Rumors are pouring in when it comes to graphics cards based on Nvidia’s latest architecture Turing, and for desktops, cards in the middle segment are on the wallpaper. In early December, it was reported that one of the big news during the CES 2019 trade fair could potentially be the Turing and Geforce RTX series for laptops.
Through the product page for a laptop, the Chinese manufacturer Cjscope now confirms that mobile Turing circuits are not far away. The models mentioned are Geforce RTX 2080, RTX 2070 and RTX 2060, where the suffix “MXM” is used to distinguish the variants for laptops.
In terms of specifications, the Geforce RTX 2080 model is very similar to the version for desktops, where 2,944 CUDA cores are accompanied by 8 GB of GDDR6 memory and a base frequency of 1,515 MHz. The turbo frequency of the mobile version is 1,847 MHz, which is about 140 MHz above the level that applies to the graphics circuit in the reference version.
The “MXM” version of the RTX 2070 also matches the variant for desktops in many ways, and they both have full-scale versions of the TU106 circuit with 2,304 CUDA cores under the hood. This is accompanied by 8 GB of GDDR6 memory and a base frequency of 1,410 MHz. The turbo frequency of the mobile graphics card is stated to be 1,740 MHz, which can be compared with 1,620 MHz for the reference model.
Geforce RTX 2060 has so far not been officially announced, but the circuit for the mobile variant differs somewhat from the expected specifications of the graphics card. The ditto laptop circuit TU106 has only 24 of 36 SM clusters activated, and therefore takes a step from 1,920 to 1,536 CUDA cores. However, the two variants have identical base and turbo frequencies of 1,320 and 1,620 MHz, respectively, as well as 6 GB of GDDR6.
For all variants, a memory frequency of 14,000 MHz applies, while the width of each memory bus is identical to the stationary counterparts, which also gives an identical theoretical memory bandwidth. With similar clock frequencies and circuits, similar power consumption is to be expected, but there is still no information about this.
The graphics cards listed will thus be in need of competent cooling solutions, and when versions for thin laptops according to Nvidia’s Max-Q Design are to be expected is unknown. It is worth mentioning that Cjscope writes that the laptop with Geforce RTX 2060 MXM will be available on 15 January, which may mean product launch during CES 2019, which runs from 8 to 12 January.
Specifications: Nvidia Geforce RTX for laptops
RTX 2080 | RTX 2080 MXM | RTX 2070 | RTX 2070 MXM | RTX 2060 | RTX 2060 MXM | |
---|---|---|---|---|---|---|
Technical | 12 nm TSMC | 12 nm TSMC | 12 nm TSMC | 12 nm TSMC | 12 nm TSMC | 12 nm TSMC |
Circuit | TU104 | TU104 | TU106 | TU106 | TU106 | TU106 |
CUDA cores | 2 944 st. | 2 944 st. | 2 304 st. | 2 304 st. | 1 920 st. | 1 536 st. |
Clock frequency | 1 515 MHz | 1 515 MHz | 1 410 MHz | 1 410 MHz | 1 320 MHz | 1 320 MHz |
GPU Boost | 1 710 MHz | 1 847 MHz | 1 620 MHz | 1 740 MHz | 1 620 MHz | 1 620 MHz |
Memory bus | 256-bit | 256-bit | 256-bit | 256-bit | 192-bit | 192-bit |
Memory amount | 8 GB GDDR6 | 8 GB GDDR6 | 8 GB GDDR6 | 8 GB GDDR6 | 6 GB GDDR6 | 6 GB GDDR6 |
Memory frequency | 14 000 MHz | 14 000 MHz | 14 000 MHz | 14 000 MHz | 14 000 MHz | 14 000 MHz |
Memory bandwidth | 448 GB/s | 448 GB/s | 448 GB/s | 448 GB/s | 336 GB/s | 336 GB/s |
TDP | 215 W | ? | 175 W | ? | <150 W(?) | ? |
* Values apply to Nvidia’s Founders Edition of each model, which comes with 90 MHz overclocking. In addition to slightly higher performance, this in all cases leads to an increase in TDP of 10 W.
Source: Cjscope via Wccftech