look at our test systems and, after that, our results.

Test notes

We underclocked a Core 2 Extreme QX9650 in order to simulate our Core 2 Quad Q9450, but performance should be exactly the same. Power consumption should be a different matter, so we left the Q9450 out of those tests. We used the same method to simulate a Core 2 Duo E8400 with an underclocked E8500 and a Phenom X3 8450 with an underclocked 8750.

Our testing methods

As ever, we did our best to deliver clean benchmark numbers. Tests were run at least three times, and the results were averaged.

Our test systems were configured like so:

Processor Core 2 Quad Q6600 2.4GHz Core 2 Duo E6750 2.66GHz
Core 2 Extreme QX6850 3.00GHz
Core
2 Extreme QX9770
3.2GHz
Dual

Core
2 Extreme QX9775
3.2GHz
Athlon 64 X2 5600+ 2.8GHz
Athlon 64 X2 6000+ 3.0GHz
Athlon 64 X2 6400+ 3.2GHz
Phenom
9500
2.2GHz
Phenom
9600
2.3GHz
Core 2 Extreme QX9650 3.00GHz Phenom
X3 8450
2.1GHz

Phenom X3 8750 2.4GHz

Phenom
X4 9750
2.4GHz

Phenom X4 9850

Black Edition 2.5GHz

Core
2 Duo E7200 2.53GHz
Core
2 Duo E8400
3.0GHz

Core 2 Duo E8500
3.16GHz
Core
2 Quad Q9300
2.5GHz
Core
2 Quad Q9450
2.66GHz
System bus 1066MHz (266MHz quad-pumped) 1333MHz (333MHz quad-pumped) 1600MHz
(400MHz quad-pumped)
1600MHz
(400MHz quad-pumped)
1GHz HyperTransport 1GHz HyperTransport
Motherboard Gigabyte GA-P35T-DQ6 Gigabyte GA-P35T-DQ6 Gigabyte
GA-X38-DQ6
Intel
D5400XS
Asus M2N32-SLI Deluxe MSI
K9A2 Platinum
BIOS revision F1 F1 F6b XS54010J.86A.0780.

2008.0110.1956

1201 VP.0B7
(No patch)

V1.2B1 (TLB patch)

F4 V1.3
F5c
F5c
North bridge P35 Express MCH P35 Express MCH X38
Express MCH
5400
MCH
nForce 590 SLI SPP 790FX
South bridge ICH9R ICH9R ICH9R 6321ESB ICH nForce 590 SLI MCP SB600
Chipset drivers INF Update 8.3.0.1013

Intel Matrix Storage Manager 7.5

INF Update 8.3.0.1013

Intel Matrix Storage Manager 7.5

INF Update 8.3.0.1013

Intel Matrix Storage Manager 7.5

INF Update
8.5.0.1009

Intel Matrix Storage Manager 7.8

ForceWare 15.01
Memory size 4GB (4 DIMMs) 4GB (4 DIMMs) 4GB (4 DIMMs) 4GB
(2 DIMMs)
4GB (4 DIMMs) 4GB (4 DIMMs)
Memory type Corsair TWIN3X2048-1333C9DHX

DDR3 SDRAM at 1066MHz

Corsair TWIN3X2048-1333C9DHX

DDR3 SDRAM at 1333MHz

Corsair TWIN2X2048-8500C5D

DDR2 SDRAM at 800MHz

Micron
ECC DDR2-800 FB-DIMM at 800MHz
Corsair TWIN2X2048-8500

DDR2 SDRAM at ~800MHz

Corsair TWIN2X2048-8500C5D

DDR2 SDRAM at 800MHz

CAS latency (CL) 8 8 4 5 4 4
RAS to CAS delay (tRCD) 8 9 4 5 4 4
RAS precharge (tRP) 8 9 4 5 4 4
Cycle time (tRAS) 20 24 18 18 18 18
Audio Integrated ICH9R/ALC889A

with Realtek 6.0.1.5449 drivers

Integrated ICH9R/ALC889A

with Realtek 6.0.1.5449 drivers

Integrated
ICH9R/ALC889A

with Realtek 6.0.1.5449 drivers

Integrated
6321ESB/STAC9274D5

with SigmaTel 6.10.5511.0 drivers

Integrated nForce 590 MCP/AD1988B

with Soundmax 6.10.2.6100 drivers

Integrated
SB600/ALC888

with Realtek 6.0.1.5532 drivers

Hard drive WD Caviar SE16 320GB SATA
Graphics GeForce 8800 GTX 768MB PCIe with ForceWare 163.11 and 163.71 drivers
OS Windows Vista Ultimate x64 Edition
OS updates KB940105, KB929777 (nForce/790FX systems only), KB938194, KB938979

Please note that testing was conducted in two stages. Non-gaming apps and Supreme Commander were tested with Vista patches KB940105 and KB929777 (nForce/790FX systems only) and ForceWare 163.11 drivers. The other games were tested with the additional Vista patches KB938194 and KB938979 and ForceWare 163.71 drivers.

Thanks to Corsair for providing us with memory for our testing. Their products and support are far and away superior to generic, no-name memory.

Our single-socket test systems were powered by OCZ GameXStream 700W power supply units. The dual-socket systems were powered by PC Power & Cooling Turbo-Cool 1KW-SR power supplies. Thanks to OCZ for providing these units for our use in testing.

Also, the folks at NCIXUS.com hooked us up with a nice deal on the WD Caviar SE16 drives used in our test rigs. NCIX now sells to U.S. customers, so check them out.

The test systems’ Windows desktops were set at 1280×1024 in 32-bit color at an 85Hz screen refresh rate. Vertical refresh sync (vsync) was disabled.

We used the following versions of our test applications:

The tests and methods we employ are usually publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.

WorldBench

WorldBench’s overall score is a pretty decent indication of general-use performance for desktop computers. This benchmark uses scripting to step through a series of tasks in common Windows applications and then produces an overall score for comparison. WorldBench also records individual results for its component application tests, allowing us to compare performance in each. We’ll look at the overall score, and then we’ll show individual application results alongside the results from some of our own application tests.

Already, we see some of the same trends as in our past articles. The cheapest offerings—AMD’s Athlon X2 5600+ and Intel’s Core 2 Duo E7200—score highest on our performance-per-dollar chart, while the outrageously expensive Core 2 Extreme QX9775 two-way setup delivers the least bang for your buck.

The scatter plot helps bring the performance question into perspective. The Core 2 Duo E7200 and Core 2 Duo E8400 fare better than alternatives at similar price points, and the Athlon 64 X2 5600+ doesn’t look so hot. Quad-core chips don’t look especially attractive here, either, which isn’t surprising considering that few WorldBench tests take advantage of more than two cores.

We should probably take a minute to say something about AMD. As you can see in the scatter plot, all of the AMD chips are clumped together below the Intel offerings. The differences in performance between the two aren’t huge—remember that we cropped our Y axes to make the results readable—but AMD’s offerings don’t perform as well in WorldBench’s application suite overall.

Productivity and general use software

MS Office productivity

The AMD CPUs do well in Microsoft Office. Our value chart and scatter plot both show the Athlon X2 5600+ beating out the Core 2 Duo E7200, and the Athlon 64 X2 6000+ and 6400+ rival the Core 2 Quad Q9300 and Core 2 Duo E8500. However, a look at the Y axis and our performance chart suggests that the Office test isn’t particularly CPU-bound, making for relatively small differences in actual performance.

Firefox web browsing

The picture is a little more mixed in Firefox, where the E7200 gets back at the 5600+ and the E8400 steps all over the 6400+. Because Firefox doesn’t seem to have much use for more than a couple of CPU threads, Phenoms and Core 2 Quads don’t really shine.

Image processing and multitasking

Photoshop

Like Firefox, this Adobe Photoshop CS2 test doesn’t benefit significantly from more than a couple of CPU cores. Our scatter plot shows the Core 2 Duo E8400 towering over pricier Core 2 Quads, not to mention AMD’s entire Phenom X4 lineup. Here again, the E7200 and E8400 both look very well positioned.

The Panorama Factory photo stitching
The Panorama Factory handles an increasingly popular image processing task: joining together multiple images to create a wide-aspect panorama. This task can require lots of memory and can be computationally intensive, so The Panorama Factory comes in a 64-bit version that’s multithreaded. We asked it to join four pictures, each eight megapixels, into a glorious panorama of the interior of Damage Labs. We’ve added up the total operation time to give us an overall measure of performance.

At last, an application that showcases the potential of quad-core processors. Intel’s Core 2 Quads and AMD’s Phenoms zoom past comparable dual-core chips in The Panorama Factory, and even AMD’s slowest quad-core Phenom manages to outperform the Core 2 Duo E8400.

Where’s the sweet spot here? When we take a gander at sub-$200 offerings, the Core 2 Duo E7200 and E8400 look to us like the best choices. If you want to spend more than $200, it’s a toss-up between the Core 2 Quad Q6600 and the Q9300—the former being a little slower but less expensive. We’d probably side with the Q9300, despite its lower value score, due to its superior performance and higher power efficiency. More on that in a moment.

Multitasking – Firefox and Windows Media Encoder

Multitasking normally showcases the strengths of multiple CPU cores, so we’re a little surprised to see the Core 2 Duo E8400 and E8500 do so well here. The E8500 may cost the same as the Q9300, but it outperforms the quad-core chip by about 13%. The E8400 looks even better, though, considering that it’s nearly as fast as the E8500, costs $80 less, and ranks third in performance per dollar.

Team Fortress 2

We’ll kick off our gaming tests with some Team Fortress 2, Valve’s class-driven multiplayer shooter based on the Source game engine. In order to produce easily repeatable results, we’ve tested TF2 by recording a demo during gameplay and playing it back using the game’s timedemo function. In this demo, we played as the Heavy Weapons Guy, with a medic in tow, dealing some serious pain to the blue team.

We used a relatively low display resolution with low levels of filtering and AA in order to prevent the graphics card from becoming a primary performance bottleneck, so we could show you the performance differences between the CPUs. We tested at 1024×768 resolution with the game’s detail levels set to their highest settings. HDR lighting and motion blur were enabled. Antialiasing was disabled, and texture filtering was set to trilinear filtering only.

Judging by both our CPU usage graph and our performance results above, Team Fortress 2 only takes advantage of one full CPU core. The Core 2 Duo E7200 and E8400 continue to do well here, occupying prime real estate in our scatter plot and delivering good performance per dollar in our value chart.

Somewhat embarrassingly for AMD, the $133 E7200 beats out the best Athlon and Phenom processors on the market. And really, the E7200 may be all you need to have a good gaming experience in TF2. The E8400 certainly looks like a tantalizing step up, though, and it’s in third place in the value chart.

BioShock

We tested BioShock by manually playing through a Big Daddy fight five times while recording frame rates using the FRAPS utility.

This method has the advantage of simulating real gameplay quite closely, but it comes at the expense of precise repeatability. We believe five sample sessions are sufficient to get reasonably consistent results. In addition to average frame rates, we’ve included the low frame rates, because those tend to reflect the user experience in performance-critical situations. In order to diminish the effect of outliers, we’ve reported the median of the five low frame rates we encountered.

For this test, we largely used BioShock‘s default image quality settings for DirectX 10 graphics cards, but again, we tested at a relatively low resolution of 1024×768 in order to prevent the GPU from becoming the main limiter of performance.

AMD salvages its honor in BioShock thanks to a strong performance from the Phenom X4 9750. However, we should point out that BioShock doesn’t seem all that bound by CPU performance—at least that’s how we’d explain why half the processors we tested produced close to the same frame rates, and why the 9750 somehow outdid the higher-clocked Phenom X4 9850.

We believe Intel’s Core 2 Duo E7200 lies in the sweet spot here, trailed closely by the Phenom X3 8450. Speedier dual-core chips aren’t much faster, and extra cores don’t seem to help a whole lot.

Lost Planet: Extreme Condition
Lost Planet puts the latest hardware to good use via DirectX 10 and multiple threads—as many as eight, in fact. Lost Planet‘s developers have built a benchmarking tool into the game, and it tests two different levels: a snow-covered outdoor area with small numbers of large villains to fight, and another level set inside of a cave with large numbers of small, flying creatures filling the air. The former doesn’t appear to be CPU-bound, so we’ll be looking at the latter.

We tested this game at 1152×864 resolution, largely with its default quality settings. The exceptions: texture filtering was set to trilinear, edge antialiasing was disabled, and “Concurrent operations” was set to match the number of CPU cores available.

Lost Planet‘s ability to harness multiple CPU cores effectively may be rather unique among today’s games, but it shows what may be in store for future titles. Here, it seems you can’t go wrong with any quad-core chip priced around $200, be it a Phenom X4 or Core 2 Quad. The Phenom X4 9500 looks particularly attractive, outdoing the Core 2 Duo E8400 at almost the same price point while nipping at the heels of a Core 2 Quad Q6600 that costs $30 more.

Audio and video encoding

LAME MT audio encoding

LAME MT is a multithreaded version of the LAME MP3 encoder. LAME MT was created as a demonstration of the benefits of multithreading specifically on a Hyper-Threaded CPU like the Pentium 4. Of course, multithreading works even better on multi-core processors. You can download a paper (in Word format) describing the programming effort.

Rather than run multiple parallel threads, LAME MT runs the MP3 encoder’s psycho-acoustic analysis function on a separate thread from the rest of the encoder using simple linear pipelining. That is, the psycho-acoustic analysis happens one frame ahead of everything else, and its results are buffered for later use by the second thread. That means this test won’t really use more than two CPU cores.

Our results focus on the 64-bit version of LAME MT built with Intel’s compiler. We are encoding a massive 10-minute, 6-second 101MB WAV file at a variable bit rate.

LAME MT’s dual-threaded execution allows Intel’s speedy 45nm Core 2 Duos to take the cake yet again. The E7200 and E8400 both offer compelling value.

Windows Media Encoder x64 Edition video encoding

Windows Media Encoder is one of the few popular video encoding tools that uses four threads to take advantage of quad-core systems, and it comes in a 64-bit version. Unfortunately, it doesn’t appear to use more than four threads, even on an eight-core system. For this test, we asked Windows Media Encoder to transcode a 153MB 1080-line widescreen video into a 720-line WMV using its built-in DVD/Hardware profile. Because the default “High definition quality audio” codec threw some errors in Windows Vista, I instead used the “Multichannel audio” codec. Both audio codecs have a variable bitrate peak of 192Kbps.

Windows Media Encoder juggles multiple CPU threads comfortably, allowing quad-core chips to stretch their legs—and the Phenom X4 9500 to trump the Core 2 Duo E8400. Further up the ladder, the Core 2 Quad Q6600 just barely edges out the Phenom X4 9850.

As prices climb, performance returns begin to diminish. Perhaps the best examples of this trend are the Core 2 Extreme QX9650 and 9770, both of which fail to outperform significantly the vastly cheaper Core 2 Quads. The two-way Core 2 Extreme QX9775 setup isn’t in our scatter plot, but our performance chart shows it’s barely any faster than the QX9770.

Incidentally, if you’re wondering why the Phenom X3s fare so poorly, it’s because Windows Media Encoder only makes use of two of their cores. AMD has told us it’s working with Microsoft on a fix.

WinZip file compression

The Core 2 Duo E7200 and Core 2 Quad Q9300 offer nearly identical performance in WinZip, suggesting that this is yet another application that doesn’t make effective use of quad-core processors. As we saw with other apps that don’t take advantage of more than two cores, the E7200 and E8400 look to be the best values south and north of the $150 mark, respectively. AMD’s Athlon and Phenom chips all trail their Intel rivals by substantial margins here.

3ds max rendering

Our 3ds max results somewhat mirror those of the Windows Media Encoder test: AMD’s Phenom X4 9500 outdoes the Core 2 Duo E8400, while the Core 2 Quad Q6600 trumps the fastest Phenoms. Moving up the price scale, Core 2 Extreme offerings don’t exactly justify their lofty price premiums with markedly better performance. Even the eight-core QX9775 setup fares poorly when we look at performance per dollar.

Cinebench rendering

Cinebench is a benchmark based on Maxon’s Cinema 4D rendering engine. It’s multithreaded and comes with a 64-bit executable. This test runs with just a single thread and then with as many threads as CPU cores are available.

Cinebench gives more love to the Phenoms—so much so, in fact, that the Phenom X4 9500 tops our value chart. Our scatter plot confirms its attractive positioning. The Core 2 Quad Q9300 is probably the next step we’d consider, since other sub-$250 quad-core offerings all lie quite close to one another.

POV-Ray rendering

We caved in and moved to the beta version of POV-Ray 3.7 that includes native multithreading. The latest beta 64-bit executable is still quite a bit slower than the 3.6 release, but it should give us a decent look at comparative performance, regardless.

POV-Ray makes good use of three and four cores, and the Phenoms perform well here as a result. In fact, the $235 Phenom X4 9850 outperforms the $266 Core 2 Quad Q9300 and sits at nearly the same performance level as the $999 Core 2 Extreme QX6850. The $215 Phenom X4 9750 ranks higher on our value chart than the X4 9850, but for the record, we really like the X4 9850’s unlocked multiplier.

Folding@Home

We have a slick little Folding@Home benchmark CD created by notfred, one of the members of Team TR, our excellent Folding team. For the unfamiliar, Folding@Home is a distributed computing project created by folks at Stanford University that investigates how proteins work in the human body, in an attempt to better understand diseases like Parkinson’s, Alzheimer’s, and cystic fibrosis. It’s a great way to use your PC’s spare CPU cycles to help advance medical research. I’d encourage you to visit our distributed computing forum and consider joining our team if you haven’t already joined one.

The Folding@Home project uses a number of highly optimized routines to process different types of work units from Stanford’s research projects. The Gromacs core, for instance, uses SSE on Intel processors, 3DNow! on AMD processors, and Altivec on PowerPCs. Overall, Folding@Home should be a great example of real-world scientific computing.

notfred’s Folding Benchmark CD tests the most common work unit types and estimates performance in terms of the points per day that a CPU could earn for a Folding team member. The CD itself is a bootable ISO. The CD boots into Linux, detects the system’s processors and Ethernet adapters, picks up an IP address, and downloads the latest versions of the Folding execution cores from Stanford. It then processes a sample work unit of each type.

On a system with two CPU cores, for instance, the CD spins off a Tinker WU on core 1 and an Amber WU on core 2. When either of those WUs are finished, the benchmark moves on to additional WU types, always keeping both cores occupied with some sort of calculation. Should the benchmark run out of new WUs to test, it simply processes another WU in order to prevent any of the cores from going idle as the others finish. Once all four of the WU types have been tested, the benchmark averages the points per day among them. That points-per-day average is then multiplied by the number of cores on the CPU in order to estimate the total number of points per day that CPU might achieve. We’ll be looking at that last figure for this value comparison. You can check detailed numbers for each unit type in our Phenom X3 review.

This may be a somewhat quirky method of estimating overall performance, but our sense is that it generally ought to work. We’ve discussed some potential reservations about how it works here, for those who are interested. We have included results for each of the individual WU types below, so you can see how the different CPUs perform on each.

These results suggest folding enthusiasts ought to at least consider AMD’s Phenom line. That said, the performance of the Core 2 Quad Q9300 towers over cheaper quad-core offerings, and it’s ranked third in our value chart.

Valve Source engine particle simulation

Next up are a couple of tests we picked up during a visit to Valve Software, the developers of the Half-Life games. They had been working to incorporate support for multi-core processors into their Source game engine, and they cooked up a couple of benchmarks to demonstrate the benefits of multithreading.

The first of those tests runs a particle simulation inside of the Source engine. Most games today use particle systems to create effects like smoke, steam, and fire, but the realism and interactivity of those effects are limited by the available computing horsepower. Valve’s particle system distributes the load across multiple CPU cores.

The results in Valve’s particle simulation benchmark are reminiscent of those we saw in Lost Planet‘s Cave benchmark a few pages back. Here, however, the Core 2 Quad Q6600 comes out ahead of the Phenom pack and sets up camp at the top of our value chart. The Q9300 isn’t far behind, offering slightly higher performance at a slightly higher price.

Valve VRAD map compilation

This next test processes a map from Half-Life 2 using Valve’s VRAD lighting tool. Valve uses VRAD to precompute lighting that goes into games like Half-Life 2. This isn’t a real-time process, and it doesn’t reflect the performance one would experience while playing a game. Instead, it shows how multiple CPU cores can speed up game development.

Here’s another test where the Core 2 Quad Q6600 excels. Intel’s cheapest quad-core CPU finds itself at the top of our value chart again, and our scatter plot shows it comfortably above similarly-priced Phenoms from AMD. The Core 2 Quad Q9300 comes in second place in the value chart, providing a good alternative that costs a little more but also runs a little faster, all while consuming less power overall.

Power efficiency

We’ve talked a little bit about power envelopes throughout the past few pages, so let’s have a closer look at how our 20 CPUs compare when you factor power efficiency into the value equation. We measured power utilization with an Extech 380803 power meter. The meter reads power use at the wall socket, so it incorporates power use from the entire system—the CPU, motherboard, memory, graphics solution, hard drives, and anything else plugged into the power supply unit. (We plugged the computer monitor into a separate outlet, though.) We measured how each of our test systems used power across a set time period, during which time we ran Cinebench’s multithreaded rendering test.

Almost all of the systems had their power management features (such as SpeedStep and Cool’n’Quiet) enabled during these tests via Windows Vista’s “Balanced” power options profile. The exception here was the dual QX9775 “Skulltrail” system, since its BIOS didn’t support SpeedStep.

For this piece, we’ve limited the scope of our results to our render energy test, which tells us how much energy each system used to render our test scene in Cinebench. For a more detailed look at our methodology and power efficiency tests for these chips, you can check out our full results in our review of AMD’s Phenom X3 processors.

For now, let’s look at power efficiency per dollar. The 1/microjoules value in our power efficiency per dollar graph is really 1/(watt-seconds/1000000), or 1,000,000 m-2 kg-1 s2. That’s a little obscure, but it quantifies power efficiency in a readable fashion based on the source data, which is in joules.

High-end, quad-core chips are the most power-efficient in this test (and they should be, since they spend less time at their fully loaded state rendering the scene). Still, the Core 2 Duo E7200 trumps other contenders in our power-efficiency-per-dollar chart. Looking at our scatter plot, we see that the E7200 outdoes all of our AMD chips and comes close to the level of efficiency of the Core 2 Quad Q6600 and Core 2 Duo E8500.

The next step up is undeniably the Core 2 Quad Q9300, which places second in our value rankings. This chip rivals the power efficiency of the QX9650 at a fraction of the price, and our scatter plot shows it sitting quite close to the magical top left corner of the plot. On the other side, the two-way Core 2 Extreme QX9775 provides similar power efficiency at more than ten times the cost.

Putting it all together

We’ve created a synthetic “overall performance per dollar” scatter plot to sum up our results. We’re presenting our performance numbers as percentages based on an average of each CPU’s score across all 18 of our performance tests. The 100% score corresponds to the slowest chip in our test lineup, the Athlon X2 5600+, and other scores are relative to it. As with all aggregate scores, we don’t recommend you draw hasty conclusions from the results. If there’s one thing we can say for sure after looking at all the data, it’s that the performance picture can vary tremendously from one application to another.

As we’ve seen all along, AMD’s processors bunch up tightly in the space under $250, making our task of sorting out relative value rather difficult. Unhappily for AMD, the Intel processors that occupy the same price range tend to sit further toward the left and top of the plot, indicating stronger performance per dollar. In fact, a range of Intel processors forms a nice line of values extending from the Core 2 Duo E7200 through the E8400, Q6600, Q9300, and Q9450. As we step up the ladder in price among these choices, performance increases by solid margins, as well.

In response to reader feedback, we’ve taken another look at overall performance per dollar that factors in total system costs. Below, we’ve modified the price figures on the X axis to factor in the cost of a complete system’s worth of components. For our theoretical AMD systems, we selected an MSI K9A2 Platinum motherboard ($149.99), 2GB of Kingston DDR2-800 memory ($43.99), a 320GB Western Digital Caviar SE16 hard drive ($69.99), and an XFX GeForce 8800 GT graphics card ($189.99). Our Intel estimates are the same, except that we selected an identically priced Gigabyte GA-EP35-DS3P motherboard instead of the MSI board. Similarly, for our dual QX9775 system, we factored in the price of Intel’s Skulltrail motherboard ($609.99) and a couple gigs worth of DDR2-667 FB-DIMMs ($149.99).

Taking the cost of other components into account helps us put the impact of CPU prices into context. After all, if you choose to spend $300 instead of $150 on a CPU, you won’t pay twice as much for your PC. Of course, your own mileage may vary depending on which components you buy for a new system.

With total system costs factored into the picture, the race becomes even tighter. The Intel “line of values” remains intact, but AMD’s offerings are right in the thick of things, although AMD’s lineup doesn’t extend as far up the price and performance axes as Intel’s.

Despite the prevalence of apps that don’t take advantage of more than a couple of cores in our benchmark suite, the Core 2 Quad Q6600 comes out looking very good, ahead of the dual-core CPUs we tested, not to mention each and every AMD processor. Once again, the Core 2 Quad Q9300 sits in a particularly nice spot, too. It lies comfortably above all cheaper and similarly priced offerings on the performance scale, but costs only $144 more than the cheapest config with an Athlon X2 5600+. Other quad-core CPUs don’t look quite as appealing, and only the very expensive Core 2 Extreme QX9775 setup offers substantially higher overall performance.

Among the AMD processors, we should single out the Phenom X4 9850 for distinction. This CPU places well in our overall scatter plots, very close to two of the best Intel values in the Core 2 Duo 8400 and Core 2 Quad Q6600, and as we’ve mentioned, it comes with an unlocked upper multiplier for easy overclocking. As a result, AMD’s highest performance processor may also be its most compelling overall value.

Of course, depending on your needs, a more (ahem) extreme solution may be a good value for you, as well. Relatively speaking, computer hardware is cheap. And time is money. If going with a more expensive CPU can save you several minutes or more each day, it may be worth every cent.

Comments closed
    • indeego
    • 11 years ago

    You should update this 2-3x a year. prices are down quite a bit. I picked up an E8400 for $130 this past thanksgiving, for example, $50 less than this articleg{<.<}g

    • Clairvaux
    • 11 years ago

    Fantastic job. Exactly what I had been looking for all along.

    Other sites doing value analysis would stop at ranking price/performance ratios, which are useless unless you map them against the absolute price, because in the real world people have budgets. That’s the reason you cannot do a proper value review without 2-D scatter graphs such as yours.

    Now for a few suggestions.

    1. Take the more ridiculously-priced Intel Quads out of the picture. Show one graph including them, then kick them out. Or find a graphical metaphor equivalent to a footnote, or zoom-in.

    Once you’ve seen one of those graphs with those 1 000 $ + processors on the right-hand side, it becomes very clear they are not intended for value-seeking buyers. And their presence muddles up the details in that 100 – 250 $ segment, which is precisely where one would like to dig deeper.

    2. Include (or make a special review for) lower-priced processors. Pentium dual-cores and cheaper AMD dual-cores are obviously prime candidates for value-seeking buyers.

    3. Keep (or increase…) the level of differentiation you have reached between applications (office vs. games), core usage, processor price vs. system price, etc. Any less detail, and value analysis becomes next to meaningless.

    I’d especially like to see you develop the system value angle, which is one of the strong points of this review. Expanding the scope to a lower price segment would probably introduce price differences between equivalent Intel and AMD motherboards, which we do not have here. So the global AMD / Intel picture could change quite a bit. Also, including the integrated graphics option would be interesting.

    4. Do this more often. An update every six months would be very useful.

    5. Even better: enable permenent update through automatic price acquisition (I think one of your competitors does that). Of course, analysis cannot be updated in real time.

    6. And finally, in the dreaming awake department, is there a way you could enable your readers to tinker with the results in order to analyse them in a specific way appropriate to their own usage, without giving away your work to competitors? (Basically, something equivalent to having the raw data in Excel format, and being able to generate your own graphs.)

      • TurtlePerson2
      • 11 years ago

      There should be a graph with just midrange processors. The others end up just being outliers and ruin the rest of the data. It’s difficult to see the value between $0-$200 because there’s a couple $500+ processors.

    • zappman
    • 11 years ago

    The test configuration used to test the Phenom X4 9850 (and all other AM2 CPUs) is bad, resulting in misleading low numbers for the top AMD CPUs.

    Some Background Info:
    The Phenom X4 9850 Black Edition 2.5GHz used in the test is an AM2+ processor.

    The MSI K9A2 Platinum motherboard used in the test, supports 1066 MHz
    Dual Channel DDR2 for AM2+ processors.

    The Corsair TWIN2X2048-8500C5D DDR2 SDRAM is 1066 MHz Memory.

    The problem with the test setup used:
    The Corsair TWIN2X2048-8500C5D DDR2 SDRAM was configured to 800MHZ not 1066MHz.

    The results of the tests are not valid and are misleading, causing the AM2+ CPUs to have lower scores then they should.

    Until the tests can be rerun, a disclaimer should be put at the top of the review.

    • oakdad
    • 11 years ago

    You only have 17 processors on the power Power efficiency page.
    Missing are the E8400, Q9450, 8450

    • bdwilcox
    • 11 years ago

    I think another benchmark with real world implications would be a Windows XP and Vista boot time chart. And I don’t just mean the OS, but the OS bloated with a bunch of startup programs like AVG anti-virus, Webroot Spysweeper, HP printer drivers, Itunes, and whatever else you can find clogging up most boxes. Microsoft has a utility for benchmarking XP’s boot times (BootVis) which they no longer offer at their website, but can be found at other download sites like Major Geeks or Softpedia.

      • crazybus
      • 11 years ago

      How often do you need to boot? Every patch Tuesday, hardware change, driver update or rare app install is it for me. Is boot time really that important of a metric?

        • bdwilcox
        • 11 years ago

        I boot every day and sometimes several times a day depending on what I’m doing. Boot time differences between old and new PCs are definitely noticeable and seem to have a direct correlation to CPU speed as long as you have at least 512MB of memory (on XP).

          • indeego
          • 11 years ago

          So don’t boot, hibernate or sleep/suspend. There’s little reason to do complete power-offs after Vista SP1 and XP SP3.

          But yeah I used a ~6 year old W2K machine (PIII 933mhz) and it beat the pants off any boot for a modern XP or Vista box I’ve seen recentlyg{<.<}g

    • thermistor
    • 11 years ago

    #80…exactly. IF you’re will to go for the reward vs. what appears to be low risk. The internets show lots of OC proponents hugging their Intel E2xxx series with a pretty stellar, stable OC.

    I would add the E2220 @ 2.4 Ghz to the article, too. It is only $90, but can demonstrate what a ‘value’ processor can do.

    I also agree that everyone buys to their own needs/wants.

    • thermistor
    • 11 years ago

    I just purchased on e-bay an Conroe-L 420 (1.6 smokin’ Ghz), to use in a file server box. I can tell ya that I wouldn’t want anything weak like an E2140, E4300, any of the X2 lineup below 5600+, in any of my daily driver PC’s, unless you overclocked them to satisfactory performance.

    Keep the mid-range, though including the E4700 @ 2.6 Ghz stock would be very revealing. It’s currently $139 at the egg.

    To the fellow who said to buy the cheapest…I can’t do that with cars, scotch, CPU’s or women – you always pay more in the end and sometimes the payment is extracted in frustration, anguish, or hangovers.

      • zealeus
      • 11 years ago

      Except that with the e2140, I’ve overclocked it on air to 2.8Ghz without any stability issues. Now consider that I can buy the slightly faster e2160 for less than half the price of that e4700 (though at this point, one should be looking at the e7200 in that price range anyways), while having clock speeds faster than the e4700’s stock speed.

      Fort budget builder like myself who are willing to overclock, there’s little incentive to spend 2-3x the money on a more expensive processor, particularly when it’s a home computer where I don’t need my CDs to rip in 30 seconds instead of 37.
      I’ve been very happy with the e2140 proc and see no reason to upgrade in the near future.

      • MadManOriginal
      • 11 years ago

      I’ve got a passively Scythe Ninja-cooled Cel 420 as well. It’s fine for a pure server box, if I ever decide to do more on the backend though I’d get an e7200, they are cool running as well. Too bad I bought it not terribly long before the dual-core Cellies, the E1200 is only $50.

    • Nictron
    • 11 years ago

    There is only one answer to your question or statement;

    *[

      • eitje
      • 11 years ago

      freezepop? rock band?

      hehe. ๐Ÿ˜›

    • eitje
    • 11 years ago

    page 4 – panorama factory scatter – E8500 is the wrong color. ๐Ÿ™‚

    also, you really should have left skulltrail out of the last few scatters. :-/

    • StashTheVampede
    • 11 years ago

    After my brother read this, he’s basically stuck on what he wants to buy: Q6600 or E8400. Not many games scale linearly across four CPUs and Ghz is still king. He can plop in some uber quad later (before nehalem, I know), so he has a little life in the rig.

      • MadManOriginal
      • 11 years ago

      GHz is king at low resolutions which no one actually uses for gameplay. I haven’t seen a solid review of CPU speeds scaling at real-world resolutions though but I’d really like to see TR do one. Yes, I am pimping my previous idea again.

      • Usacomp2k3
      • 11 years ago

      He wants an e7200 or so. Spend the money on a better video card.

    • henfactor
    • 11 years ago

    Ha! My math teach said I would never see a scatter plot in the real world… I’m sending her the link right now, AND bringing this up on the overhead at school! That’ll show her!

    Great article, very informative, thanks for all the hard work you guys put in these things ๐Ÿ™‚

    Keep’em coming!

      • cynan
      • 11 years ago

      /[

    • kleinwl
    • 11 years ago

    One issue about the motherboards is that typically the AMD motherboards can be less expensive. If I’m about to order a $65 processor, I’m not going to shell out for a $150 motherboard… a $60 or $70 motherboard is going to be all I need. This scaling of costs will kew your graphs even more.

    • Willard
    • 11 years ago

    Two comments.
    First, the overall performance/system price graph on the last page does not add any information to the previous graph (performance vs cpu price). You’re basically just adding a constant to all cpu prices and rescaling the x. One could wish it were otherwise, but constants are not interesting to anyone but *the* most die-hard geeks.

    Maybe what you’re trying to get at is the *share* of the total system price that the cpu takes up, and how much performance varies by changes in the *share* of the system price? Since the system price is a constant, though, I think this would again boil down to differences in cpu price.

    Second, the 100% baseline in the graphs means the bottom third of the graph is unnecessary :).

      • MrJP
      • 11 years ago

      Actually the addition of a constant makes a significant difference in this case, especially where the graph is fully displayed with zero visible on both axes. It’s the relative gradient from the origin to the different points that determines the value of the different processors, which is quite different using pure CPU prices compared with system prices.

      For example, CPU A gives 25% more performance than CPU B but also costs 25% more. So, Perf(A)=1.25xPerf(B) and Price(A)=1.25xPrice(B). On a plot with just CPU price, the gradient from these points to the origin is equal, so can be considered to have equal value. If the rest of the system costs 4 times more than CPU B (for example), then SysPrice(B)=5xPrice(B) and SysPrice(A)=5.25xPrice(B), so SysPrice(A)=(5.25/5)xSysPrice(B)=1.05xSysPrice(B). The performance increase going from CPU B to CPU A is still 25%, but the system price increase is only 5%. It’s then pretty clear that in the context of the complete system price, CPU A is much better value. This is a crude example, but shows why more expensive processors are more easily justifiable in the context of a more expensive system.

        • Willard
        • 11 years ago

        I agree that this is what they’re trying to get at, and great job formalizing the relationship in a way that was beyond me this a.m. However I’m still perplexed by how to scale this relationship. In your example, aren’t you really just changing the cost/performance graph slopes by an arbitrary amount, rather than presenting new data? You pick a system-cost factor of 5, but I think you’ll find the linear relationship between system price and performance price holds, no matter your factor.

        For example, your factor of 5 means A costs 5% more than B, while a factor of, say, 10, would means A costs 2.5% more than B (I think). And this new relationship would hold throughout the range of cpu prices at a factor of 10. So which is better, a cost change of 2.5% or 5%?

        Answer – It doesn’t matter, because the performance/cost changes across cpus are driven by the arbitrary cost of the system and nothing else.

        So really you’re right back to A costs more than B and also performs better.

        In any event, the graphs on page 10 don’t accomplish much – I was struck by how similar they were.

          • MrJP
          • 11 years ago

          I suppose it just depends on how you go about weighing up value if you’re shopping for a whole new system. Extra money (usually) means extra performance, and so the break point comes where you consider the performance gain to be not worth the extra cost. That’s where the value of the rest of the system becomes important, and that’s why the best-value system at any given price point should be well balanced.

          Just looking at CPU performance vs CPU cost is interesting, but it’s not a universally-applicable measure of value because otherwise everyone would buy the same peak-value CPU. The point on the curve where each person finds their “ideal” CPU will be different depending on their personal performance/cost tolerance, which largely depends on their total system budget.

          I suppose what I’ve ended up arguing is that while it’s important to consider the total system cost in a discussion like this, using a single system cost is still not going to work for anyone who’s budget is dramatically different. Which is back to your original point about an arbitrary constant offset…

    • danny e.
    • 11 years ago

    hmm there’s something that really bothers me about charts with arbitrary starting points and how that shapes the imaginary diagonal line and gives weird impressions.

    also… i’ve proved with my own testing that for games the CPU barely matters anymore. it would be much more interesting to me if in this type of comparison 1920×1200 was used for games and the best GPU.

    this would show a better “gaming value” because for gaming you’re not going to play at low res. run the benchmarks at what you’d actually play at and then see how much the CPU really makes a difference.

    • IntelMole
    • 11 years ago

    I’m a bit torn on the logarithmic price axis thing.

    I considered suggesting it last time around, but my reasoning then was that in addition to compressing the high end, it arguably more accurately represents price sensitivity. If someone is after the best value processor, they are going to be more sensitive to smaller fluctuations than someone who has much more “money no object” opinion at the very high end. Hence the expansion of the low end and compression at the high end might be appropriate.

    But using it then creates an arbitrary opinion of value. So you can argue that either way.

      • Damage
      • 11 years ago

      I think one of the problems with the price scale in this case, which people criticizing our choices seem to miss, is that prices are compressed right now due to AMD’s failure to execute with Phenom. They can’t range above ~$235 because they don’t have the performance to justify it. We’ve chosen to test a nice cross-section of the AMD lineup even though the price differences between the products are relatively small, AND we’ve tested quite a few competing, similarly priced Intel products against them. Hence things seem to be bunched up in the graphs because, well, these products are bunched up pretty closely in reality.

      We couldn’t include everything, but Intel’s product line does have some additional range between the Q6600/Q9450 and the Extreme editions. The fact that AMD’s doesn’t really isn’t our fault! Clipping the pricing at $320 or using a log scale would give a false impression of separation, in my view.

      And we *did* clip a true CPU market outlier in the form of the dual QX9775 Skulltrail system.

      Of course, whatever we do, folks will find reasons to complain. Just wanted to point out that we do think about these things and have reasons for the choices we make.

        • Kharnellius
        • 11 years ago

        Actually, I really like these scatter plots as I think they really do show how close everything actually is. You guys know how to actually use graphs without skewing the results.

        For instance, I prefer seeing the graph start at 0 on both axes and have each tick be consistent without any jumps in the increments to “save space”. I personally believe the activity of “saving space” is often intentionally done to skew perception of the graph.

        I don’t know if I am making any sense. If not just say so and I will try to clarify.

        All in all great job on what I consider to be a very useful and informative article (as usual).

        (Why isn’t this more prominent on the front page?)

        • flip-mode
        • 11 years ago

        I don’t think anyone is blaming you, Cyril, or Geoff for AMD not having a high-end CPU. The point that I and I think some others were trying to make is that for an article that it looking at CPU /[

        • IntelMole
        • 11 years ago

        I think the choices you made are all sensible good choices.

        I guess I was just musing a little there; I didn’t really intend it to be a criticism of the article.

    • Fighterpilot
    • 11 years ago

    Why not make a separate article for high end Extreme Editions and avoid skewing the graphs so severely?
    That would address most of the concerns posted so far and also make for a
    an occassional “Best of the best” article which I’m sure would be appreciated by all the horsepower geeks here ๐Ÿ™‚

      • ybf
      • 11 years ago

      If you did that you’d miss that the Q9450 is a very good deal compared with the QX6850. I can read the graphs fine, and having it all on one chart means you don’t have to read two articles and skew it mentally.

      I do agree we should see the E2XXX and C2Duo E4XXX in here, as long as we’re showing the 2’s and 3’s from AMD.

      But I think overclocking data should go in a separate article. You can’t get consistent overclocking from any CPU binning. Overclockability has a random factor to it that makes the “value” to the general public a gamble. The stock-speed numbers are very important on their own. A separate article would give the adventurous a handle on the “expected value” and leave the general public with this article to make their retail decisions.

    • FireGryphon
    • 11 years ago

    What happened to all of the scientific simulations that TR used to run?

    • willyolio
    • 11 years ago

    why did the q9450 disappear from the power consumption page?

    • MadManOriginal
    • 11 years ago

    Nice article.

    I have a proposal for an article, as if you’re bored without enough to do ๐Ÿ˜‰ Some time I know there will be a lull in hardware releases though. What I’d like to see is a test of CPU speed scaling in gaming for resolutions that are actually used by people playing games. CPU scaling at 1024×768 is all well and good for what it shows but it’s not very real-world. Something along the lines of resolutions and AA/AF used in GPU testing would be nice, just with a high-end GPU setup and scaling the CPU frequency instead of scaling GPUs. I want to see if there’s anything to the ‘faster dual core is better than slower quad core’ for resolutions and graphics settings at which people actually play.

    • SNM
    • 11 years ago

    You…you broke your graphs! What happened to zero-basing /[

      • Cyril
      • 11 years ago

      See the first page. We love zero-basing graphs too, but some of these would have turned out completely unreadable otherwise. Check this out:

      ยง[< https://techreport.com/mockups/office_nocrop.png<]ยง

        • MBIlover
        • 11 years ago

        As quoted by the knight in The Last Crusade (or the accompanying Pepsi commercial),”You have chosen…wisely.”

        • MrJP
        • 11 years ago

        If you took the Extreme Editions off that plot, I think it would probably be quite readable, and show the big picture that *[

        • SNM
        • 11 years ago

        I think that graph admirably demonstrates what it should be showing:the difference is practically nonexistent.

          • FireGryphon
          • 11 years ago

          Showing everything bunched up may not let you tell micro differences between each processor, but it tells you that there’s really no practical difference in choosing any CPU over the other for that application. I like zero graphs better. We need both.

        • JoeBobCPU
        • 11 years ago

        Here’s the thing. The unreadable, overlapping mess of dots in the zero-based graph is actually the perfect depiction of the essential takeaway from that benchmark: that for Microsoft Office it doesn’t matter at all what processor you buy.

        Looking at the original graph in the article, even being aware that the performance axis begins at an arbitrary place, a person instinctually tries to create meaning from the various points, as in, “Oh look, Processor X gets 150 more Microsoft Office synergies-per-second,” as if it matters. But it doesn’t. For that application suite, these differences are meaningless. The axis-adjusted graph shows off your hard work, but makes it difficult to extract any wisdom from the results.

        In fact, these results show that you should just dump this benchmark. Why waste the time? Just handwave it away with a little explanation about any CPU being good enough for Office and add something more useful, like a Handbrake DVD —> H.264 encoding test.

        You guys do great work. I’ve been a reader for years, and this kind of article is exactly why I’m so loyal. But I think you stumbled a bit with the execution on this one.

        • notfred
        • 11 years ago

        It would be nice to juggle the scales so that a linear scaling corresponds to a 45 degree line i.e. if you move up 10% in price and you get 10% more performance then the processor is 45 degrees from where you are now.

        It would then become immediately obvious if a processor family gets better bang-for-buck as you move up the family, or worse, or where the sweet point is in the middle.

        • oldDummy
        • 11 years ago

        Point well taken.

        • indeego
        • 11 years ago

        You can do both, they are called zoom graphs. Just put a ‘picture in picture’ of the smaller area in the white space of the graph.

        As for the legibility, you can draw lines.

        Come on, look at the comments here, it appears you are getting pretty universally panned hereg{<.<}g

        • Dagwood
        • 11 years ago

        I wish you would have included that rejected graph in the finished article as well. It paints a much clearer picture to me than the non-zero one. In fact I would argue that it shows a completly different result than the one you used. A low end AMD processor looks like it should be the winner of the price performance crown, just as it was on the last CPU value article.

        However, unlike the last value article you included the total system cost. As your graph on the last page shows this changes the picture. By adding the price of the system you can see that a cheep processor is not such a good deal. The difference between a paying $700 dollars for a system with a AMD 5600 and $750 for a Intel 7200 does not seem worth the performace loss. And the winner looks to be a Q9300.

        The real take home message here, is that it depends what your doing with the computer as to what is the best value. If your doing general MS office stuff, an intell 7300 or an AMD X2 5600 look like very good choices. I would argue that these are also good choices for a gaming computer as well. If the number one use for your computer is to fold@home a Q9300 is the clear choice.

      • oldDummy
      • 11 years ago

      Normally when you have a cluster where each point is within a very small area the graph is telling you that there is no defined attribute that distinguish these items.

      Hence:…erhrmmm…..change scale and exaggerate the delta.

    • Willard
    • 11 years ago

    question: what is the “percent” x axis on the last page? I can’t find an explanation in the text.
    Comment: In every page but the folding at home etc. page, the $/performance champs are mostly the 5600 and the 7200, with a few stray others in the mix.
    This fact deserves to appear in the conclusions on the last page since it is more or less the common, er, thread of the article.
    That is all.

      • Cyril
      • 11 years ago

      As we said on that page, the percentages are averages of performance scores in all 18 benchmarks. We did neglect to mention exactly what the 100% figure represents, though — that’s our baseline score that corresponds to the slowest chip, the Athlon X2 5600+. We’ve updated the text to clarify that.

        • Willard
        • 11 years ago

        Thanks. I’m glad you got around my getting the axes backward.

    • themattman
    • 11 years ago

    Nice analysis.

    My suggestions would be:

    1)Possibly put a separate graph not including the extreme edition processors, since people usually don’t consider those chips “values”.

    2)It would require more work, but I would like to see how the lower-end chips compare with the latest chips.

    • MrJP
    • 11 years ago

    Very nice summary of all your recent CPU testing from a value point of view, but I’ll add yet another voice asking for the Extreme Editions to be removed from the scatter plots. Just a couple of plots with those included would have been enough to make the point that you’d be nuts to buy any of them unless you’re after a professional rendering workstation where time literally is money.

    The final full system price comparison is very interesting. Arguably the system is a bit low on RAM and HDD capacity for the average system being built by people around here (take a look at the SBA forum), and probably should include typical case, PSU, and operating system costs to yield a more representative total system price. This would tend to favour the more expensive processors even more, which is perhaps unfair if people’s primary use will be gaming where performance will most-likely be GPU-limited. Perhaps that’s the next article, where the CPU & GPU analyses are combined to scientifically identify best-value systems for different usage scenarios – free 3D glasses included for viewing the plots. ๐Ÿ˜‰

    Sorry if this comes over as too critical. I really like these articles and hope they continue to be updated.

    • absinthexl
    • 11 years ago

    wow, we really have the worst-looking money in the world.

    • bdwilcox
    • 11 years ago

    I would have to concur that three things are amiss in this article. The Extreme processors are out of place and make everything more difficult to read, the exclusion of the Pentium E2XXX and Core2Duo E4XXX series is a serious oversight, and overclocking potential is an important factor in a processor’s perceived “value”. The overclocking potential could have been skimmed from HWBot with a bunch of disclaimers slapped on it saying YMMV.

    • Prototyped
    • 11 years ago

    You did an article on the value angle and completely omitted the Pentium Dual Core E2xx0 series? Wow. That’s a pretty egregious omission. (Not to mention the Celeron 4×0 series, Celeron Dual-Core E1x00 series, Sempron LE series and Athlon LE series, which are among the cheapest processors and can be expected to perform disproportionately better than the more expensive dual- and quad-cores you’ve chosen.)

    I’d rather see the currently available E2xx0 processors, the E2160, E2180, E2200 and E2220, rather than the E6750, which is essentially superseded /[

      • Willard
      • 11 years ago

      l[

        • axeman
        • 11 years ago

        Ummm, where? I’d be really interested to find out if the small cache of the E2180 makes that much difference. Pretty hard to tell if the huge caches on the Core Duos are doing much for performace if you have no basis for comparison.

          • Willard
          • 11 years ago

          I only have the vaguest recollection that celerons etc. are not what enthusiasts should be looking at. Must’ve been at least 4 years ago. ??. Sorry.

            • Prototyped
            • 11 years ago

            Not useful, ktnx.

    • Silus
    • 11 years ago

    Hey Cyril!

    Great article! These are definitely what I consider the most informative articles, since they pretty much tell people what is best bang for their buck!

    I found a few typos btw, in the WodlBench page:

    “but AMD’s offterings don’t peform as well in WorldBench’s application suite overall.”

    should be

    “but AMD’s offerings don’t perform as well in WorldBench’s application suite overall.”

      • Cyril
      • 11 years ago

      You’re right, thanks. Should be fixed now.

    • ludi
    • 11 years ago

    Saw the front-page image for this article and thought for a second that TR was taking a DIY foray into counterfeiting.

    • TravelMug
    • 11 years ago

    I agree about the need to remove the the $1000+ CPUs. I don`t think they bear any relevance for any selfrespecting TR reader and just make the relevant information hard to read. Especialy the last two overall graphs would improve by much. Or just add there another two where the performance axis is 100%-200% and the price is $0-$500.

      • muyuubyou
      • 11 years ago

      The best compromise would be to add a pair of graphs like those in the last page but zooming in by the X-axis and leaving the 1000+ chips out.

      90% of the units are crowding the same area because of a few rogue ones

    • cocobongo_tm
    • 11 years ago

    Hi everyone!

    Very interesting article! I’m actually amazed by the overall result and by the level of dominance that the E7200 displays. I was kind of hoping that AMD would manage to climb back into the saddle after the initial Phenom fiasco. It’s tough times for them: cheap and (a little) off the pace.

    I might be off topic for what I’m about to write and I apologize for this. But I’ve been thinking of writing it on one of the topics of the TR forum so here goes.

    What strikes me as odd in this “CPU war” that is being fought since the mid 90s (I think), is that whenever Intel makes a flop of a CPU, the plethora of big system integrators flock together around it and statements like “we know, AMD is faster but our customers are not into gaming and the Intel NNNNN is more than up to the job for that” start coming in.

    In the days of the first K7 (The Rise of AMD as I like to call it ๐Ÿ™‚ ), the Pentium III was behind AMD much like the current Athlons X2 and Phenoms X4 are behind the Core 2 family. Did we see Dell or HP rushing to AMD’s gates to buy everything they could get their hands on? No.

    They stared in disbelief, but not for long since their marketing departments quickly got up to speed and suddenly even if Intel was behind, it was somehow still THE choice if you’re serious about a computer. (If not – get and AMD). That made me soo mad! I mean…what more should be done about Intel! AMD had a faster chip on all counts! What do you want more??? “Well it was faster but it was not from Intel so we’re not interested.”

    When AMD was on top, it couldn’t boost sales because of people’s inertia (or should I call it “INTEL-TIA”) and now when it is not doing so hot anymore, the same people say “I told you so.”.

    I’m not an AMD lover. I was back in the K2-II era. But two months after I bought it, I was already fed up with it. The VIA MVP3 motheboard I was using coupled with the K6 was not stable at all. Plus, all the games were Intel optimized anyway (isn’t this odd??? they had 3DNow! first and no one was looking at it, but when Intel produced SSE everybody joined the chorus!)

    I then vowed never to buy an AMD again. I wanted stability. In the summer of 2002 I bought a Celeron 1.2 which I still have. My next rig will be an Intel as well. Why? Because after a wining streak with the X2, AMD failed to be consistent and to me, Phenom is a flop. AMD has to produce a cpu which running at lower freq. will beat and Intel running at a faster freq. And even then they will not be accepted by the system integrators league, but they will survive.

    All the best, lads!

    • mwg
    • 11 years ago

    Another awesome read. Thanks!

    • flip-mode
    • 11 years ago

    OK, these articles are terrific, but if there is one overarching request I’d like to make, it’s that you drop the high end CPUs and stretch the graphs in the X-axis (price) so that price differences in the all too crammed $100-$300 range are more perceptible. I’d argue that if you are buying a CPU above $500 that you aren’t as much interested in getting value as you are in getting the fastest CPU that you can.

    If you just excluded the three highest priced processors, look how much more dynamic your graphs would be.

      • ssidbroadcast
      • 11 years ago

      flip-mode is on point.

        • deruberhanyok
        • 11 years ago

        ssidbroadcast is on point about flip-mode being on point.

      • ew
      • 11 years ago

      I completely agree. Another potential solution is to have the x-axis show log(price). Although that add its own kind of confusion.

      I’d have also like a few <$100 processors.

      • Jambe
      • 11 years ago

      I agree. You ought to nip exorbitant high-end offerings & flesh out the “sweet spot” your audience cares about. You could include flagship chips in separate graphs as tools to illustrate disparity in price/perf (i.e., “you can get .90% of the performance of this chip for .15% the cost,” or something like that).

      Very nice graphs & good presentation overall, though. I just think you’d benefit from pruning the top-right side of things.

    • Thresher
    • 11 years ago

    Just a suggestion for the scatter graphs.

    You might consider putting in a big X in the background, then light colors in each quadrant to indicate low price/good performance, low price/poor performance, low price/good performance, and high price/poor performance.

    Additionally, because these are all pretty much clustered in the left side, I would probably scale the bottom line differently, perhaps logarthmically.

      • JustAnEngineer
      • 11 years ago

      Agreed. The EXTREME price of the Intel Extreme processors compresses the scale so much that it makes it hard to see the detail for the processors that are more sensibly priced.

        • Thresher
        • 11 years ago

        Statistically, it would be an outlier, so its value to the chart is very low, other than to show how bad price/value can be.

    • Hattig
    • 11 years ago

    Nice article, you can see where the profit comes in for the CPU makers, and AMD isn’t in that part of the curve. Here’s hoping that they can sort themselves out. How about throwing in a couple of Opterons on a dual-CPU motherboard as well? Of course, the RAM cost would go up there…

    • indeego
    • 11 years ago

    /[<"Volume prices aren't always representative of what consumers pay, but the two are close enough right now, and selecting "official" figures saves us the trouble of working out averages from fickle online retailers."<]/ weaksauce. You partner with one of the largest retailers online, wouldn't getting numbers from them be more representativeg{

      • Damage
      • 11 years ago

      Not really, no. Prices tend to fluctuate pretty widely depending on supply and demand in short order, making a single snapshot unreliable. I believe list pricing is a better source of info in this case.

        • indeego
        • 11 years ago

        ok. The Q9450 currently prices out at NE $360 and your chart has it at $316. Other recommended proc on the price/perf are not as significantly different… So I guess the only gripe would be the Q9450 isn’t /[https://techreport.com/articles.x/14756/9<]ยง

          • Damage
          • 11 years ago

          We didn’t include “emulated” CPUs in power testing. Our Q9450 was an underclocked QX9650, as we noted.

    • Firestarter
    • 11 years ago

    I find these graphs somewhat flawed, as the region of interest most most people doesn’t extend way up to $1600 processors. By including those, it becomes harder to accurately judge the value of the cheaper processors.

      • flip-mode
      • 11 years ago

      Gotta agree it would kinda make sense to just dismiss them, disqualify them as not to be considered “value” products in any way.

        • Mithent
        • 11 years ago

        I agree.. removal of all the Extreme processors would make the graphs significantly easier to read. You can sort of take for granted that you’re getting the fastest things around but you’ll pay a hefty price for the relatively small improvements.

          • titan
          • 11 years ago

          Yes, outliers are often excluded from results when they depart too far from the norm. The outliers are usually mentioned at some point.

            • Mithent
            • 11 years ago

            I know the most expensive system was excluded, but the existence of the other Extreme CPUs still bunches up the graph a lot.

    • flip-mode
    • 11 years ago

    The Force is strong with the E8400, E8500, and Q6600. My journey to the Dark Side is almost complete.

    Edit:

    Big high-five to TR for doing this article and for including total system price. Down low? My x2-3600 is too slow…

    • jjj
    • 11 years ago

    while some factors can be ignored when making such a chart i don’t think overclocking should be ignored,after all there are significant differences and (i guess) most ppl reading this are enthusiasts and would OC
    also i would have liked if an E21xx and maybe E1200 would have been included instead of some of the extreme cpu’s

      • Richie_G
      • 11 years ago

      I agree, it’d be interesting to see how the graph would look given an average overclocking of each chip, but then I guess we’re adding too many more variables to the equation. I’m sticking to my guns on my E8400 modestly OC’d to 3.6ghz.

      • flip-mode
      • 11 years ago

      Disagree. Overclocking results are far to variable and depend on various factors such as heatsink, motherboard, ambient temperature, voltage, and limits of the individual CPU. There are just too many vairables and no guarantees. The article as is has a well defined scope that has few variables, i.e., it’s results are applicable and meaningful to the widest range of people.

      For overclockers there are typically only a few CPUs that are even interesting – those with the highest multiplier for the dollar for the given microarchitectural generation and included feature set (i.e. amount of cache, virtualization support, yadda yadda) and then the number of cores as well.

        • titan
        • 11 years ago

        I agree with flip-mode. Overclocking is certainly a YMMV statistic, ergo totally illogical for inclusion in a comparison such as this.

    • Dposcorp
    • 11 years ago

    Nice Job guys.
    I am running a AMD 9500 now, and am looking to build a 2nd Quad Core.
    Looks like I may end up going with another AMD CPU.

      • flip-mode
      • 11 years ago

      Which one?

        • deruberhanyok
        • 11 years ago

        dunno about dposcorp but, to me, that 9750 is looking mighty tempting for the price… and the 9850 with the unlocked multi looking pretty good, too.

  • Pin It on Pinterest

    Share This