If you’re waiting for real-time ray tracing to show up on gaming laptops, the wait might soon be over.
Nvidia’s latest RTX-series GPUs, codenamed Turing, are of course already available in desktop form factors in the shapes of the high-end RTX 2080 and RTX 2070.
However, the mid-range and mobile segments are still missing, although we’ve heard rumors that this could change soon.
Now some additional – and even more interesting – details regarding the laptop Turing chips have surfaced via Chinese manufacturer CJSCOPE, and the leak was picked up by wccftech.com.
The Turing chips mentioned are the Geforce RTX 2080, RTX 2070 and RTX 2060, where the suffix MXM is used to distinguish the variants for laptops. MXM is short for Mobile PCI Express Module – a socket available on certain gaming laptop motherboards, allowing for easy installation of a GPU that supports the standard.
Google’s cache of the page in question was still available at the time of writing and included tidbits such as this (auto translated):
Interestingly, this even includes a shipping date for the RTX 2016, which is apparently ‘tentatively scheduled’ for January 15, 2019.
GeForce RTX: Desktop VS Mobile
So if the listed specs are correct, a comparison to the desktop lineup would look like this:
In terms of specifications, the Geforce RTX 2080 model looks very similar to the desktop design. 2,944 CUDA cores are accompanied by 8 GB of GDDR6 memory and the base frequency is 1,515 MHz. The turbo frequency for the mobile version is at 1,847 MHz, which is about 140 MHz higher than the reference version.
The mobile version of the RTX 2070 also mostly matches the desktop GPU. Both use full versions of the TU106 chip with 2,304 CUDA cores, in combination with 8 GB of GDDR6 memory. Its base frequency is 1,410 MHz, but again, the turbo frequency is higher in the mobile GPUs at 1,740 MHz (reference model: 1,620 MHz)
While the GeForce RTX 2060 has not even been officially announced for desktops, the mobile variant differs somewhat from the 2060’s expected specifications. The laptop chip appears to have only 24 out of 36 SM clusters enabled, and consequently 1,536 instead of 1,920 CUDA cores. Base and turbo frequencies are however identical, at 1,320 and 1,620 MHz, respectively. Both are also paired with 6 GB of GDDR6.
For all variants, an effective memory frequency of 14 GHz applies. Each memory bus is identical to its desktop counterpart, which also results in an identical theoretical memory bandwidth.
All else being equal, the estimated power consumption should also be the same, although there is no information available about this. But if so, the RTX 2080 and 2070 in particular will require very competent cooling solutions when they arrive in laptops, even in putative Max-Q incarnations.
The most likely time for an official launch of the new mobile GeForce RTX lineup is still CES 2019, which runs from January 8 to 12.