Geek Patrol

Overclocked Geekbench Comparison (December 2006)

Back at the beginning of December I posted a Geekbench comparison featuring a number of current (and not-so-current) processors. For those of you keeping score at home, you might’ve noticed that the comparison only looked at non-overclocked processors. Since Geekbench is popular with the overclocking crowd, I thought it’d be interesting to look at non-overclocked and overclocked results for some of the more popular processors in the Geekbench Result Browser database.

Setup

Much like December’s Geekbench comparison, I’ve collected results from the Geekbench Result Browser database for the 32-bit Windows version of Geekbench running on systems with at least 512MB of RAM.

Instead of averaging the overall scores for each processor (as was done in previous comparisons), I’ve plotted each score for each processor in a graph that shows the effect of processor frequency on Geekbench score (each processor gets its own graph). Higher scores are better.

Results

Conclusion

I’m not at all surprised to see that Geekbench scores for the same model of processor tend to increase with processor speed; after all, Geekbench is a processor benchmark! I’m also not surprised to see that Geekbench scores don’t just depend on processor frequency; looking at the graphs it’s clear that other factors play a role (such as memory configuration).

What is interesting is the clusters of results at certain frequencies for the Core 2 Duo processors. There’s a cluster of results around the stock frequency, and a second cluster around a higher frequency; I wonder if this second cluster shows the overclocking limit of each particular Core 2 Duo processor (without resorting to exotic cooling systems).

So, if you want to improve your Geekbench score, the most obvious way to do it is by overclocking your processor. However, as the graphs show, there are other, less obvious ways to do it, too (such as investing in faster memory).

Better Living Through Open Source

Every month I get a number of requests for the Geekbench source code (for a variety of different reasons, ranging from curiosity to a desire to port Geekbench to a different platform). For now, though, Geekbench is closed-source software. Releasing Geekbench as open-source software would cause a number of problems, the major one being it’d make comparing Geekbench results difficult, if not impossible. Part of the value of Geekbench is that it makes it easy to compare results across different configurations and different platforms when run by different users, in part because Geekbench is built with a similar set of compiler and optimization options across all platforms.

If alternate builds of Geekbench were made available, there’s no guarantee what options or optimizations it was built with (let alone whether it was built with changes made to the source code); comparing scores generated by different binary versions of Geekbench would become pointless.

That said, there are certainly parts of Geekbench that might benefit from being released as open-source software. There’s a lot of cross-platform code that’s not specific to Geekbench that might be of interest to other developers.

One particular example is the system information gear in Geekbench. Gathering system information can be tricky; you need to know where to look for the system information, and when the system information can be wrong. Whenever I port Geekbench to a new platform, most of the time is spent writing and debugging the system information code for the new platform; porting the benchmarks and Geekbench really doesn’t take very long at all.

So I created Geekinfo, which is based on the cross-platform system information gear in Geekbench. While it’s got some limitations (such as a C++-only API), it works well and it’s small and easy to understand. Right now it builds on Windows using Visual C++ 2005, on Solaris using Sun Studio 11, and Linux and Mac OS X using gcc 4.0.

Geekinfo is hosted over on Google Code. You can grab the latest source code using subversion, or download a .tar.gz file containing one of the point releases. Both should just build and run; let me know if you have problems with either.

Geekinfo is licensed under the LGPL; which means closed-source applications can use Geekinfo provided any changes made to Geekinfo are also released under the LGPL. Of course, if this license (for whatever reason) doesn’t appeal to you, I’m willing to consider exceptions on a case-by-case basis.

If you’re interested in contributing, there are a number of ways you can help. New features or bug fixes are always welcome, as are ports to other platforms (such as Solaris/SPARC or FreeBSD/x86). As always, comments and suggestions are welcome, too.

I hope you find Geekinfo interesting and useful, and if you end up using Geekinfo in a project I’d love to hear about it.

Eight Core Mac Pro Benchmarks

As we’ve known for a while now, it’s possible to upgrade the processors in the Mac Pro. Now that quad-core Xeon processors are available now, some enterprising Mac user has gone out and upgraded the processors in their stock Mac Pro from two dual-core Xeons to two quad-core Xeons, and run Geekbench on their new eight-core Mac Pro. I thought it’d be interesting to see how the performance of this eight-core Mac Pro compares to a couple of the stock four-core Mac Pros.

Setup

• Mac Pro (Quad-Core Xeon @ 2.33GHz)

  • Two Intel Xeon E5345 quad-core processors @ 2.33GHz
  • 4096MB RAM
  • Mac OS X 10.4.8
  • Geekbench 2006 64-bit (Build 242)

• Mac Pro (Dual-Core Xeon @ 3.00GHz)

  • Two Intel Xeon 5160 dual-core processors @ 3.00GHz
  • 4096MB RAM
  • Mac OS X 10.4.8
  • Geekbench 2006 64-bit (Build 242)

• Mac Pro (Dual-Core Xeon @ 2.66GHz)

  • Two Intel Xeon 5150 dual-core processors @ 2.66GHz
  • 4096MB RAM
  • Mac OS X 10.4.8
  • Geekbench 2006 64-bit (Build 242)

I’m reporting the baseline score, rather than the raw score, for each benchmark (where a score of 100 is the score a Power Mac G5 1.6GHz would receive). Higher is better.

Conclusion

When it comes to processor-intensive multi-threaded tests, the eight-core Mac Pro is certainly faster than both four-core Mac Pros. However, the eight-core Mac Pro is slower than both four-core Mac Pros when it comes to single-threaded tests, thanks to the fact that its cores run at 2.33GHz (compared to 3.00GHz and 2.66GHz for the four-core Mac Pros). Heck, any processes that uses four threads or less will be faster on the four-core Mac Pros than the eight-core Mac Pro; you’re only going to notice a performance benefit when processes take advantage of the extra cores the eight-core Mac Pro offers.

Should you run out and upgrade your four-core Mac Pro to an eight-core Mac Pro? If you have to ask that question, then probably not. This is the sort of upgrade that appeals to hardware enthusiasts and users that really need more than four processor cores.

Would it make sense for Apple to offer an eight-core Mac Pro? Maybe, if the clock speed of the quad-core processor was comparable to the clock speed of the dual-core processor. Even then, I’m not sure how many users would find the four extra cores worth the extra money.

That doesn’t mean I can’t covet an eight-core Mac Pro, though.

Quad Core Benchmarks

CDRinfo posted some Intel quad-core processor benchmark results earlier this month, including results from Geekbench. While I’ve got concerns about CDRinfo’s benchmark methodology (they under-clocked CPUs to emulate CPUs missing from their lineup, like the E6600 and QX6600), it’s still interesting to see how benchmarks like Geekbench perform on a quad-core CPU.

Geekbench IRC Channel

Since I’m a huge geek, I’ve set up an IRC channel for Geekbench on macspeak.net. Feel free to drop by #geekbench and ask questions or just say hi. If you need an IRC client, I recommend Colloquy on Mac OS X and mIRC on Windows.

Forums Back Online

I spent some time this evening and set up some new forums (for those of you just tuning in, the old forums died when we upgraded our hosting plan). Give ‘em a shot and let me know what you think of them!

32-bit vs 64-bit Vista Benchmarks

The folks over at 64-bit-computers.com have published an article comparing the performance of 32-bit and 64-bit Windows Vista, and they’re using Geekbench as their benchmarking tool.

From the article:

Both 32-bit and 64-bit tests were done on a budget AMD Sempron 2800+ powered desktop PC, featuring Asus K8U-X motherboard with 512MB of main memory.

The 32-bit Windows Vista Ultimate achived 99.8 GeekBench points, while the 64-bit version of the same OS gained 110.7 points.

(via Digg)

Geekbench Comparison (December 2006)

Our obsession with benchmarks and Geekbench continues! Earlier this year we did a couple of Geekbench Comparisons using a number of different machines. While the comparisons were popular, a number of people thought we were being, well, less than fair since we didn’t have enough high-end CPUs from AMD and Intel.

So, for this Geekbench comparison, I’m including most of the recent AMD and Intel desktop processors (along with some not-so-recent desktop processors) along with a wide variety of recent Mac models, ranging from the PowerBook G4 to the Mac Pro.

This time ’round, instead of using results from a single computer, I’m using average results from all the computers in the Geekbench Result Browser database (provided the results meet certain criteria outlined below). To make sure I’ve got a good sample size for each computer, I’m only including computers (or processors) that have at least three distinct results in the result browser. This should help compensate for differences in components (such as RAM) between different computers.

Setup

For Intel and AMD processors, results were collected from Geekbench for Windows (32-bit) running on systems with processors running at stock speed (i.e., no over-clocked processors) and with at least 512MB of RAM.

For Mac systems, results were collected from Geekbench for Mac OS X (32-bit) running on systems with standard processors (i.e., no over-clocked processors or processor upgrades) and with at least 512MB of RAM.

I’m reporting the average overall score for each processor, where 100 is the score a Power Mac G5 @ 1.6GHz would receive. As always, higher scores are better.

AMD Processors

You can see differences in performance for each processor depending on its socket type (thanks to the type of RAM each socket uses); Socket 754 (which uses single-channel DDR RAM ) is slower than Socket 939 (which uses dual-channel DDR RAM), which in turn is slower than Socket AM2 (which uses dual-channel DDR2 RAM)

You can also see that AMD model numbers are fairly consistent; an AMD Athlon 64 3800+ offers roughly the same performance as an AMD Athlon 64 X2 3800+ (180.2 vs 172.5 for Socket AM2), Unfortunately both are slower than a Pentium 4 3.8GHz (211.0), which suggests the model numbers are somewhat optimistic when compared to Intel CPUs.

Intel Processors

Looking at these numbers, Pentium D processors don’t seem particularly attractive; on one end, high-end single-core Pentium 4 processors offer better performance, while on the other end dual-core Core 2 chips offer better performance and better power-consumption. I’m not surprised that it’s getting harder and harder to buy Pentium D processors these days (save for the low-end Pentium Ds).

Another interesting thing to note is the jump in performance between the Core 2 Duo E6400 and the Core 2 Duo E6600; moving from 2×1MB to 2×2MB of L2 cache makes a big difference for the Core 2 Duo processor. Speaking of Core 2 Duo processors, the Core 2 Duo X6800 is the fastest dual-core processor in this comparison (with only the dual-processor dual-core Mac Pro @ 3.0GHz being faster overall).

Mac Models

Mac mini performance almost doubled with the switch from a PowerPC G4 @ 1.5GHz to an Intel Core Solo at 1.5GHz with no increase in clock speed. Zomg! iBook and PowerBook (er, MacBook and MacBook Pro) performance also increased dramatically with the switch to the Intel Core Duo (and later to the Intel Core 2 Duo), especially since both went from single-core to dual-core processors.

iMac and PowerMac (er, Mac Pro) performance didn’t benefit as much with the switch to Intel processors since the PowerPC G5 isn’t nearly as old and busted as the PowerPC G4, but the high-end Mac Pro does have a considerable performance advantage over the top-end Power Mac G5. Even the low-end Mac Pro is competitive against the high-end Power Mac G5 (and is substantially cheaper to boot).

Final Thoughts

It’s interesting to see where the bottlenecks in processor and memory performance have occurred over the years (hello, PowerPC G4!) and how processor and computer manufacturers worked around these bottlenecks (hello, Intel!). Really, though, it’s graphs like these that make me want to “accidentally” spill a gallon of water on my three-year-old PowerBook G4 so I can go out and get a shiny new MacBook Pro.

Wii Impressions

aListair picked up a Wii last week and he brought it over this evening so Deborah and I could give it a try.

Wii Sports, which comes bundled with the Wii, is an interesting game. It’s really easy to pick up! Both Deborah and I managed to figure out the game mechanics fairly quickly, and we were able to stay somewhat competitive with aListair. I’m not sure how long it’d hold my interest (I can see it becoming old fairly quickly), but it’s a great game for casual gamers.

Both the Wii Remote and the Nunchuk work okay. There were times when the Wii seemed to miss gestures I was making, and pick up gestures I wasn’t making, which makes for a frustrating gaming experience. The speaker in the Wii Remote is a nice touch; I think it’s better at adding feedback than force feedback. It is a little unnerving, though, when you hear someone else’s Wii Remote whistling a few inches from your ear.

The Wii’s graphics looked good on my standard definition TV, although I don’t think Wii Sports is the best game to use when evaluating the Wii’s graphics.

After playing with one for a couple of hours, I’ve decided that I like the Wii, but not enough to rush out and get one. I think I’ll wait until there are more games available, and developers have figured out how best to use the Wii Remote and the Nunchuk controllers (sort of like how they had to figure out what to do with the two screens on the Nintendo DS).