Technical Data / Specifications
With the Gigabyte GeForce GTX 780 OC V2 you get a factory overclocked card which features
a custom PCB as well as a custom cooler. The ASIC quality
measured on our sample was 77.4 % which is a bit higher than the average.
About ASIC quality:
Gigabyte's GeForce GTX 780 OC V2 is based on NVIDIA's massive Kepler GK110
chip, which is manufactured by TSMC using the latest 28 nanometer process
technology. With the GeForce GTX 780 you get twelve SMX units which results in
2'304 CUDA cores. Following the specs further, there are 192 TMUs, 48 ROPs, a
384 bit wide memory interface and 3 GB of GDDR5 memory.
Since Gigabyte didn't bother overclock the memory, the latter stays clocked
at 1'502 MHz (effective 6'008 MHz). In the end there is a memory bandwidth of
288 Gbps.
On the GPU you find quite a decent factory overclock with a base clock of 954 MHz and a
typical boost clock of 1'006 MHz. During our tests the highest boost clock we
measured was 1'097 MHz and the average boost clock was also 1'097 MHz.
The card also supports GPU Boost 2.0. The first version of GPU Boost stopped overclocking the GPU when a certain power target was hit. This new second version of GPU Boost stops overclocking the card, when a certain temperature is being reached. This makes sense since the temperature is the bigger inhibitor than the power target in most cases. In this case the temperature target is 80 degrees Celsius. In other words, as long as the GPU
runs at less than 80°C it will keep overclocking until the maximum frequency has
been reached. This is why in our case the average boost clock was 91 MHz higher than the typical boost clock
the manufacturer advertises.
nVidia GeForce |
GeForce GTX Titan |
Gigabyte GTX 780 OC
V2 |
ASUS GTX 780 DCU 2 OC |
GeForce GTX 780 |
Chip |
GK110 |
GK110 |
GK110 |
GK110 |
Process |
28 nm |
28 nm |
28 nm |
28 nm |
Transistors |
7.10 billion |
7.10 billion |
7.10 billion |
7.10 billion |
GPU
clock |
837 MHz |
954 MHz |
889 MHz |
863 MHz |
GPU Boost
clock |
876 MHz |
1'006 MHz |
941 MHz |
900 MHz |
Memory GDDR5 |
6'144 MB |
3'072 MB |
3'072 MB |
3'072 MB |
Memory
clock |
1'502 (6'008) MHz |
1'502 (6'008) MHz |
1'502 (6'008) MHz |
1'502 (6'008) MHz |
Memory
interface |
384 Bit |
384 Bit |
384 Bit |
384 Bit |
Memory
bandwidth |
288'400 MB/s |
288'400 MB/s |
288'400 MB/s |
288'400 MB/s |
TMUs |
224 |
192 |
192 |
192 |
Shader Cores |
2'688 (14 SMX) |
2'304 (12 SMX) |
2'304 (12 SMX) |
2'304 (12 SMX) |
ROPs |
48 |
48 |
48 |
48 |
TDP |
250 Watt |
>250 Watt |
>250 Watt |
250 Watt |
PCB Type |
Reference Design |
Custom Design |
Custom Design |
Reference Design |
Lenght (PCB - Total) |
27.0 - 27.0 cm |
26.9 - 28.8 cm |
26.7 - 28.7 cm |
27.0 - 27.0 cm |
Height (PCB - Total) |
11.0 - 11.0 cm |
11.0 - 12.6 cm |
12.0 - 14.2 cm |
11.0 - 11.0 cm |
Slots |
2 |
2 |
2 |
2 |
Cooler |
NVIDIA Reference |
WindForce 3X |
ASUS DirectCU II |
NVIDIA Reference |
MSRP |
$999 |
$XXX |
$XXX |
$649 |
Gigabyte decided to equip their GeForce GTX 780 OC V2 with their very
own cooling solution answering to the name WindForce 3X 450W. This one features
two 8mm and four 6mm copper heatpipes. Soldered to the heatpipes you find the fin stack
which is being fed with fresh air via three 80mm fans. They carry the model
number T128010SU and have been inclined in order to blow the air away better at
the top of the card.
Overall the cooler is well manufactured but the copper base that is far from the
mirror finish, since it looks industrial. Around the base you find an aluminium part
which is used to fix the cooler to the PCB and also to cool the memory thanks to
thermal pads.
The rear fin stack has been equipped with an aluminium plate to cool the
MOSFETs of the GPU power design. A thermal pad has also been used in this
case.
A closer look at the PCB shows that Gigabyte equipped its card with a
8+2 phase power design. The GPU gets its current from eight phases and the two
phases left take good care of the memory.
Checking the
voltage regulation chip we find a digital 8-phase synchronous buck converter
NCP4208
from ON (ON Semiconductor) that support I2C. Furthermore there is an
unidentified Richtek 2-phase analog PWM labelled 0T=FA V2A taking care of a stable
current supply for the the
memory.
On the PCB we find another chip labelled INA 3221. The INA3221 from Texas
Instruments is a three-channel, high-side current sensor and bus voltage monitor
with an I2C interface.
The memory chips used are made by Samsung and carry the model number
K4G20325FD-FC03. They are specified to run at 1'500 MHz (6'000 MHz effective).