An Nvidia Graphics card score has been leaked on Geekbench, which has shocked everyone. The two massive cards are rumoured to be Fat Boy and Little Boy from the upcoming Ampere architecture.

The graphics card represented an Nvidia GPU, featuring a massive 33 teraflops performance. Furthermore, there was two graphics card in Geekbench that featured 118 streaming Multiprocessors in one, and 108 in the other. If these specs are true, then we are looking at 7,552 CUDA Cores in primary, and 6,912 CUDA cores in the secondary card. The CUDA Cores really define how strong these two cards are.

Nvidia

Apart from CUDA Cores, the primary card also featured 24GB of Memory, running at 1.11Ghz, scoring 184,096 OpenCL in Geekbench. This result is uncomparable to any Nvidia card right now. If you compare it to Nvidia’s strongest card, the Quadro RTX 8000 and TITAN RTX, both fell short for about 50000 scores. The secondary card which is sort of the non-Ti version still managed to score ahead of Quadro RTX 8000 and TITAN RTX. The weaker card or Little Boy features 48GB VRAM and 1.01Ghz clock speed scoring 141,654.

Nvidia

Many reports claim that these two cards from the Nvidia Ampere Generation. We have never seen this much power from Nvidia over the past years. The Fat boy is ahead of all others by a significant margin, and Little Boy is still ahead of all other cards in the market.

This is bit shaky, here is why?
The Geekbench might not be reported number of CU’s correctly. According to a report by PCGamesN, Nvidia might change the way they are filling Sm’s in the upcoming GPUS. The rumors include Nvidia, doubling up on the floating-point units with the new Architecture, potentially leaving a number of integer units the same. Hence, we will get to see around 7,552 INT32 Units inside the Fat Man, featuring a total of 15,104 FP32 units.

The Little Boy, featured 48GB however Geekbench shows 46.8 GB only. This might be a bug, because we are clearly able to see that there are 108 compute units. If this is the case, then we might have 6.912 CUDA Cores.

Conclusion


Ampere Specs and Leaks have been on the internet from the past few months now. The specs are incredible and interesting. Nvidia is planning something out of the box this time, that would really bring them on top comparing their rival AMD. The OpenCL Scores are very higher than top order Nvidia cards, which has given us an idea about the upcoming generation of Nvidia cards. The upcoming months will be challenging for both Nvidia and AMD. Suppose, AMD releases its big Navi, and the rumors of Navi sort to be true. Then, this is the actual Nvidia card that will be used as a response to AMD’s latest Big Navi.
Lastly, Nvidia is set to host Graphics Technology Conference in San Jose this March. We will hear from Jen-Hsun-Huang, introducing a newer generation of the graphics architecture.

Source: PCGames, Geekbench – Fat Boy & Little Boy.

Half-Life: Alyx won’t be delayed
Lost Words: Beyond the Page gets New Gameplay Trailer.
Latest Update Allows iPad Pro to run Fortnite at 120 frames per second
Riot could end up paying $400 Million to settle the lawsuit.