Review Nvidia GeForce GTX 860M Maxwell vs. Kepler
Florian Glaser, 04/14/2014
Gaming Laptop Windows
The unequal duel. Nvidia offers two different versions of its latest high-end GPU GeForce GTX 860M. Besides a version based on the Maxwell architecture, there is also a GPU from the old Kepler generation. We compared the performance of the two DirectX 11 chips and found a winner.
...
Specifications
It is easy to understand why Nvidia produces two versions of the
GeForce GTX 860M. Smaller notebooks use the soldered GPUs with the Maxwell chip GM107, whereas bulkier gaming cases with MXM slots use the significantly bigger Kepler sibling (GK104 chip), which is also manufactured in a 28 nm process.
Other similarities of the two cards include the memory interface: The GTX 860M uses modern GDDR5 video memory, but a 128-bit interface is not up to date anymore in 2014. It does not really matter if Nvidia equips the graphics card with 2 or 4 GB memory based on our experiences.
Much more important is the number of shader units. Thanks to the huge efficiency improvement, 640 CUDA cores are sufficient for the Maxwell GTX 860M to compete with the in-house rivals. Compared to the
GTX 765M, the Kepler GTX 860M gets 50% more CUDA cores, more precisely 1,152 instead of 768. This is even superior to the
GTX 770M (960 CUDA cores).
The Maxwell chip is particularly superior in terms of clocks. 1,029-1,097 MHz during 3D operation is significantly higher than the Kepler version. The latter manages 797-915 MHz under load, which is still decent.
Graphics card | GeForce GTX 765M | GeForce GTX 860M Kepler | GeForce GTX 860M Maxwell | GeForce GTX 770M | GeForce GTX 870M | GeForce GTX 780M | GeForce GTX 880M |
---|
Manufacturing process | 28 nm | 28 nm | 28 nm | 28 nm | 28 nm | 28 nm | 28 nm |
Chip | GK106 | GK104 | GM107 | GK106 | GK104 | GK104 | GK104 |
Architecture | Kepler | Kepler | Maxwell | Kepler | Kepler | Kepler | Kepler |
CUDA Cores | 768 | 1,152 | 640 | 960 | 1,344 | 1,536 | 1,536 |
Core clock base | 850 MHz | 797 MHz | 1,029 MHz | 811 MHz | 941 MHz | 823 MHz | 954 MHz |
Core clock Turbo | 902 MHz | 915 MHz | 1,097 MHz | 862 MHz | 967 MHz | 849 MHz | 993 MHz |
Memory clock | 2,000 MHz | 2,500 MHz | 2,500 MHz | 2,000 MHz | 2,500 MHz | 2,500 MHz | 2,500 MHz |
Memory | 2,048 MB | 4,096 MB | 2,048 MB | 3,072 MB | 6,144 MB | 4,096 MB | 8,192 MB |
Memory type | GDDR5 | GDDR5 | GDDR5 | GDDR5 | GDDR5 | GDDR5 | GDDR5 |
Memory interface | 128-bit | 128-bit | 128-bit | 192-bit | 192-bit | 256-bit | 256-bit |
...
Verdict
GTX 860M Maxwell (left) vs. GTX 860M Kepler (right)
The Kepler GTX 860M does actually leave a good impression when you compare it with the last generation. Around 30% more performance than a
GTX 765M certainly deserve credit.
However, the old chip has a hard time keeping up with the Maxwell version. Despite less shaders, the new architecture is around 10% faster. The
GeForce GTX 770M is also slightly ahead of the Kepler chip, which is not least caused by the larger memory interface. We are pretty sure that the Kepler GTX 860M would be faster with a 192-bit memory interface.
The Kepler GPU offers a decent, but not brilliant high-end performance. If possible, gamers should choose the Maxwell chip or a more powerful GPU. The
GTX 860M is usually sufficient for 1600x900 pixels, high details and – depending on the game – subtle anti-aliasing settings.
GPU | Overall Performance |
---|
GeForce GTX 880M | +65% |
GeForce GTX 780M | +47% |
GeForce GTX 870M | +37% |
GeForce GTX 860M Maxwell | +10% |
GeForce GTX 770M | +3% |
GeForce GTX 860M Kepler | Base (Average of 16 games) |
GeForce GTX 765M | -23% |