Hands-On Review and Speed Tests--15" 2.4GHz MacBook Pro LED

We ordered a new 15" MacBook Pro Core 2 Duo with LED-backlit screen the day they were announced, i.e. Tuesday, June 5, with 2-day delivery. It arrived from Shanghai on June 12 and we have been using it heavily ever since. The new 15" MacBook Pro LED comes with Mac OS X 10.4.9 but with an updated build number (8Q1058 vs 8P2137).

It was a BTO from the online Apple Store because we wanted the larger 200GB hard drive even though it is slower at 4200rpm vs 5400 or 7200rpm. With more music, videos and Parallels Desktop for Mac with two Windows virtual machines crowding the 160GB drive in our former 15" MacBook Pro Core 2 Duo, we badly needed the additional space.

Apple promised a slightly faster CPU at 2.4 vs 2.33GHz and 800 vs 667MHz frontside bus, much faster graphics (Nvidia GeForce 8600M GT vs ATI Radeon X1600), better battery life from the LED-backlit screen and improved CPU management, more RAM capability (4GB vs 3GB) and a 0.2lb lighter Mac. What's not to like? The decision seemed a no-brainer to us, partly because we're a die-hard gamer and the faster graphics were especially attractive and partly because we are a road warrior and the longer battery life and reduced weight are both highly welcome.

This time we got the glossy screen instead of the standard matte screen. After receiving positive comments from many glossy-screen MacBook owners, comparing the two screens on MacBook Pros at the local Apple Store and seeing so many high-end Windows machines with glossy screens, we decided to try them. When it comes to screens, we are more of a consumer than a pro. We really like the glossy's color saturation better and decided we probably wouldn't mind the glossy's seemingly minor reflection issue all that much.

In summary, the salient features that differentiate our new MacBook Pro are:

  • 2.4GHz Intel "Santa Rosa" Core 2 Duo CPU with 800MHz frontside bus
  • 15" LED-backlit glossy screen
  • Nvidia GeForce 8600M GT graphics with 256MB VRAM
  • 200GB Toshiba 4200rpm hard disk drive

The macBook Pro that it replaced has:

  • 2.33GHz Intel "Merom" Core 2 Duo CPU with 667MHz frontside bus
  • 15" CCR standard matte screen
  • ATI Radeon X1600 graphics with 256MB VRAM
  • 160GB Hitachi 5400rpm hard disk drive

Both have 2GB of RAM, an 8x slot-loading SuperDrive, the same complement of ports and were priced the same. The biggest difference is in the graphics systems.

The LED Screen

The only externally visible new feature is, of course, the new 15" MacBook Pro's LED-backlit glossy screen. We tried and failed to get some useful photos of the screens of the former and newer MacBook Pros. Qualitatively, the LED-backlit glossy screen appears to be a bit brighter than the CCR matte screen at full brightness but just barely. This difference may, in fact, be due simply to the CCR-backlit matte screen being 6 months older. The LED-backlit glossy screen also appears to be a bit more evenly lighted over the whole screen. And, as we noted before, we do like the better color saturation of the glossy screen over the matte screen when viewed directly. When viewed from an angle, the white or light gray colors shift to a light greenish/yellowish tint on the glossy LED-backlit screen much more so than the matte CCR-backlit screen.

Outside on a deck in the shade with the screens turned up to full brightness, an image on the LED-backlit glossy screen is little more visible than on the CCR-backlit matte screen. In direct sunlight, the image on the LED-backlit glossy screen is clearly more visible than on the CCR matte screen. On the other hand, when the screen is dimmed to extend battery life, the glossy LED-backlit screen's reflection issue becomes more prominent. In a room with lots of windows and overhead lighting, like an airport terminal, a fully dimmed glossy LED-backlit screen is much more difficult to read than the matte CCR screen. On a darkened airplane, the fully dimmed glossy LED-backlit screen is more easy to read in our experience. So, there are trade-offs to consider.

Our view is that the LED-backlit screen is slightly superior in visual quality to the CCR-backlit screen when both are relatively new. The gap is expected to widen as both screens age. After a week's use, all things considered, we are happy with the glossy screen choice.

Performance

The major performance difference is expected to be in graphics. Based on Apple's posted tests, the new MacBook LED with an Nvidia GeForce 8600MGT ran 37, 50 and 57 percent faster than a 2.16GHz 17" MacBook Pro Core Duo with the ATI Mobility Radeon X1600 when respectively running Motion, Doom 3 and Quake 4. It is curious and perhaps ominous that Apple chose to benchmark against the earlier MacBook Pro Core Duo instead of the second generation MacBook Pro Core 2 Duo. We posted gains of 33 and 30 percent for Doom 3 and Quake 4 between those two older models. Taking that into account, we would still expect a gain on the order of 17 and 27 percent for the new MacBook Pro LED over our MacBook Pro Core 2 Duo.

The benchmarks posted by NotebookCheck.com positions the Nvidia 8600M GT graphics with only 128MB VRAM way ahead of the older Radeon X1600 with 128MB VRAM and very nearly the same as the very fast ATI Mobility Radeon x1900 with 256MB VRAM.

Comparison of Graphics Cards by NotebookCheck.com
Card
VRAM Pixel Shaders Vertex Shaders 3DMark03 3DMark05 3DMark06
ATI Mobility Radeon X1900
256
36
8
11800
7080
3460
Nvidia GeForce 8600M GT
128
16
16
10840
6029
3323
ATI Mobility Radeon X1600
128
12
5
8000
3500
1800

So our expectations were high that the new Nvidia GeForce 8600M GT with 256MB VRAM would out-perform handily the older ATI Mobility Radeon X1600.

To find out, we ran our standard set of speed tests on our Intel-based MacBook Pros.The tests consist of four canned benchmark applications, five so-called real world tests and six 3D games.

Here are the results for the MacBook Pro LED vs the MacBook Pro C2D:

Test*
MacBook Pro LED
MacBook Pro C2D
Performance Change (%)
Benchmarks (scores)
Altivec Fractal v1.3 (GFLOPS)**
10.0
9.9
+1
Cinebench v9.6 - Rendering (2X)
738
724
+2
- OpenGL Hardware Lighting
3903
4321
-10
Let1kWindowsBloom v1.0 (sec)**
8
8
0

Xbench v1.3 - CPU

122
121
+1
- Threading
236
231
+2
- Memory
135
129
+5
- Quartz graphics
175
143
+22
- OpenGL graphics
137
207
-34
- User interface
432
377
+15
- Disk
31
34
-9
"Real World" Tests (seconds)
Startup
31
25
-24
Shutdown
10
9
-11
Duplicate 500MB Folder
61
41
-49
Scroll Adobe Reader v8.1
11
10
-10
Export DV from iMovie v6.0.3
52
48
-8
3D Game Tests (framerate)
Quake III Arena v1.32UB
407
626
-35
Call of Duty v1.5**
43
46
-7
Call of Duty 2 v1.3
~25
~75
-67
Halo v2.0.1
85
69
+23
Doom 3 v1.3.1
85
90
-5
Quake 4 v1.2
66
67
-1
*Average of three runs. **Non-Universal. Blue is much slower. Red is much faster.

There's good news and bad news stemming from our speed tests. The graphics results are mixed and the slower Toshiba drive is noticeably slower.

On the good side, the new MacBook Pro LED with Nvidia graphics is much faster at Quartz graphics according to Xbench 1.3 leading to a significantly improved User Interface score in Xbench 1.3. It is also much faster at the 3D game Halo v2.0.1, consistent with other comparisons between Nvidia and ATI graphics that we have made. On the other hand, it does not do better at Doom 3 and Quake 4 at the resolutions and settings we use. At least it does not do worse.

But the bad side is very bad. The new MacBook Pro LED with the supposedly hot new Nvidia GeForce 8600M GT graphics does much poorer with OpenGL graphics, the foundation of many Mac games, than its predecessor with ATI Mobility Radeon X1600 graphics. It does not even come close to the expectations from previous benchmarking. This can be seen in the OpenGL benchmarks of Cinebench 9.6 (-10%) and Xbench 1.3 (-34%) and in the 3D games Quake II Arena (-35%) and, especially, Call of Duty 2 (-67%). The latter is extremely unfortunate for us because Call of Duty 2 is our favorite 3D game and it is totally unplayable in multiplayer mode at any settings with the new MacBook Pro LED. Call of Duty 2 plays great on our 24" iMac Core 2 Duo with Nvidia graphics so its poor performance on the new MacBook Pro is a mystery. It is possible that the Call of Duty 2 performance issue may be at least partially in the game itself. Either way, we hope that Apple and/or Aspyr fix this glitch quickly.

[UPDATED--6/20/07--Several readers and members of Apple's MacBook Pro Discussion Forums have confirmed the unplayability of Call of Duty 2. Unfortunately, Mac OS X 10.4.10 update released today does not fix the problem with OpenGL performance nor with Call of Duty 2.]

We expected the 200GB 4200rpm Toshiba hard disk drive to be slower than the predecessor's 160GB 54000rpm Hitachi hard disk drive and it is. It scores 9% lower in Xbench 1.3. It also contributes significantly to the 24% increase in startup time and to the 49% increase in the time to duplicate a folder with 20 files totaling 500MB. If one can get by with 160GB of disk space, the MacBook Pro will run significantly faster for many operations with the 160GB 5400rpm drive. But we need the extra space and do not do a lot of drive-sensive work. We also expect to get another couple of positives that are important to us, i.e. bit more battery life and a little less heat generated.

Heat

Speaking of heat, the utilities CoreDuoTemp, smcFanControl and Temperature Monitor all indicate that the new 2.4GHz MacBook Pro LED's "Santa Rosa" CPU runs about the same temperature as the "Merom" CPU of its predecessor 2.33GHz MacBook Pro Core 2 Duo on battery and on AC with no applications open and maybe a few degrees more depending on which applications are open--Safari and Dreamweaver C3 seem to raise the temperature the most. We thought the "Santa Rosa" CPU would run cooler based on Intel's announced specifications. Anyway, the fans in the new MacBook Pro LED run at a lower speed and are quieter at all speeds than in its predecessor.

Another curious observation is that the speed of the new "Santa Rosa" CPU has never exceeded 2.0GHz as measured by the CoreDuoTemp utility even though it is a 2.4GHz CPU. No matter how we tried to stress the CPU it has never registered above 2.0GHz. Hopefully, this is because CoreDuoTemp needs to be updated for the new CPU.

Battery Life

Our experience is that battery life is significantly improved in the new MacBook Pro LED. We have not conducted any rigorous timed tests as yet but at full useable conservation, we have been getting 3.5-4 hours use regularly on the new model vs 2.75-3.25 hours on its predecessor.

WiFi

Finally, the Intel 802.11n WiFi unit in the new "Santa Rosa" CPU works great. We've had no problems connecting with our Apple AirPort Extreme "n" Base Station or other WiFi transmitters in various AirPorts and cafe's.

In Summary

Overall, we are generally pleased with the new 15" 2.4GHz MacBook Pro Core 2 Duo LED. The only exception is its very poor OpenGL graphics performance. Given the benchmarks of the Nvidia graphics unit published elsewhere and noted above and Apple's advertising claims, the graphics performance is a huge disappointment. Hopefully, this simply stems from the Nvidia drivers in Mac OS X needing to be updated for the 8600M series chip and that they will be updated shortly.

We like the new LED screen and are pleased with the glossy option. The longer battery life is a real plus. We don't really notice the 0.2lb weight loss when handling the two MacBook Pros but we are happy about it anyway. And, we're glad to get almost an extra 40GB of disk space even if it is a bit slower as a result.

If Apple can fix the new MacBook Pro's OpenGL performance, it is a real winner. If not, there are still plusses that will appeal to many, although we would recommend they carefully consider the $1999 2.2GHz model along with the $2499 2.4GHz model. [Bill Fox]

Saw something? Send a tip

The archive ran on reader tips. What did you see, where, and do you want the credit?

Read by the editor. Never published without your say.

More in Reviews · This month in the archive