Today the presentation of the new NVIDIA graphics cards has taken place, but a detail that the company has not had the diligence to clarify is that the two versions of the RTX 4080 presented do not have the same performancesuggesting that the 12GB RTX 4080 really should have been an RTX 4070.
This case, as we indicated in the main news, is reminiscent of the NVIDIA GeForce RTX 1060, where the 3GB models had significantly lower performance than the 6GB modeland in this case we are faced with the same thing.
|
RTX 4090
|
RTX 4080 16GB
|
RTX 4080 12GB
|
---|---|---|---|
CUDA kernels
|
16384
|
9728
|
7680
|
base frequency
|
2230MHz
|
2210MHz
|
2310MHz
|
boost frequency
|
2520MHz
|
2510MHz
|
2610MHz
|
Memory
|
24GB GDDR6X
|
16GB GDDR6X
|
12GB GDDR6X
|
memory bandwidth
|
384 bit
|
256 bit
|
192 bit
|
Consumption
|
450W
|
320W
|
285W
|
Although the operating frequencies are favorable to the 12GB model, it is worth seeing how the 16GB model has 9728 cores, compared to the 7680 cores that the 12GB model has, in the case of a 21% reduction in the number of coreswhich we could extrapolate to the rest of the components of the chip in the absence of a clear specification about them.
Similarly, this reduction in performance will not be 21 percent either, since the operating frequencies are slightly higher, but, in any case, we find two graphics cards with the same name and appearance that will perform quite differently from each otherso buyers should be careful.
End of Article. Tell us something in the Comments!
Jordi Bercial
Avid technology and electronics enthusiast. I’ve been messing around with computer components almost since I could walk. I started working at Geeknetic after winning a contest on their forum about writing hardware articles. Drift, mechanics and photography lover. Feel free to leave a comment on my articles if you have any questions.
Add Comment