Monday, April 6th 2020
NVIDIA GeForce RTX 2060 SUPER Mobile Specs Leaked
NVIDIA last week debuted its GeForce RTX 20-series SUPER Mobile GPU family with four SKUs: the RTX 2080 SUPER (mobile), RTX 2070 SUPER (mobile), and their Max-Q variants. An ASUS ROG marketing slide just revealed the third, unannounced SKU from the series, the RTX 2060 SUPER (mobile), courtesy of a Hardware Unboxed news bulletin. Likely based on the "TU106" silicon, this SKU has an identical core-configuration to its desktop counterpart, with 2,176 "Turing" CUDA cores, 136 TMUs, 64 ROPs, 272 tensor cores, 34 RT cores, and a 256-bit wide GDDR6 memory interface, holding 8 GB of memory.
Where the RTX 2060 SUPER (mobile) differs from the desktop part is clock speeds and aggressive power management. It's clocked at 1305 MHz GPU base, 1485 MHz GPU Boost (1470/1650 MHz for the desktop SKU). The memory clock is untouched at 14 Gbps (GDDR6-effective). Although the slide doesn't say so, it's highly likely that there is a Max-Q variant of this SKU with tighter clock speeds and significantly lower power targets. Find the Hardware Unboxed presentation in the link below.
Sources:
Hardware Unboxed (YouTube), Tom's Hardware
Where the RTX 2060 SUPER (mobile) differs from the desktop part is clock speeds and aggressive power management. It's clocked at 1305 MHz GPU base, 1485 MHz GPU Boost (1470/1650 MHz for the desktop SKU). The memory clock is untouched at 14 Gbps (GDDR6-effective). Although the slide doesn't say so, it's highly likely that there is a Max-Q variant of this SKU with tighter clock speeds and significantly lower power targets. Find the Hardware Unboxed presentation in the link below.
2 Comments on NVIDIA GeForce RTX 2060 SUPER Mobile Specs Leaked
Based on the clockspeeds there would be <10% speed difference between the RTX 2060M Super and RTX 2070M Super? :confused:
We Need Max-Q Plus, with even more carefully fine-tuned undervolts, not Super editions that run with higher clocks, damnit!!