While reviews for Nvidia’s RTX 2060 and RTX 2070 SUPER graphics cards have been public for three weeks now, we’ve had to wait to get a look at the RTX 2080 SUPER. The wait is over, however, and today we can present our full review of this new card. Built on the same TU104 GPU as the original 2080, the SUPER model ships with two extra SMs, meaning it has an extra 128 CUDA cores. On top of that, Nvidia is using all-new 15.5Gbps GDDR6 memory with this card, increasing total memory bandwidth to almost 500 GB/s.
Watch video via Vimeo (below) or over on YouTube at 2160p HERE
The Nvidia RTX 2080 SUPER isn’t just a spec bump over the original RTX 2080, though, as the Founders Edition card has also received an £80 price cut, from £749 to £669. This is due to the fact that, with the original 2080, the Founders Edition was a factory overclocked model priced above MSRP, but with 2080 SUPER the Founders card is a reference model priced at MSRP.
This still makes it £194 more expensive than the RTX 2070 SUPER, however, so what sort of performance increase can we expect to justify the additional cost? Alongside the Founders Edition, we also have a review of the MSI Gaming X Trio 2080 SUPER model for launch today, so be sure to check that out over here if you are interested in seeing what a high-end custom card can do.
|GPU||RTX 2080 SUPER||RTX 2080 (FE)||RTX 2070 SUPER||RTX 2070 (FE)||RTX 2060 SUPER|
|Rays Cast||8 Giga Rays||8 Giga Rays||7 Giga Rays||6 Giga Rays||6 Giga Rays|
|GPU Boost Clock||1815 MHz||1800 MHz||1770 MHz||1710 MHz||1650 MHz|
|Memory Data Rate||15.5 Gbps||14 Gbps||14 Gbps||14 Gbps||14 Gbps|
|Total Video Memory||8GB GDDR6||8GB GDDR6||8GB GDDR6||8GB GDDR6||8GB GDDR6|
|Memory Bandwidth||496.1 GB/sec||448 GB/sec||448 GB/sec||448 GB/sec||448 GB/sec|
To go over the core specs of the RTX 2080 SUPER in a little more detail, we first need to cast our mind back to September 2018, with the launch of the OG RTX 2080. For this card, Nvidia used its Turing-based TU104 GPU to offer 46 Streaming Multiprocessors, 2944 CUDA cores and 46 RT cores. This version of TU104 used for RTX 2080, however, was never actually a full implementation of the GPU, as full TU104 caps out at 48 SMs/3072 CUDA cores.
And that is what we get with RTX 2080 SUPER – a full TU104 with the accompanying two extra SMs. This is a relatively small jump in core spec, though – 2060 SUPER has four more SMs than original 2060, and 2070 SUPER also has four more SMs than original 2070.
To help further improve this new card, Nvidia has also outfitted the 2080 SUPER with new memory chips, clocked at 15.5Gbps. The GDDR6 memory we’ve seen with the rest of the RTX series is slower, clocked at 14Gbps. It’s important to stress that the 2080 SUPER memory is new, and Nvidia hasn’t simply taken the existing memory and factory overclocked it. We will look at manually overclocking the new memory later in this review.
Lastly, the extra cores and faster memory does result in an increase in total graphics power (TGP), with the 2080 SUPER rated for 250W, or 25W more than the original 2080 Founders Edition.