NVIDIA is launching three mobile versions of its GeForce RTXs: the RTX 3060, RTX 3070 and RTX 3080. Even the less powerful of the three is faster (on paper) than the current top of the range!
Rumors about the arrival on the market of mobile versions of the RTX 3000 Series were becoming more and more pressing lately, NVIDIA has finally put an end to them: the manufacturer has just announced three new GPUs intended for laptops, the GeForce RTX 3060 Mobile, RTX 3070 Mobile et RTX 3080 Mobile. As a bonus, NVIDIA takes the opportunity to unveil the GeForce RTX 3060 « desktop », little sister of the current RTX 3060 Ti.
Three new mobile GeForce RTX GPUs
With its three new mobile GPUs, NVIDIA brings the Ampere architecture to laptops. We therefore find all the technologies already present in desktop chips: second generation RT cores, third generation Tensor Cores, as well as DirectX 12 Ultimate, Vulkan RT and OpenGL 4.6 compatibility. DLSS is of course in the game, as is HDMI 2.1, all on a PCI-Express Gen 4 bus. These GPUs also include a seventh generation video encoder and a fifth generation decoder.
On the TDP side, the RTX 3060 Mobile consumes between 60 and 115 watts (depending on the settings defined by the manufacturer of the platform), while the TDP of the RTX 3070 Mobile is between 80 and 125 watts. More powerful, the RTX 3080 Mobile displays a TDP of 80 to 150 watts, or even more. Interesting detail regarding GDDR6 video memory: the RTX 3080 Mobile can be associated with 8 GB or 16 GB of memory on a 256-bit bus, while the RTX 3070 Mobile and RTX 3060 Mobile can only be associated with 8 GB respectively ( on a 256-bit bus) and 6 GB (on a 192-bit bus).
|Model||GeForce RTX 3080 Mobile||GeForce RTX 3070 Mobile||GeForce RTX 3060 Mobile|
|Boosted frequency (MHz)||1245 – 1710 MHz||1290 – 1620 MHz||1283 – 1703 MHz|
|GPU Subsystem Consumption (Watts)||80 – 150+ W||80 – 125 W||60 – 115 W|
|Standard memory configuration||16 Go GDDR6
8 Go GDDR6
|8 Go GDDR6||6 Go GDDR6|
|Memory interface size||256-bit||256-bit||192-bit|
On the performance side, NVIDIA says the RTX 3060 Mobile is faster than any (laptop) model currently on the market, and performs 30% better than a PS5. This GPU is designed for Full HD, with a framerate 90 FPS target (in Ultra). The RTX 3070 Mobile is more suitable for 1440p, still with a framerate 90 FPS target in Ultra. It is advertised as 50% more efficient than an RTX 2070 Mobile. Finally, the RTX 3080 Mobile becomes the fastest model on the market, capable of running games at 1440p Ultra at over 100 FPS.
The first laptop models equipped with these new GPUs are expected to hit the market from January 26, with prices starting at $ 999 for models with RTX 3060 Mobile, $ 1299 for those equipped with an RTX 3070 Mobile. and $ 1999 for high-end models with an RTX 3080 Mobile.
NVIDIA takes the opportunity to announce its third generation Max-Q platform, with some new features and improvements. Dynamic Boost 2.0 is therefore part of the game, with the possibility of balancing consumption between the GPU, the CPU but also the graphics memory, to optimize the output and therefore the performance of the entire machine. The WhisperMode goes to version 2.0, and we note the appearance of the Resizable BAR, the equivalent of AMD’s SAM (Smart Access Memory): this technology makes it possible to present to the CPU all of the graphics memory, accessible at once , improving (as appropriate) performance.
And a new GeForce RTX 3060
The manufacturer finally unveils a “small” GeForce RTX 3060, a lighter and modified version of the RTX 3060 Ti. The number of Cuda Cores goes from 4864 to 3584, with a slightly lowered base frequency (1.32 GHz, against 1.41 GHz for its big sister). In return, the Boost frequency climbs from 1.67 GHz to 1.78 GHz.
The memory bus is also revised downwards, going from 256-bit on the Ti version to “only” 192-bit. Surprisingly, we find 12 GB of GDDR6 memory on this RTX 3060 where the 3060 Ti has only 8 GB. The HDMI 2.1 and DisplayPort 1.4a outputs are finally still in the game. This new graphics card should make its appearance at the end of February. We will not advance on its price in store, the prices of new NVIDIA cards still have not stabilized at the time of this writing …