AMD revealed its strategy to confront Qualcomm on the GPU playground for smartphones 2AMD revealed its strategy to confront Qualcomm on the GPU playground for smartphones 2

Last June, Samsung and AMD announced a strategic partnership to bring AMD’s `Next Gen` GPU architecture to mobile devices.

Considered the successor to AMD’s GCN (Graphics Core Next) microarchitecture, RDNA not only features changes in the number of small cores to the memory and interconnects within it, it also includes

According to AMD’s white paper, GPUs built on the RDNA architecture will span a wide range of devices, including notebooks, smartphones as well as some of the world’s largest supercomputers.

Does AMD’s GPU meet the requirements of smartphones?

Although it is difficult to predict the performance of AMD GPUs through technical descriptions in the white paper, we can see that RDNA brings optimizations suitable for use on mobile devices.

The L2 cache can also be configured to provide levels ranging from 64KB to 512KB depending on the performance, power of the application, and the silicon area it targets.

AMD revealed its strategy to confront Qualcomm on the GPU playground for smartphones

AMD’s mobile GPU architecture will move from supporting 64 work items with the GCN architecture to a more optimized 32 work items with RDNA.

AMD says the benefit of this parallel computing is to distribute the workload to more cores, improving performance and energy efficiency.

This shows that AMD is paying close attention to memory and power consumption – the two most important areas for any successful GPU on smartphones.

Take advantage of Radeon’s AI tasks

AMD’s GCN architecture, the predecessor to RDNA, also has a unique advantage in machine learning or artificial intelligence workloads.

With the new architecture, RDNA still retains high-performance machine learning components, with the ability to support 64-bit, 32-bit, 16-bit, 8-bit, or even 8-bit parallel arithmetic calculations.

AMD revealed its strategy to confront Qualcomm on the GPU playground for smartphones

The FMA calculation is so popular in machine learning applications that ARM’s Mali-G77 GPU even has a separate hardware block inside to process those calculations.

In addition, RDNA also introduces the ACE (Asynchronous Compute Tunneling) compute channel to manage shader workloads.

RDNA is designed to be more flexible

Besides the above advantages, AMD’s white paper also mentions a series of other improvements implemented by this new microarchitecture.

AMD’s white paper says: `To increase performance from low to high levels, other GPUs can increase the number of shader arrays and change the resource balance within each array.`

Nvidia and ARM also have similar practices on their CUDA and Mali GPUs when increasing or decreasing the number of processing cores depending on the required power and performance goals.

AMD revealed its strategy to confront Qualcomm on the GPU playground for smartphones

This will result in a more flexible platform, with a design optimized for better expansion or contraction than previous AMD products.

When will the collaborative GPUs between Samsung and AMD be released?

According to Samsung’s announcement in its most recent earnings call, we are still `about 2 years away` from when the company launches a new GPU based on the RDNA architecture.

However, with the details in the RDNA white paper, we have an initial look at AMD’s plans to bring its famous GPU architecture to low-power devices and smartphones.

AMD revealed its strategy to confront Qualcomm on the GPU playground for smartphones

Refer to Android Authority

Avatar of Logan

By Logan

Leave a Reply

Your email address will not be published. Required fields are marked *