PS4 Success In 2013 Helped AMD Steer Clear Of Bankruptcy

91

The success of Sony Interactive Entertainment’s previous-gen console, the PS4, in the year 2013 helped chip manufacturer AMD in steering clear of bankruptcy.

According to the LinkedIn profile of Renato Fragale, Senior Director of Consumer and Gaming Client Business at AMD, the company benefited greatly from the successful launch of the PS4 back in 2013, allowing it to avoid bankruptcy. As the Senior Manager of Product Development Engineering at the time, he managed a 15-person product development team to handle the manufacturing of the PS4 Accelerated Processing Unit (APU).

Fragale’s resume further mentions that the PS4 launch in 2013 is viewed as one of the most successful launches in AMD history, helping the company in avoiding bankruptcy. In doing so, the company went on to achieve success in the CPU market with its Ryzen lineup of processors. A custom version of one of these CPUs would ultimately make its way into the PS5 SoC in 2020. Meanwhile, AMD continues to lag behind NVIDIA in the GPU market, though it has been competitive in the price/performance ratio with its mid-range graphics cards. It’s clear that these contracts have played a major role in the company’s financial turnaround.

In terms of hardware specs, the PlayStation 4 features the aforementioned APU at its heart, which was developed by AMD in collaboration with Sony. It combines the CPU, GPU, memory controller, and video decoder. The CPU includes 8 64-bit x86-64 cores, with 7 cores available for development. The GPU has 18 compute units, achieving a theoretical peak performance of 1.84 TFLOPS. The system’s 8 GB GDDR5 memory runs at a maximum clock frequency of 2.75 GHz (5500 MT/s) and has a maximum bandwidth of 176 GB/s.

Despite being based on AMD’s GCN architecture, the PS4’s GPU has several distinguishing features compared to current-gen PC graphics cards with first-gen GCN architecture. Among them is an additional dedicated 20 GB/s bus that bypasses the L1 and L2 GPU cache for direct access to system memory, reducing synchronization challenges during fine-grained GPGPU compute tasks, simultaneous graphical and asynchronous compute tasks in the L2 cache with a ‘volatile’ bit tag, and an upgrade from 2 to 64 sources for compute commands, enhancing compute parallelism and execution priority control.