Shulou(Shulou.com)11/24 Report--

Thanks to CTOnews.com netizen OC_Formula for the clue delivery! CTOnews.com Sept. 23, Nvidia recently released RTX 4090, 4080 16GB and 4080 12GB with AD102, AD103 and AD104 GPU, respectively. According to VideoCardz, Nvidia has now released complete information about the chip size and number of transistors for AD102, AD103 and AD104 GPU.

AD102 (RTX 4090): 608 mm ², 76.3 billion transistor

AD103 (RTX 4080 16G): 378.6 mm ², 45.9 billion transistor

AD104 (RTX 4080 12G): 294.5 mm ², 35.8 billion transistor

It is reported that the transistor density of the new GPU is higher than 121 million per square millimeter. The AD104 with 35.8 billion transistors has 7.5 billion more transistors than the previous generation's flagship GA102 GPU.

In addition, the number of render output units (ROP) of the new generation GPU is much higher than that of the previous generation, with an AD102 of up to 192 ROP,AD103 with a ROP of 112, as many as the previous generation's flagship GA102, while the AD103 has 80 ROP. A higher number of ROP improves rasterization performance.

The second-level cache of the new generation of GPU has also been improved, with 96MB of AD102 bit AD103 and 64MB of 48MB AD104. Nvidia has confirmed that both RTX 4080 models will be equipped with 48MB L2 cache.

Related readings:

Nvidia releases Ada Lovelace architecture GPU: TSMC 4N process, performance doubling, new ray tracing technology

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.