NVIDIA is working on a Titan-class graphics card based on the next-gen Ada Lovelace microarchitecture. Leveraging a triple-fan heatsink, it’ll mark the return of the Titan brand after a four-year-long absence. Although we don’t know anything about the core specifications, I’d bet on a full-fledged implementation of the AD102 core with 18,432 FP32 shaders spread across 144 SMs and 12 GPCs. It’ll be paired with 48GB of GDDR6X memory across a 384-bit bus for a total bandwidth of over 1 TB/s.
Totally fake. The Titan doesn’t look like this. It has three fans.
The GeForce RTX Titan will likely be preceded by the RTX 4090 Ti, a higher-binned AD102 core. According to AGF on Twitter, it will feature 18,176 FP32 cores across 142 SMs. These dies will be cherry-picked bins, boosting up to 3GHz, a 200-300MHz bump over the RTX 4090.
The L2 cache on the Ti will also be a significant upgrade over the RTX 4090. While a 72MB LLC buffer complemented the latter, the 4090 Ti will utilize the 96MB available on the AD102. The memory will likely be pushed to 24Gbps, from 22.5Gbps on the current flagship. The RTX 4090 Ti is expected to have a TGP of 475W, a 25W bump over the 4090. On the other hand, the Titan should be rated at a TBP of 550W.
The GeForce RTX 4090 Ti should land early 2023, possibly with a CES 2023 reveal. The Titan usually precedes the corresponding Titanium variant. Still, since a full-fat die will be hard to produce in viably quantities this early, I’m betting on a mid to late 2023 launch.
Processors, PC gaming, and the past. I have written about computer hardware for over seven years with over 5000 published articles. I started during engineering college and haven't stopped since. On the side, I play RPGs like Baldur's Gate, Dragon Age, Mass Effect, Divinity, and Fallout. Contact: areejs12@hardwaretimes.com.
