Nvidia released its RTX 30-Series of GPUs back in September of 2020, and it’s been nearly impossible to purchase these graphic cards since then, thanks to the global chip shortage. However, this hasn’t stopped Nvidia, as the company is reportedly already planning to announce its upcoming RTX 40-series GPUs sometime next year.
According to the latest reports by DigiTimes (via Wccftech), Nvidia will be developing its next-gen GPUs with TSMC. If these reports are accurate, it means that Nvidia will be shifting its strategy drastically for its RTX 40-series cards.
The current GeForce RTX 30-series GPUs are made using Samsung’s 8-nanometer fabrication process, while the TSMC’s 7nm process is reserved for data centre chips. But it is expected that Nvidia will be moving to a more advanced process, while it would rely entirely on TSMC’s 5nm process for its upcoming product lines.
Speaking of which, thanks to earlier reports, we already know that Nvidia will be launching two new GPU architectures codenamed ‘Ada Lovelace’ for its GeForce RTX 40 Series lineup of gaming GPUs and ‘Hopper’ for data centres and other enterprise-related applications.
While in terms of performance, the RTX 40 chips are rumoured to run at a whopping 64 teraflops (TFLOPs), which exactly doubles the power provided by the current top-of-the-line RTX 3080 chip which can hit up to 34 TFLOPs. For reference, the Xbox Series X, the most powerful console at the moment, provides 12 TFLOPs of computing power, while the PS5 can hit up to 10.3 TFLOPs.
This fantastic performance is only possible as ‘Ada Lovelace’ combines the smaller node size (5nm down from 8nm) with up to 18,432 CUDA cores, up from the 8,704 CUDA cores found in the RTX 3080. This would also result in up to 71% faster performance over chips based on Nvidia’s Ampere architecture.
Apart from the things mentioned above, very little is known about the GPUs power requirements. However, we suspect that the power requirement of the next-gen cards won’t be drastically higher than its predecessor, as going to a smaller node usually leads to better power efficiency while providing noticeable performance gains.
What are your thoughts on Nvidia’s upcoming RTX 40 Series GPUs? Let us know in the comments below. Until then, stay tuned to Oyprice and subscribe to our YouTube channel. Also, follow us on Telegram, Facebook, and Twitter.