Earlier this month Nvidia announced its latest GeForce, GTX 1080. Based on its Pascal architecture and built on the 16nm FinFET process from TSMC, the GTX 1080 will be the first 16nm graphics card / based on 14nm. The result is that this board will be the new brand for a single performance GPU.
Contrary to what happened earlier NVIDIA is taking to launch the GTX 1080 for the 27th, next Friday, with pre-orders to be made from this Friday. It is almost guaranteed that the first batch of cards will sell out, but potential buyers will have a few days to mull over their data, and decide whether to pay $ 699.
The analysis of AnandTech, I’ll hold it, rather than hasten it as a less complete article. How do I know everyone is anxious to know what will be the performance, I wanted to look at the card and the numbers as a preview of what’s to see. The performance was made available in the new GPU section Anand Tech Bank, for those who want to see results in resolution and other details regarding settings.
Regarding the Pascal architecture is a mix of new and old, is not a new revolution but it’s like a major improvement. Maxwell to an architecture was very successful for NVIDIA BOTH at the level and the consumer level professional, and for the consumer iterations of Pascal, NVIDIA has not made any radical changes. For other details, we ALUs, texture units, ROPs, and caches all that execute similar to how they did in GM2xx.
NVIDIA’s focus on this generation has been the clockspeed to increase the transfer rate to 9 TFLOPs and update the memory subsystem.
To que end NVIDIA has sort of run with the idea of boosting clockspeeds, and relative to Maxwell They have done additional work at the level chip design to allow for higher clockspeeds at The Necessary critical paths. All this will help to achieve clockspeed without blowing the power budget of 1080 GTX.