NVIDIA’s next-gen RTX 40 series GPUs are just around the corner. But there is one rumored GPU that has even better specs than the flagship. Previous rumors suggested that NVIDIA was going flat out with its Ada Lovelace AD102 card with some insane specifications.
However, the information we have now says that the GPU will be rocking the Ada Lovelace architecture & feature a slightly cut-down configuration, “AD102-450-A1“, rocking 142 Streaming Multiprocessors which is around 11% more than RTX 4090. According to kopite7kimi, the CUDA core count will be a whopping 18,176(out of 18432 CUDA cores in the flat-out design).
Based on a clock speed of around 3 GHz, this graphics card will easily break past the 100 TFLOPs Compute barrier. The graphics card is said to be equipped with 48 GB of GDDR6X memory running across a 384-bit bus interface.
It not only features a higher capacity than RTX 4090 (24GB) but also a faster memory (24Gbps vs. 21 Gbps). Such configuration enables a maximum bandwidth of 1.1 TB/s.
Coming to Power consumption, the GPU has a TBP rating of 800W, almost double of the flagship RTX 4090. It is worth noting that board power is the maximum power the board may require from the power supply, it is not the same metric as TDP.
Considering that a single 16-pin connector can only provide 600 Watts of power, a dual 16-pin connector config will have to be utilized for this monster of a card if it ever becomes a reality. The graphics card will use the PG137-SKU0.
The leaker did not mention the name of this graphics card but nicknamed it ‘The beast’, although previous rumors hinted that the GPU might be called ‘Titan’ or could be the Ti version of RTX 4090.
What is known for sure is that this GPU will launch after the RTX 4090, which is scheduled to launch this fall. With specs like this, and the RTX 40 series already launching this year, we might have to wait till next year to see this Titan beast.
| NVIDIA RTX Beast and 40-series Rumored Specs | RTX Beast/TITAN | RTX 4090 | RTX 4080 | RTX 4070 |
| Architecture | Ada (TSMC N4) | Ada (TSMC N4) | Ada (TSMC N4) | Ada (TSMC N4) |
| GPU | AD102-400 | AD102-300 | AD103-300 | AD104-400 |
| Board Number | PG137-SKU0 | PG139-SKU330 | PG139-SKU360 | PG141-SKU341 |
| SMs | 142 | 126 | 80 | 56 |
| CUDA Cores | 18176 | 16384 | 10240 | 7168 |
| Memory | 48GB G6X | 24 GB G6X | 16 GB G6X | 10 GB G6 |
| Memory Bus | 384-bit | 384-bit | 256-bit | 160-bit |
| Memory Speed | 24 Gbps | 21 Gbps | 21 Gbps | 18 Gbps |
| Power | ~800W (TBP) | ~450W (TDP) | ~420W (TDP) | ~300W (TDP) |
| Launch Date | TBC | October 2022 | November 2022 | December 2022 |
About Nvidia
NVIDIA Corporation is an American multinational technology company incorporated in Delaware, based in Santa Clara, California. They design graphics processing units (GPUs) for the gaming and professional markets, as well as system on chip units (SoCs) for the mobile computing and automotive market.
Best known for the “GeForce” lines of GPUs, they are a direct competitor to AMD’s “Radeon” series. NVIDIA has also expanded its offerings with its handheld game consoles Shield Portable, Shield Tablet, and Shield Android TV and its cloud gaming service GeForce Now.
No Comments on NVIDIA “Beast” Ada Lovelace GPU to Feature 18176 Cores, 48 GB Memory, & 800W TBP