Unconfusing GeForce RTX 20 Mobile
This month is particularly interesting for gaming laptops enthusiasts. Both CPU and GPU makers have released new hardware for mobile gamers, including AMD Ryzen 4000H&U, Intel 10th Gen H&U-series and NVIDIA RTX SUPER series. Sadly, there is still no word from OEMs about RX 5700M and RX 5600M laptop launches.
NVIDIA has quite a few variants of its GeForce RTX models. Their recent launch has delivered not only the RTX SUPER models, but also refreshed SKUs of already existing mobile graphics cards.
Check this list for better understanding of RTX Mobile series:
RTX 20×0 Max-P
In January 2019 NVIDIA released its first RTX graphics card for the mobile market. The series includes three main SKU: 2080, 2070 and 2060. The cards are equipped with Turing TU104 and TU106 GPUs. These were the first mobile graphics cards to feature GDDR6 memory. Internally these cards are known as Max-P designs (aka max performance).
RTX 20×0 Max-Q
Additionally to the original RTX series, NVIDIA introduced Max-Q designs (aka max efficiency). These are power-limited versions of Max-P graphics, so in other words, they are slower.
The thing to note is that there are many different variants for each Max-Q model. NVIDIA allows OEMs to control the TGP from 90W to 65W (for RTX series). So just because your laptop has Max-Q graphics, it does not mean it will perform the same as other Max-Q designs. For this reason, pay attention to GPU specifications in laptop reviews.
RTX 20×0 SUPER
On April 2nd, 2020, NVIDIA introduced the SUPER RTX series. These cards feature more CUDA cores, low-voltage GDDR6 memory, and a new regulator. Additionally, these cards support Dynamic Boost (a clone to AMD SmartShift except it works with both AMD and Intel CPUs) and also Advanced Optimus (iGPU/dGPU switching technology with G-Sync support). The AO requires additional hardware and it will take some time to see laptops with this technology (Lenovo Legion 5i and 7i is currently the only series to officially support this version of Optimus).
RTX 20×0 Refresh
On ComputerBase forums, Schenker (XMG) representative revealed that NVIDIA has also released new refreshed designs of existing RTX 2070 and RTX 2060 models. Unlike the SUPER variants, these cards do not feature more cores, they are essentially the same cards (but slightly higher-clocked) as Max-P models.
The key difference is that they feature hardware optimizations from the SUPER series, which are low-voltage GDDR6 memory (1.25V vs 1.35V) and a new regulator. Since the memory can consume up to 20% of total TGP (total graphics power), such an upgrade provides a noticeable performance boost over the original series.
For the RTX 2070, the maximum TGP has been unchanged and remains at a value of 115W, but the TGP of RTX 2060 has been increased up to 115W (the original was only up to 90W).
According to Notebookcheck, RTX 2070 Mobile Refresh features TU106-735 GPU (aka N18E-G1R) and a boost frequency of 1485 MHz (+45 MHz). As for the 2060R (with N18E-G1-B GPU), the clock speeds remain unconfirmed.
NVIDIA GeForce RTX 20 Mobile Series | |||||||
---|---|---|---|---|---|---|---|
GeForce RTX | GPU | CUDAs | Boost Clock | Memory | TDP | 1.25V GDDR6 & New Regulator | Dynamic Boost & Advanced Optimus |
2080S | N18E-G3R | ? | ? | ||||
2080S Max-Q | N18E-G3R | ? | ? | ||||
2080 | N18E-G3 | ? | ? | ||||
2080 Max-Q | N18E-G3 | ? | ? | ||||
2070S | N18E-G2R | ? | ? | ||||
2070S Max-Q | N18E-G2R | ? | ? | ||||
2070 Refresh | N18E-G1R | ? | ❔ | ||||
2070 | N18E-G2 | ? | ? | ||||
2070 Max-Q | N18E-G2 | ? | ? | ||||
2060S * | N18E-G1R | ❔ | ❔ | ? | ❔ | ||
2060 Refresh | N18E-G1-B | ❔ | ? | ❔ | |||
2060 | N18E-G1 | ? | ? | ||||
2060 Max-Q** | N183E-G1 | ? | ? | ||||
* Not yet released; ** Never released |
Source: ComputerBase