GeForce GT 650M 

Released March, 2012
  • 850 MHz
  • 1 GB GDDR5
6.4 Out of 10

Although the reference GeForce GTX 650 has two dual-link DVI ports and one mini-HDMI output, Gigabyte adds an analog VGA connection.
by Don-Woligroski (Sep, 2012)
Although Zotac employs more conservative clock rates than Gigabyte, its 993 MHz core and 1059 MHz average GPU Boost speeds still represent 13 and 26 MHz improvements over Nvidia's specification.
by Don-Woligroski (Sep, 2012)

GPUBoss Review Our evaluation of the GeForce GT 650M among Laptop GPUs over 35W TDP


GeForce GT 650M
Battlefield 3, Battlefield 4, Bioshock Infinite and 21 more


GeForce GT 650M
T-Rex, Manhattan, Sky Diver Factor and Fire Strike Factor


GeForce GT 650M
Face Detection, Ocean Surface Simulation and 3 more

Performance per Watt

GeForce GT 650M
Battlefield 3, Battlefield 4, Bioshock Infinite and 31 more


GeForce GT 650M
Battlefield 3, Battlefield 4, Bioshock Infinite and 31 more

Noise and Power

GeForce GT 650M
TDP, Idle Power Consumption, Load Power Consumption and 2 more


Overall Score

GeForce GT 650M

Benchmarks Real world tests of the GeForce GT 650M

Bitcoin mining Data courtesy CompuBench

GeForce GT 650M
23.2 mHash/s
GeForce GT 740M
44.05 mHash/s
GeForce GT 750M
27.37 mHash/s

Face detection Data courtesy CompuBench

GeForce GT 650M
9.82 mPixels/s
GeForce GT 750M
11.72 mPixels/s
GeForce GT 740M
9.33 mPixels/s

T-Rex (GFXBench 3.0) Data courtesy CompuBench

GeForce GT 650M
GeForce GT 750M
GeForce GT 740M

Manhattan (GFXBench 3.0) Data courtesy CompuBench

GeForce GT 650M
GeForce GT 750M
GeForce GT 740M

Fire Strike Factor Data courtesy FutureMark

GeForce GT 650M

Sky Diver Factor Data courtesy FutureMark

GeForce GT 650M

Bioshock Infinite

GeForce GT 650M

Crysis 3

GeForce GT 650M

In The News From around the web


NVIDIA Quadro M6000 Announced

by Scott Michaud |
I wonder what this means for the Tesla line, and whether we will see a variant of Maxwell with a large boost in 64-bit performance, or if that line will be in an awkward place until Pascal.
The memory, but not the cache, is capable of ECC (error-correction) for enterprises who do not want a stray photon to mess up their computation.

NVIDIA Announces Quadro M6000 & Quadro VCA (2015)

by Ryan Smith |
However the company is also pushing the use of Physically Based Rendering (PBR), a compute-intensive rendering solution that uses far more accurate rendering algorithms to accurately model the physical characteristic of a material, in essence properly capturing how light will interact with that material and reflect off of it rather than using a rough approximation.
However as FP64 compute is not widely used in graphics, this is not something NVIDIA believes will be an issue.
Sorry folks but unless I am wrong on this one, you wont be seeing Maxwell in 800M Series GPUs.
We also know that the epitome of the Maxwell architecture is power efficiency and Performance Per Watt.
Nvidia would also most probably release a 880M SLI Configuration which would be the king of mobile graphics for a year or so.
Details about Maxwell’s Mobility series surfaced a few days back and now predicted benchmarks have also surfaced.

Specifications Full list of technical specs


GPU brand Nvidia
GPU name GK107
Market Laptop
Clock speed 850 MHz
Is dual GPU No
Reference card None

raw performance

Shading units 384
Texture mapping units 32
Render output processors 16
Pixel rate 6.8 GPixel/s
Texture rate 27.2 GTexel/s
Floating-point performance 652.8 GFLOPS


Memory clock speed 900 MHz
Effective memory clock speed 1,800 MHz
Memory bus 128 bit
Memory 1,024 MB
Memory type GDDR5
Memory bandwidth 28.8 GB/s

noise and power

Report a correction


comments powered by Disqus