Introduction

NVIDIA today refreshed its GeForce RTX 20-series graphics card lineup with the new "Super" RTX 20-series. The RTX 20-series graphics cards have already been in the market for over eight months, and AMD is finally addressing the performance segment with its 7 nm PCIe Gen 4 Radeon RX 5700 "Navi" series. AMD claims that the Navi-based Radeon RX 5700 and RX 5700 XT are faster than the GeForce RTX 2060 and RTX 2070 respectively, so NVIDIA decided to upscale the two SKUs with new models. We hence have the new GeForce RTX 2070 Super and GeForce RTX 2060 Super, which will be available in stores starting on the 9th of July. The new models will sit at the same price points as the RTX 2060 and RTX 2070, and the existing cards will drop slightly in price. The two will be joined by the high-end GeForce RTX 2080 Super later this month, on the 23rd.

The new RTX 2070 Super is a whole different beast from the original RTX 2070, which had maxed out the "TU106" silicon. NVIDIA is having to tap into its larger sibling, the 12 nm "TU104". This is the same chip on which the RTX 2080 and upcoming RTX 2080 Super are based. The RTX 2070 Super is endowed with 2,560 CUDA cores, 40 RT cores, and 320 Tensor cores. The memory subsystem is unchanged, with 8 GB of GDDR6 memory across a 256-bit interface ticking at 14 Gbps (448 GB/s). The GPU clock speeds have been increased to 1605 MHz base and 1770 MHz boost, compared to 1410 MHz base and 1620 MHz boost of the original RTX 2070. Being based on the "TU104" has another perk—NVLink support enabling 2-way SLI. The lack of multi-GPU capability was one of the major criticisms of the RTX 2070, especially since some of the custom-design cards are sold for upward of $600.



NVIDIA is targeting the upcoming Radeon RX 5700 XT with the RTX 2070 Super. However, at $499, it comes at a $50 premium over the disclosed $449 MSRP of the RX 5700 XT. NVIDIA is also allowing its add-in card (AIC) partners to release custom-design graphics cards based on this chip, which will begin rolling out later this month. Today, only reviews of the Founders Edition are allowed.


As we stated earlier, NVIDIA has carved out the RTX 2070 Super from the "TU104" silicon by disabling four TPCs or eight SMs (streaming multiprocessors). Four TPCs also happens to make up a GPC (graphics processing cluster) on the "TU104". We don't know whether an entire GPC is disabled to achieve the SM count of 40 or one TPC is disabled across four GPCs (a more optimal way to do this). The RTX 2070 Super ends up with a CUDA core count of 2,560, an RT core count of 40, a Tensor core count of 320, 160 TMUs, and 64 ROPs.

In this review, we take the RTX 2070 Super Founders Edition card for a spin. This card looks similar to the RTX 2080 Founders Edition since it's based on the same silicon and reference PCB. A chrome top-plate with green "Super" badging next to "RTX 2070" and another badge on the back-plate are all that set this card apart. As with its original RTX 20-series cards, the Founders Edition cards by NVIDIA don't tick at reference clock speeds. This card has an overclocked GPU Boost frequency of 1770 MHz.

Packaging and Contents

You will receive:

• Graphics card
• Documentation
• DisplayPort to DVI-D adapter

The Card

Visually, the RTX 2070 Super looks identical to the RTX 2080 with the exception of the green "Super" badge on a mirror-reflective background. The backplate seems identical, too, just another "Super" badge has been added. Dimensions of the card are 27.0 x 11.5 cm.


Installation requires two slots in your system.


Display connectivity options include three standard DisplayPort 1.4a, one HDMI 2.0b, and a VirtualLink connector, which is basically USB-C with DisplayPort routing and USB-PD, so a single cable can power, display, and take input from your VR HMD.

NVIDIA has updated their display engine with the Turing microarchitecture, which now supports DisplayPort 1.4a with support for VESA's nearly lossless Display Stream Compression (DSC). Combined, this enables support for 8K@30Hz with a single cable, or 8K@60Hz when DSC is turned on. For context, DisplayPort 1.4a is the latest version of the standard that was published in April, 2018.

At CES 2019, NVIDIA announced that all their graphics cards will now support VESA Adaptive Sync (aka FreeSync). While only a small number of FreeSync monitors have been fully qualified for G-SYNC, users can enable the feature in NVIDIA's control panel regardless of whether the monitor is certified or not.


The board uses an 8-pin and a 6-pin power connector. This input configuration is specified for up to 300 watts of power draw.


Unlike the regular RTX 2070, the RTX 2070 has support for SLI through the NVLink interface, which is because the card is based on the TU104 GPU (like the RTX 2080), whereas the RTX 2070 is based on TU106.

Disassembly

Disassembly is just as complicated as on the RTX 2080, which does make it a little bit easier compared to disassembling the RTX 2070. Our original RTX 2080 review has more details on that.


An aluminium base plate touches a vapor-chamber plate, which in turn is soldered to an aluminium fin stack that is ventilated by the pair of fans. Quite impressive how a single piece of cooler can cool all the components on the PCB.


The backplate is made out of metal and protects the card against damage during installation and handling. It slightly wraps around the back of the card to create a stellar curved look.

On the next page, we dive deep into the PCB layout and VRM configuration.

High-resolution PCB Pictures

These pictures are for the convenience of volt-modders and those who would like to see all the finer details on the PCB. Feel free to link back to us and use them in your articles or forum posts.



High-res versions are also available (front, back).