Upcoming Nvidia RTX 3080 Ti GPU performance leaked

We already know that Nvidia is gearing up to launch its RTX 3000 series. The specifications of the 2080 Ti successor has not been made public by the manufacturer of the graphics. We, however, have an alleged specification leak of the upcoming RTX 3080 Ti.

According to sources, the new card will be based on the companies GA102 architecture. The specifications will lie between that of the RTX 3080 and 3090. The architecture of the chip is GA102-250-A1 and features a 384 bits bus along with GDDR6X memory.

The card is expected to be launched with either the 12 GB or 24 GB buffer; most of the high-end games are bottlenecked on the 11 GB ram of RTX 2080 Ti. Thus the company needs to launch its new Graphics processor with at least more than 11 GB of RAM to avoid disappointment.

The RTX 3000 series is based on Samsung’s 8nm process node, but the company is soon thinking of migrating to nodes manufactured by TSMC. According to sources, the new card will be a powerhouse with 34 TFLOPs of power. But the leaker claims that the current API, driver, and application infrastructure cannot fully take advantage of the card’s raw power.

Nvidia’s RTX 3080 Ti beats the predecessor RTX 2080 Ti in performance. It indicates that the performance of the Ampere GPUs is not inferior compared to the Turing GPUs.

AMD will be launching its big Navi series GPU today, and it seems that Nvidia is gearing to give its competitor a run for its money. NVIDIA needs to work with developers to properly fix the API and driver stacks to take advantage of the Ampere series’ raw performance properly. Nvidia and AMD are launching their GPUs this year, and it would be interesting to see who will win this race.

Source

Do check out: