Nvidia GeForce GTX 1080 V/S GTX 980 Ti – The Impact Of Pascal Architecture!
Pascal was developed by Nvidia after the Maxwell microarchitecture as successor to it in April 2016 with the GP100 chip. Later came the GP104 chip which had many changes as compared to the GP100 chip. This new chip shows similarities in design to the Maxwell Architecture, but what exactly is the difference?
The Nvidia GeForce GTX 1080 is based on the new Pascal Architecture while the GeForce GTX 980 Ti is based on the older Maxwell Architecture.
Advancements in the Pascal Architecture
- The streaming processor is partitioned into two processing blocks of each having 32 single precision CUDA cores.
- Improved Instruction buffer and warp scheduler.
- High Bandwidth memory
- Twice the registers as compared to the Maxwell architecture.
- GDDR5X is a new memory standard supporting 10Gbit/s data rates and an updated memory controller.
GTX 1080 v/s GTX 980
Nvidia was using the Maxwell Architecture for their earlier GPU’s like the GTX 980. Then the Pascal architecture came into existence with the GP100 chip. The GTX 980 and the new GTX 1080 differ a lot both in their architectures and manufacturing processes.
Let us compare them both on different parameters
Nvidia claims that the GTX 1080’s Pascal Architecture is the most efficient GPU ever built and guarantees you high performance and improved power efficiency. Some time back it was said the same about the GTX 980.
Manufacturing Process and Power
Both the Graphic cards differ a lot in manufacturing and the amount of power they consume. As said earlier the GTX 980 uses the Maxwell architecture as compared to the new GTX 1080 uses the Pascal Architecture.
With this change, Nvidia can now boost the graphic performance. The new architecture allows 2 billion more transistors to fit enabling improved performance output of its graphic cores. Even with the 2 billion more transistors in play the GTX 1080 draws only 15 W more power. Not only this, the GTX 1080 has more Cuda cores as compared to the GTX 980 Ti. The Cuda cores work the best under the Pascal Architecture and better organization and task management can be achieved.
As it can be seen above with the clock speed comparison, the GTX 1080 has improved clock speed (around 1600 Mhz) as compared to the GTX 980 Ti (around 1100 Mhz).Now, that’s a clock speed increase of almost 40%. All thanks to the new Pascal Architecture!
Talking about memory, the GTX 980 has a 6GB GDDR5, while the GTX 1080 deploys 8GB of GDDR5X, which delivers huge speed and bandwidth. The GTX 980 Ti has a bandwidth of 336 GB/s and memory bus is 384-bit, whereas the GTX 1080 has a bandwidth of 320 GB/s and memory bus is 256-bit.
Apart from having more memory, GTX 1080 doesn’t seem to impress on other fronts. But the memory speed of GTX 1080 is 10 GB/s whereas of GTX 980 Ti is only 7 GB/s. Therefore, while GTX 1080 has more memory, it can operate at much higher speeds compared to the GTX 980 Ti. Therefore, overall performance is much higher.
We will try and test two games GTA 5 and The Rise Of The Tomb Raider. We will study the graphic card performance of these both in frames per second. The test was conducted at a resolution of 3440×1440 and the conclusions can be seen in the charts below.
The Pascal architecture has been a huge leap for Nvidia in terms of both performance and usability. This is visible from the performance of the GTX 1080. This new architecture has also allowed Nvidia to build smaller, thinner graphics cards which can operate on low power. These graphics cards can now be used in a laptop which will give the almost same performance as that of a desktop grade graphics card. Therefore, You can now enjoy true power of graphics performance in a portable laptop.
The GTX 1080 costs something around 60Kin India while the GTX 980 costs around 40K. With this difference in price ranges, we will still suggest the GTX 1080 to go with because the manufacturing and the architectural changes are worth it.