Primate Labs Blog

After a lot of nudging (sorry it took so long!) I’ve updated Primate Labs’ Mac Benchmark chart. The updated chart adds the following Macs:

• iMac (Late 2009)
• iMac (27-inch Late 2009)
• MacBook (Late 2009)

The Mac Benchmark chart is generated by gathering results from the Geekbench Result Browser and averaging scores for each Mac model and processor combination. However, not every combination is included; for a combination to be included it has to meet a number of criteria which includes the total number of results for that combination. If the number is below a certain threshold the combination isn’t included.

Which leads me to something surprising; I was hoping to include Apple’s latest Mac Pro update, the single-processor Mac Pro @ 3.33GHz, but there are only two results for it in the Geekbench Result Browser. I was expecting more results for this particular model.

There are, however, a large number of dual-processor Mac Pros @ 3.33GHz in the Geekbench REsult Browser even though Apple doesn’t offer this configuration for sale. Apparently adding a second processor (or replacing both processors) in a Mac Pro is a popular upgrade. It’s a powerful upgrade, too, as one of these upgraded Mac Pros currently has the sixth highest score in the Geekbench Result Browser.

Late last week Geekbench results from what appeared to be a yet-unreleased MacBook Pro appeared on the Geekbench Result Browser. These results generated a lot of excitement, along with a lot of speculation; people wanted to know if these results were legitimate.

I spoke with Wired’s Brian Chen earlier today and shared my thoughts regarding the legitimacy of these results:

Geekbench creator John Poole told Wired.com that Geekbench results can be faked, but he believes the Core i7 MacBook Pro benchmark is real.

“If the system information for this system has been faked, it’s a convincing fake; there aren’t any inconsistencies in the system information, the Mac OS X build number matches Apple’s numbering scheme for new hardware builds, and the processor matches people’s expectations for what will appear in the refreshed MacBook Pro,” Poole said. “In other words, I believe this result is authentic.”

I have a feeling we’ll find out sooner rather than later whether this result is real or not!

I’ve updated the Mac Benchmark chart with Geekbench results for the new MacBook Pros announced at WWDC earlier this month. I’ve included the results for these new MacBook Pros below, along with results for the previous generation of MacBooks and MacBook Pros.

Results were collected from the Geekbench Result Browser for Macs with standard processors (i.e., no processor upgrades or overclocked processors). I’ve reported the average overall score for each model and processor combination. A score of 1000 is the score a Power Mac G5 @ 1.6GHz would receive, and higher scores are better.

Results

A few years ago I bought a 15″ PowerBook with an illuminated keyboard. At the time, I didn’t think much of the keyboard:

The keyboard backlight is an interesting idea, and looks very cool, but I’m not sure how practical it is. It’s only really visible when in a pitch-black room, and I can’t remember the last time I used a laptop in a room that dark (save when I was testing the keyboard backlight).

A few days ago, I bought a 15″ MacBook Pro with an illuminated keyboard and unlike the first time around I can’t say how much I’ve missed, and how much I appreciate, having a backlit keyboard.

You see, in between the PowerBook and the MacBook Pro I had a MacBook with a regular keyboard. That keyboard was hard enough to read in dim lighting, let alone complete darkness. Now, though, I have no problem reading the keys.

If you’re considering between an illuminated and a regular keyboard, might I recommend going with the illuminated keyboard. Even if it doesn’t seem like much at first, you’ll soon come to appreciate it.

Welcome to the most recent installment of Primate Labs‘ Mac Performance chart, a carefully crafted compilation of Mac performance that’s generated from the user-submitted results in the Geekbench Result Browser.

Like other updates, this update includes the latest Apple hardware (in this case the MacBook, the MacBook Air, and the MacBook Pro Apple released in November). Unlike other updates, this update doesn’t just present all of the Macs together in one large chart; it also splits the Macs into four different categories making the results easier to view and understand:

• Pro Laptops (MacBook Pros, MacBook Airs, and PowerBooks)
• Pro Desktops (Mac Pros and PowerMacs)
• Consumer Laptops (iBooks and MacBooks)
• Consumer Desktops (eMacs, iMacs, and Mac minis)

Setup

Results were collected from the Geekbench Result Browser for Macs with standard processors (i.e., no processor upgrades, overclocked processors, or Hackintoshes) with at least 512MB of RAM.

I’ve reported the average overall score for each model and processor combination. If you’re unfamiliar with Geekbench and how it measures performance, a score of 1000 is the score a Power Mac G5 @ 1.6GHz would receive. Higher scores are better.

Also, keep in mind that Geekbench 2 only measures processor and memory performance which is why, for example, MacBook and MacBook Pro scores are so similar, despite both having radically different graphics adapters.

Pro Laptop Performance

What strikes me every time I look at this chart is how stable (or stagnant, depending on how charitable you’re feeling) PowerBook performance was after the introduction of the aluminum PowerBooks; it’s a reminder of how out-of-date the PowerPC G4 was then. When Apple switched to Intel processors, Pro laptop performance almost tripled overnight. Even the slowest MacBook Air, which many people regard as “too slow” to do “real work” on, is over twice as fast as the fastest PowerBook.

Pro Desktop Performance

Pro desktops weren’t held back by the PowerPC G4 because they were able to take advantage of IBM’s PowerPC G5. While the PowerPC G5 never reached the 3GHz promised by Steve Jobs in 2003, they did provide better performance than the PowerPC G4 and served the PowerMac well until Apple switched to Intel.

Now, with the switch to Intel, the latest Mac Pro is the fastest Mac available (and one of the fastest computers available) thanks to its two quad-core processors.

Consumer Laptop Performance

Much like Apple’s pro laptops, Apple’s consumer laptops were held back by the out-of-date PowerPC G4; switching to Intel almost tripled MacBook performance overnight.

Consumer Desktop Performance

While the switch to Intel brought performance benefits to consumer desktops thanks to faster processors, they also brought performance benefits thanks to multi-core processors. Think about it; none of Apple’s PowerPC-based consumer desktops are multi-core, while almost all of Apple’s Intel-based consumer desktops are multi-core. With so much multi-core aware software available today for Mac OS X, running a single-core Mac seems almost quaint.

Mac Performance

Final Thoughts

I’m always amazed at the performance drought the PowerPC G4 inflicted on Apple in general and on Apple laptops in particular before the switch to Intel; laptop performance is generally unchanged between 2003 and 2006 (desktops were at least able to use the PowerPC G5).

Once Apple switched to Intel, performance has increased (and increased significantly) with almost every update. Apple shouldn’t have to worry about getting stuck in another performance drought again.

When Apple released the new MacBooks and MacBook Pros last month, Apple also updated the MacBook Air. While the changes for the MacBook Air weren’t as radical, there were still a number of improvements made such as a new chipset, faster RAM, better graphics cards, and faster processors (at least for the high-end model).

What do these improvements mean for processor performance? My initial guess is that, like the new MacBooks and MacBook Pros, it doesn’t mean much, but let’s look at some Geekbench results to see if that’s the case.

Setup

I used Geekbench 2, our cross-platform benchmark, to measure the processor and memory performance of the new MacBooks and MacBook Pros. Here is the configuration of the four test machines.

• MacBook Air (Late 2008)

  • Intel Core 2 Duo @ 1.86GHz or
  • Intel Core 2 Duo @ 1.60GHz
  • 2.00 GB 1067 MHz DDR3
  • Mac OS X 10.5.5 (Build 9F2523)

• MacBook Air

  • Intel Core 2 Duo @ 1.80GHz or
  • Intel Core 2 Duo @ 1.60GHz
  • 2.00 GB 667 MHz DDR2 SDRAM
  • Mac OS X 10.5.5 (Build 9F33)

Geekbench 2 scores are calibrated against a baseline score of 1000 (which is the score a Power Mac G5 @ 1.6GHz would receive). Higher scores are better.

Results

Overall Performance

Integer Performance

Floating Point Performance

Memory Performance

Stream Performance

Conclusions

I thought the new MacBook Air wouldn’t be that much faster than the old MacBook Air, and it turns out I’m partly right (or partly wrong, depending on how you look at it). While the high-end MacBook Air is only slightly faster, the low-end MacBook Air is almost 10% faster (a dramatic improvement).

Keep in mind, too, these benchmarks ignore one of the big improvements to the MacBook Air — the new graphics chipset. While Geekbench isn’t able to measure graphics performance (an unfortunate limitation) benchmarks elsewhere show the huge gains NVIDIA’s integrated graphics have over Intel’s integrated graphics.

So if you’re thinking about buying a MacBook Air, don’t be too concerned that processor performance hasn’t improved for the high-end MacBook Air; the new graphics cards will certainly make up for it. Plus, while many complain about the low performance of the MacBook Air, remember it’s faster than the first round of Intel-based laptops and all of the dual-processor PowerMac G5s.

Last week when I looked at the performance of Apple’s latest MacBook Pros I didn’t have results for the high-end CTO MacBook Pro (with the 2.8GHz Core 2 Duo processor).

Now that I’ve got results for the high-end MacBook Pro I thought I’d revisit MacBook Pro performance to see how the high-end MacBook Pro compares against the last two generations of MacBook Pros. Sure, it’s the fastest laptop Apple’s ever produced, but just how much faster is it?

Setup

Like last time, I used Geekbench 2, Primate Labs’ cross-platform benchmark, to measure the processor and memory performance of the new MacBook Pros. Here’s the configuration of the test machines:

• MacBook Pro (Late 2008)

  • Intel Core 2 Duo T9600 @ 2.80GHz or
  • Intel Core 2 Duo T9400 @ 2.53GHz or
  • Intel Core 2 Duo P8600 @ 2.40GHz
  • 4.00 GB 1067 MHz DDR3
  • Mac OS X 10.5.5 (Build 9F2114)

• MacBook Pro (Early 2008)

  • Intel Core 2 Duo T9500 @ 2.60GHz or
  • Intel Core 2 Duo T8300 @ 2.40GHz
  • 4.00 GB 667 MHz DDR2 SDRAM
  • Mac OS X 10.5.5 (Build 9F33)

Geekbench 2 scores are calibrated against a baseline score of 1000 (which is the score a Power Mac G5 @ 1.6GHz would receive). Higher scores are better.

Results

Overall Performance

Integer Performance

Floating Point Performance

Memory Performance

Stream Performance

Earlier this week Apple released new MacBooks and MacBook Pros. While the biggest changes are the new enclosures and the new graphics cards, the new laptops also have a new chipset (an NVIDIA chipset instead of an Intel chipset) and faster DDR3 memory. These last two features might improve processor performance even though the new MacBooks and MacBook Pros don’t really have upgraded processors.

So, do the new MacBooks and MacBook Pros have better processor performance than the old MacBooks and MacBook Pros? Does the new memory and chipset make a difference? Let’s find out!

Setup

I used Geekbench 2, our cross-platform benchmark, to measure the processor and memory performance of the new MacBooks and MacBook Pros. Here is the configuration of the eight test machines.

• MacBook Pro (Late 2008)

  • Intel Core 2 Duo T9400 @ 2.53GHz or
  • Intel Core 2 Duo P8600 @ 2.40GHz
  • 4.00 GB 1067 MHz DDR3
  • Mac OS X 10.5.5 (Build 9F2088)

• MacBook Pro (Early 2008)

  • Intel Core 2 Duo T9500 @ 2.60GHz or
  • Intel Core 2 Duo T8300 @ 2.40GHz
  • 4.00 GB 667 MHz DDR2 SDRAM
  • Mac OS X 10.5.5 (Build 9F33)

• MacBook (Late 2008)

  • Intel Core 2 Duo P8600 @ 2.40GHz or
  • Intel Core 2 Duo P7350 @ 2.00GHz
  • 2.00 GB 1067 MHz DDR3
  • Mac OS X 10.5.5 (Build 9F2114)

• MacBook (Early 2008)

  • Intel Core 2 Duo T8300 @ 2.40GHz or
  • Intel Core 2 Duo T8100 @ 2.10GHz
  • 2.00 GB 667 MHz DDR2 SDRAM
  • Mac OS X 10.5.5 (Build 9F33)

Geekbench 2 scores are calibrated against a baseline score of 1000 (which is the score a Power Mac G5 @ 1.6GHz would receive). Higher scores are better.

Results

Overall Performance

Integer Performance

Floating Point Performance

Memory Performance

Stream Performance

Conclusions

Somewhat suprisingly, processor performance is virtually unchanged in the new MacBooks and MacBook Pros; the new NVIDIA chipset and the faster DDR3 memory haven’t managed to help increase processor performance.

Why is this surprising? Well, with other hardware revisions Apple’s managed to make some modest gains in processor performance with chipset or memory upgrades without upgrading the processor itself. Here, though, performance is practically the same (and acutally down in some cases) with the new chipset and memory. I think NVIDIA has a ways to go (at least compared to Intel) when it comes to implementing chipsets.

Does this mean the new laptops are a bad upgrade? Hardly! While the NVIDIA chipsets might not be as good as the Intel chipsets when it comes to processor performance, they far surpass Intel when it comes to graphics performance. This increased graphics performance is a huge gain for the MacBook, and makes the MacBook a compelling machine for people looking for a smaller MacBook Pro.

Throw in the new, sturdier chassis and you’ve got a compelling upgrade for almost any Mac laptop user.

Back in April Apple released new iMacs which means the previous Mac Performance chart is out of date. Time for an update!

Setup

Results were collected from the Geekbench Result Browser for Macs with standard processors (i.e., no processor upgrades, overclocked processors, or Hackintoshes) with at least 512MB of RAM.

As always, I’ve reported the average overall score for each model and processor combination. If you’re unfamiliar with Geekbench and how it measures performance, a score of 1000 is the score a Power Mac G5 @ 1.6GHz would receive. Higher scores are better.

Also, keep in mind that Geekbench 2 only measures processor and memory performance which is why, for example, MacBook and MacBook Pro scores are so similar, despite both having radically different graphics adapters.

Results

Apple reminded me this week that my ADC Select membership is about to expire, and suggested that I renew my membership so I can continue to take advantage of all the benefits it offers.

I’m not sure I’m going to, though. While the hardware discount was nice (hello, MacBook!), and access to the Leopard Early Start Kit was incredibly useful, I didn’t take advantage of any of the other benefits (like the technical support incidents). Plus, now that Leopard will be in everyone’s hands tomorrow, the major benefit (pre-release software) disappears (at least until Apple announces 10.6, which won’t be until WWDC 2008 at the earliest).

Also, like other Mac developers, I’m annoyed that we weren’t given access to the final version of Leopard before it was released. While I’m fairly confident both Geekbench and Wiinote will work with the final version of Leopard, I’ve only been able to test them with the last pre-release version of Leopard. It’s entirely possible that either one (or both) will not work with the final version of Leopard.

I’d be less annoyed if Apple didn’t give anyone access to Leopard before it was released, but a number of journalists have access to the final build of Leopard. Plus, the final build is all over the internet. I could download it from BitTorrent or Usenet and while I don’t think that would be morally wrong (I’ve paid for a copy with my ADC membership), it’s certainly legally wrong. So why’s Apple not giving everyone with the Leopard Early Start Kit early access to the final build? I can’t figure it out.

Compare Apple with Microsoft; developers with an MSDN subscription received copies of Windows Vista well before it was released to the general public, giving the developers a chance to hammer out any final compatibility issues before consumers encountered them.

So, will I renew my ADC Select membership? Probably not. There’s not a lot of value in it right now for me, and I’m kind of annoyed that Apple treats me (and other developers) as necessary evils rather than trusted partners.