Exactly one month ago, we noted the presence of a "MacBookPro10,2" entry in the Geekbench results database, suggesting that a 13-inch version of the Retina MacBook Pro is indeed in the works, as has been rumored for October. In comparison, the 15-inch Retina MacBook Pro carries a "MacBookPro10,1" identifier, while current non-Retina versions of the MacBook Pro carry "MacBookPro9,x" identifiers.
While much of the data presented to Geekbench by machines can be faked and many Hackintosh enthusiasts go to great lengths to match their machines' details to those of actual Macs, the details of the MacBookPro10,2 were consistent enough with expectations that both we and Geekbench developer Primate Labs believed the entry to be genuine.
Following the original entry, a second MacBookPro10,2 has now appeared in the results database, with the data bearing very strong similarity to the original entry but with some differences that suggest this entry may also be legitimate.
Like the machine that appeared first, the new machine runs a 2.9 GHz Core i7-3520M, the same processor found in the current high-end model of the non-Retina 13-inch MacBook Pro. The machine's motherboard identifier also matches the previous machine, which itself matched a previously unaccounted for identifier seen in early developer builds of OS X Mountain Lion. And as would be expected for machines running the same specs, the two machines have very similar Geekbench scores of roughly 7800.
There are a few differences, however, with one of the key ones being a custom 12A2066 build of the unreleased OS X 10.8.1. Just two days ago, Apple gave the first indication that it will begin seeding OS X 10.8.1 to developers, but it has yet to do so. In addition, the four-digit build number suffix is sometimes used internally by Apple on its test machines, although such four-digit build numbers are sometimes also used publicly for special builds such as the version of OS X 10.7.4 that shipped on new 15-inch Retina MacBook Pro models.
In addition, this new entry reflects a MacBook Pro with 8 GB of RAM, in line with what ships paired with the Core i7-3520M on the corresponding non-Retina model. The previous MacBookPro10,2 entry showed a machine with only 4 GB of RAM. Finally, the BIOS string for the new entry appears to show an updated date code of July 18, 2012, compared to the previous one registering as June 11, 2012.
Geekbench results have on several occasions revealed upcoming Mac models, most recently in mid-May with the MacBookPro9,1 that turned out to be the non-Retina 15-inch MacBook Pro upon its release the following month. Notably, an iMac13,2 entry that also appeared to be legitimate surfaced around the same time, but Apple has yet to release updated iMacs.
(Thanks, Matthew!)
Top Rated Comments
Windows 8 is aweful - think its about time I finally switch
I'll have to ask u guys where the best place to sit in Starbucks so my new mbp gets the most attention though
http://en.wikipedia.org/wiki/Comparison_of_Nvidia_graphics_processing_units#Pre-GeForce
Yep, that's before the GeForce series. That's a 90s GPU, before we even called them GPUs (the popular term back then was 3D accelerator).
The problem isn't the actual pixel count, that's nothing once the framebuffer is built, it's just one big copy operation. The problem is probably that Apple does multiple pass rendering of its UI with too many memory read/write operations instead of maximising buffer passes. That's the big problem with modern GPUs, they are plenty fast to push the pixels, it's feeding them the information to do it with (all the layers to blend together, the compositing effect) that slows them down.
The Intel HD graphics is probably just as able as the GT650M (the reason it kicks in on Facebook and some other sites is the use of HTML 5 Canvas/Flash and other graphic "intensive" operations. Webkit enables the dedicated GPU for that), Apple just needs to optimize the pipeline better. Why you're seeing better performance out of the dedicated GPU is basically just brute forcing an inefficient rendering pipeline through, and has nothing to do with fill rate.
Also, when getting to 5% battery, it's not just the dedicated GPU that gets turned off, your CPU also probably gets throttled through Intel Speedstep technology.
My '08 uMB 13" (the '08 13 MBP! @haushinka : haha, kidding!) was able to drive both its internal 1280x800 and an external 30" ACD at 2560x1600. That's very close to the same pixel count : 307,200,000 pixels with a 60 hz refresh. It did so on a 9400m nVidia IGP.
Obviously not the same as the 15" rMBP (2880x1800)..
:rolleyes:
Obvious MacRumors bot is obvious.