Following the latest product releases by NVIDIA, the tech giant has unveiled its latest professional GPU, the NVIDIA RTX 5880 Ada 48GB. While similar in performance to the RTX 5000 Ada, this new graphics card offers a significant boost in memory capacity and computing capabilities.
In terms of compute power, the RTX 5880 Ada features 14080 CUDA cores, 440 Tensor cores, and 110 RT cores, closely resembling the specifications of the RTX 5000 Ada with an approximate 10% improvement. This results in a 6% increase in single precision TFLOPs, RT TFLOPs, and Tensor TFLOPs. Notably, NVIDIA incorporates sparsity numbers in its Tensor TFLOPS, enhancing the overall efficiency of the GPU.
What sets the RTX 5880 Ada apart from its predecessor is its memory subsystem, which resembles that of the flagship RTX 6000 Ada. Additionally, the power consumption of the new GPU is more aligned with the RTX 6000 Ada rather than the RTX 5000 Ada.
This release addresses a gap in NVIDIA’s product lineup, as the previous RTX 5000 Ada with 32GB of memory felt lacking in comparison to the NVIDIA GeForce RTX 4090 with only 24GB, considering the notable price difference. By introducing the RTX 5880 Ada with 48GB, NVIDIA aims to bridge this gap and provide users with a more robust and cost-effective option. Furthermore, the enhanced compute resources of the new GPU offer further differentiation from its counterparts.
It is worth applauding NVIDIA for its emphasis on increasing memory capacity, particularly GDDR6, which significantly reduces costs compared to HBM technology found in higher-end cards like the A100 and H100.
In conclusion, NVIDIA’s relentless commitment to innovation is evident with the introduction of the RTX 5880 Ada. While maintaining performance levels close to the RTX 5000 Ada, the improved memory capacity and enhanced computing capabilities make this GPU a valuable addition to the professional graphics card market.
The source of the article is from the blog oinegro.com.br