After announcing the show-stealing RTX 3050 at their CES 2022 conference, NVIDIA proceeded to finally lift the curtains off two new RTX 30-series mobile GPUs. Those being the RTX 3070 Ti and the RTX 3080 Ti. Both of these represent the high-end segment and will come paired with best-in-class components across many laptops from different manufacturers. NVIDIA didn’t give us much info but we do have a lot of prior leaked info to help us out, so let’s break these down.
Both of these Ti GPUs come in standard and Max-Q versions so thin notebooks and power-hungry laptops are covered altogether. NVIDIA even announced the 4th Gen Max-Q Technologies alongside these showing off just what these GPUs could do even in a small form factor. We’ll talk about the Max-Q aspect at the end as it’s quite interesting what the company has developed.
RTX 3070 Ti Mobile
Starting off with the more conservative of the two options, the RTX 3070 Ti Mobility is a high-end mobile GPU that is 1.7x (or 70%) faster than the RTX 2070 Super, according to NVIDIA. It features 8GB of GDDR6 memory clocked in at 14Gbps running across a 256-bit interface, giving it a memory bandwidth of 448Gb/s. Moreover, it will feature 5888 CUDA Cores along with 46 SM units, and it’s based on a cut-down GA-104 die.
The company did not disclose clock speeds or thermal limits at all, but we can guess it will be probably around 150W considering the performance numbers NVIDIA gave us. Speaking of which, the GPU can push 100+ frames at 1440p resolution with Ultra settings on most games, including AAA titles. That’s a lot of horsepower for a mobile GPU. Laptops with the RTX 3070 Ti will be available starting February 1st, 2022 and the cheapest one is going to run you $1499.
RTX 3080 Ti Mobile
NVIDIA fortunately told us a little bit more about this one, perhaps because it’s the most powerful mobile GPU ever. Really, NVIDIA says this mobile GPU is faster than the desktop Titan RTX and the new benchmark for flagship mobile graphics. It achieves 120+ FPS at 1440p with Ultra settings on almost all games. Moreover, we see 16GB of GDDR6 memory clocked in at 16Gbps on a 256-bit bus, that gives this GPU a memory bandwidth of 512Gb/s, the fastest in the notebook segment, ever.
Accordingly to leaks, the 3080 Ti mobile will have 7,424 CUDA Cores and 58 Compute Units, using the GA-103 die. The clock speeds were also not disclosed by NVIDIA but reports suggest a 1125Mhz base clock and 1590Mhz boost clock. With that kind of power, this is one hungry GPU with a TGP of 175W, 10W more than the RTX 3080. This number can be further tuned by manufacturers and NVIDIA requires laptop-makers to be clear about the power limits of its GPUs.
RTX 3080 Ti laptops will begin at $2499 and will be available alongside the first 3070 Ti laptops on February 1st, 2022. NVIDIA has said that the 3080 Ti (and 3070 Ti) is 10-20% faster than its non-Ti counterpart, it all depends on the manufacturer’s input changing the power limits. Hence, the requirement to be transparent to customers.
4th Gen Max-Q Technologies
NVIDIA’s Max-Q is making a return this year with its new mobile GPUs and this time it’s powered entirely by artificial intelligence. In the 4th iteration of Max-Q, the company is introducing three new features that will work to drastically improve performance while using less power. In other words, it will make your laptops as efficient as possible by balancing power between multiple components and provide respectable battery life.
CPU Optimizer is the name of the first technology and it’s an exciting feature. This allows your NVIDIA GPU to control your GPU by putting the power management into the GPU’s hands. Almost all the times, the CPU needs significantly less power than the GPU while gaming so this feature helps the GPU optimize power and transfer it over to the GPU so it can provide better performance. This feature only works with the latest Alder Lake mobile and Ryzen 6000 “Rembrandt” chips as NVIDIA has developed a special low-level infrastructure for these.
Then there’s Rapid Core Scaling which helps the GPU sense real-time demands of compute-heavy applications like Blender and put load on only the cores needed for the task at hand. This frees up other cores that don’t really need to be used and that power is distributed to the cores that are needed, boosting them further and significantly improving performance, up to 3x according to NVIDIA.
Lastly, we have Battery Boost 2.0 which, as the name suggests, aims to boost your laptop’s battery life. With Battery Boost 2.0 on, AI controls your whole laptop, distributing power between the CPU and GPU to manage it as efficiently as possible. It tries to find the best balance of image quality, frame rate, temperatures and power usage to provide optimal [performance. This all happens in real-time, improving the battery life by up to 70%.
You’ll see these two GPUs in over 160 laptops this year from various manufacturers and we’ve already covered one here. NVIDIA also announced four new desktop GPUs, from a new flagship all the way down to a new entry-level option. Apart from the green team, AMD also announced a plethora of new Radeon mobile GPUs and Intel launched its new Alder Lake lineup of mobile CPUs, both of which we’ll be covering in detail as well.