Editor's Desk GPUs Graphics Reviews

Gainward GeForce RTX 2060 Super Ghost 8 GB review – How good is the RTX entry in cheap? | Profile

Gainward would like to prove to us that RTX is also cheap with the GeForce RTX 2060 Super Ghost. Today’s test will have to show us whether the concept of the very cost-effectively constructed card can work. The Nvidia-owned RTX alternatives without super in the appendix are slowly becoming rarer and those who buy new will probably have to arrange themselves super soon. It doesn’t have to be a disadvantage, but I don’t want to spoil anything yet. For this we have our new form of Plant-Profile, in order to be able to bring the facts, as desired (and also intended), shorter and more concise to the man, the woman or It, depending on the.

But back to the map, which I would like to take a closer look at today. What makes it interesting is the price that also makes it one of the cheapest GeForce RTX 2060, even if, of course, as always, everything depends on the miseral form of the day. But for that we love the street and its prices. You’re never sure of surprises.



The chip: TU106-410 of the GeForce RTX 2060 Super

The TU106-410 of the GeForce RTX 2060 Super is reduced to 2176 CUDA cores, 272 tensor cores, 34 RT cores, 136 TMUs and 64 ROPs compared to the full version TU106-400 on the RTX 2070. The card will have clock frequencies in the range of 1470 MHz (base clock) and 1650 MHz (boost) and can thus simultaneously deliver up to 7 TFLOPS computing power. The 34 RT cores can deliver about 6 giga-rays per second in ray tracing performance.

Compared to the RTX 2060, the RTX 2060 Super again has 4 MB L2 cache instead of the three MB of the RTX 2060. Otherwise, everything I have already written in the launch articles for the GeForce RTX 2070 FE and the GeForce RTX 2060 FE applies. And if you want to find out more about all the RTX features in detail, please refer to my long basic article “Nvidia GeForce RTX 2080 Ti and RTX 2080 – what is really hidden behind Turing” on the subject of RT. This, too, is read-proof and certainly worthwhile.

Uncircumcised, the TU106-410 also retains the same 256-bit memory bus as the TU106-400, which has eight 1 GB GDDR6 memory modules with 14 Gb/s data rate, resulting in bandwidth of up to 448 GB/s. As with the GeForce RTX 2070, NVLink support in this price range is long gone. A rogue who thinks evil in it, but what’s the point?

Although the TU106 is the least complex Turing-based GPU to date, its 445 mm2 chip contains no less than 10.8 billion transistors. That’s still huge for what Nvidia might once have considered the center of its portfolio. In comparison, the GP106, i.e. the “Midrange Pascal”, was a small 200 mm2 chip with 4.4 billion transistors.

Key data: unboxing, dimensions, weight and features

The palit card weighs only 582 grams and measures 23.6 cm from the outside edge of the slot panel to the outer edge of the radiator cover. With 11 cm from the top of the motherboard slot to the top edge of the cover, it is relatively flat and also the installation depth (“thickness”) of 3.8 cm makes it a real dual slot card. The radiator cover made of black ABS is rather inconspicuous, the backplate made of plastic is purely optical in nature. It cannot cool.

The slot bezel lets out a smaller part of the warm exhaust air directly, as the cooling fins are horizontally aligned. However, the rest disappears as usual in the depths of the housing. With a DVI port (digital only), a DisplayPort 1.4 and an HDMI 2.0 port, there are enough variants to connect to the monitor.

Installation length (gross) 23.6 cm
Installation height (gross) 11.0 cm
Installation depth front (gross) 3.8 cm
Installation depth rear (gross) 0.5 cm (backplate)
Weight: 582 g
Connections: 1x HDMI 2.0
1x DisplayPort 1.4
1x DVI (digital)
1x 8-pin PCIe Power Supply
Cooler cover: ABS injection moulding, solid black
Fan: 2x 8.5 cm rotors with 9 rotor blades each

A first overview of the electrical data here is the latest version of GPU-Z:

The overview of the relevant comparison maps then looks like this:

  GeForce RTX 2060 Super
GeForce RTX 2060 FE GeForce RTX 2070 FE
GeForce GTX 1070 FE
Architecture (GPU)
Turing (TU106-410) Turing (TU106-300) Turing (TU106-400) Pascal (GP104)
CUDA Cores
2176 1920 2304 1920
Tensor Cores
272 240 288 N/A
RT Cores
34 30 36 N/A
Texture Units
136 120 144 120
Base Clock Rate
1470 MHz 1365 MHz 1410 MHz 1506 MHz
GPU Boost Rate
1650 MHz 1680 MHz 1710 MHz 1683 MHz
Storage expansion
Storage bus
256-bit 192-bit 256-bit 256-bit
448 GB/s 336 GB/s 448 GB/s 256 GB/s
64 48 64 64
L2 Cache
4 MB 3 MB 4MB 2MB
175 W 160 W 185W 150w
10.8 billion 10.8 billion 10.8 billion 7.2 billion
The size
445 mm2 445 mm2 445 mm2 314 mm2
No No No Yes (MIO)

Test system and setup

Elegant transition and also a view of the test system, which this time relies on AMD’s socket AM4 and X570, Intel’s socket 2066 on the X299 and the socket 1151 including Z390. I used only tried-and-tested boards from MSI (X570, Z390), as well as Aorus (X299). The final question of what to do with RAM has long been a matter of concern to me. Should I run the 32 GB DDR4 on the same clock as the processor manufacturer specifies in the specs, or should I run all CPUs with the same clock?

Symbolic Picture from igorsLAB: GPU, Motherboard and CPU Testing

The 32 GB extension is new and fits the velvet system. I have listed this tabularly again in detail:

Test System and Equipment

Intel Core i9-9900 K
MSI MEG Z390 Godlike

4x 8GB G.Skill FlareX DDR4 3200
1x 2 TByte Aorus (NVMe System SSD, PCIe Gen. 4)
1x Seagate FastSSD Portable USB-C
Seasonic Prime 1200 Watt Titanium PSU

Alphacool Ice Block XPX (1151), XPX Pro (AM4, 2066)
Alphacool Ice Grinder (modified)
Thermal Grizzly Kryonaut
Lian Li T70, Raijintek Paean
Open Benchtable
Monitor: BenQ PD3220U
Power Consumption:

Non-contact direct current measurement on PCIe slot (riser card)
Non-contact direct current measurement at the external PCIe power supply
Direct voltage measurement at the respective connectors and at the power supply unit
2x Rohde & Schwarz HMO 3054, 500 MHz multichannel oscilloscope with memory function
4x Rohde & Schwarz HZO50, current clamp adapter (1 mA to 30 A, 100 KHz, DC)
4x Rohde & Schwarz HZ355, probe (10:1, 500 MHz)
1x Rohde & Schwarz HMC 8012, digital multimeter with memory function

Thermal Imager:
1x Optris PI640 + 2x Xi400 Thermal Imagers
Pix Connect Software
Type K Class 1 thermal sensors (up to 4 channels)
NTI Audio M2211 (with calibration file)
Steinberg UR12 (with phantom power for the microphones)
Creative X7, Smaart v.7
Own anechoic chamber, 3.5 x 1.8 x 2.2 m (LxTxH)
Axial measurements, perpendicular to the center of the sound source(s), measuring distance 50 cm
Noise emission in dBA (slow) as RTA measurement
Frequency spectrum as graphic
Os: Windows 10 Pro (1909, all Updates)

Gainward GeForce RTX 2060 SUPER Ghost, 8GB GDDR6, DVI, HDMI, DP (1198)

Danke für die Spende

Du fandest, der Beitrag war interessant und möchtest uns unterstützen? Klasse!

Hier erfährst Du, wie: Hier spenden.

Hier kannst Du per PayPal spenden.

About the author

Igor Wallossek

Editor-in-chief and name-giver of igor'sLAB as the content successor of Tom's Hardware Germany, whose license was returned in June 2019 in order to better meet the qualitative demands of web content and challenges of new media such as YouTube with its own channel.

Computer nerd since 1983, audio freak since 1979 and pretty much open to anything with a plug or battery for over 50 years.

Follow Igor:
YouTube Facebook Instagram Twitter