An early engineering prototype of NVIDIA's RTX 2070, known as the GeForce GTX 2070, has been spotted and features fewer cores than the original variant.

NVIDIA's Single “GTX” 20 Series GPU Spotted: An Engineering Sample Labeled GeForce RTX 2070

NVIDIA's GeForce “GTX” labeling was used up to the GeForce 10 and GeForce 16 series of GPUs. Everything after these GPUs was labeled under the GeForce “RTX” series, but it seems that the company internally labeled the RTX 20 series as the GTX 20 series because a similar model is now seen online. It has become common.

Photos posted by @unnatural__log Show a strange NVIDIA GeForce RTX 2070 GPU that mentions an engineering sample that explains why it doesn't use “RTX” labeling.

In terms of design, the NVIDIA GeForce GTX 2070 engineering sample has the same design as the GeForce RTX 2070. It has the same dual axial fan cooler with dual slot design but the label on the sides reads “GeForce GTX” and there is none. “GTX 2070” or “RTX 2070” in front. The card has some missing labels but other than that, no big difference. Display ports include the same HDMI, 2x DP, USB Type-C Virtual Link port, and DVI output. The card should have a standard 8-pin connector as it did before NVIDIA switched to a 12-pin or 16-pin interface.

The NVIDIA GeForce GTX 2070 engineering sample's PCB also matches the reference design, but it's the specifications that are the biggest change. While the RTX 2070 has 2304 cores, the GTX 2070 ES GPU has 2176 cores and is based on the TU106-400A-A1 GPU with the device ID “10DE-1F07 – 10DE-12AD”. This is the same core configuration used by the RTX 2060 SUPER which includes the TU106-410-A1 GPU.

Based on user testing, the NVIDIA GeForce GTX 2070 engineering sample GPU was flashed with the RTX 2070 “400A BIOS” BIOS which does not change the device ID or unlock additional cores. But it provides some extra power. to the GPU core which can be used for manual tweaking. With OC, the GPU ended up with about 95 percent of the RTX 2070's performance, or a 16 percent increase over its default profile. That's a pretty decent bump considering we don't see that kind of advantage on modern GPUs anymore because they already clock speeds above their listed “peak boost” clock speeds. are running

Overall, this shows that there was a time when NVIDIA decided not to use the “RTX” branding for their upcoming Turing GPUs that became the first “RTX” GPUs, thus the GeForce “GTX” series. The era is over. NVIDIA has also sized the output of The last GeForce GTX GPUs Which means it's all RTX from now on unless NVIDIA decides to switch to a new branding in the future.