Nvidia mobile RTX 3080 Max-P versus Max-Q: specs and estimated performance
Nvidia will officially unveil the RTX 3000 laptop GPUs on January 12, and just like the last few generations, the mobile SKUs are expected to get segmented into Max-P and Max-Q variants. It goes without saying that the mobile GPUs will not be as fast as the desktop versions, and, according to some new spec and performance data published by TechPowerUP, it looks like the laptop SKUs are quite far behind performance-wise, at least for the RTX 3080 models.
According to the TechPowerUP database entries, both Max-P and Max-Q versions of the RTX 3080 mobile GPU will feature the same GA104 SKU with 6144 CUDA cores, 192 TMUs, 96 ROPS and 8 GB or 16 GB of GDDR6 VRAM with 256-bit bus. However, the two versions will feature different TGPs, core frequencies and possibly different GDDR6 speeds. As such, the core clocks for the Max-P variant appear to be set at 1110 MHz (base) / 1545 MHz (boost), while the Max-Q version gets 780 MHz (base) / 1245 MHz (boost) frequencies. The Max-P version might also get 14 Gbps GDDR6 RAM, but, for now, both versions appear to pack 12 Gbps GDDR6 VRAM clocked at 1500 MHz. As far as TGPs go, the Max-P gets a 115 W default rating, and the Max-Q is rated for 80 W by default, but Videocardz is suggesting that the situation could get even more confusing, as laptop OEMs could choose to raise or lower the TGPs for each version.
TechPowerUP also provides expected performance charts for each version, and it looks like the RTX 3080 Max-P would be 3% faster than the RTX 3060 Ti desktop model, but 40% slower than the RTX 3080 desktop GPU. On the other hand, the Max-Q is even slower, as it nearly matches the GTX 1080 Ti desktop, yet it is 74% slower than the RTX 3080 desktop GPU. Now, we know there will be RTX 3070 and 3060 mobile versions, as well, and we reckon the performance difference between these and the desktop models is not going to be as big.