So what is under the hood?
Product name PlayStation®4
Product code CUH-2000 series
Single-chip custom processor
CPU : x86-64 AMD “Jaguar”, 8 cores
GPU : 1.84 TFLOPS, AMD Radeon™ based graphics engine
Memory GDDR5 8GB
Storage size* 500GB, 1TB
External dimensions Approx. 265×39×288 mm (width × height × length)
(excludes largest projection)
Mass Approx. 2.1 kg
BD/ DVD drive
(read only) BD × 6 CAV
DVD × 8 CAV
Input/ Output Super-Speed USB (USB 3.1 Gen1) port × 2
AUX port × 1
Ethernet（10BASE-T, 100BASE-TX, 1000BASE-T）×1
IEEE 802.11 a/b/g/n/ac
Power AC 100-240V, 50/60Hz
Power consumption Max. 165W
Operating Temperature 5 ºC – 35ºC
AV output HDMI™ out port (HDR output supported)
The PlayStation 4 uses a semi-custom Accelerated Processing Unit (APU) developed by AMD in cooperation with Sony and is manufactured by TSMC on a 28 nm process node. Its APU is a single-chip that combines a central processing unit (CPU) and graphics processing unit (GPU), as well as other components such as a memory controller and video decoder/encoder. The console also includes secondary custom chips that handle tasks associated with downloading, uploading, and social gameplay. These tasks can be handled seamlessly in the background during gameplay or while the system is in sleep mode.
Though not much is publicly known of the PS4’s audio capabilities, the console also contains a dedicated hardware audio module, which can support in-game chat with minimal external resources as well as “a very large number” of MP3 streams for use in in-game audio.
The main APU (2013 release) had a die size of 19 by 18.3 mm (0.75 by 0.72 in), with GPUs, CPUs and memory controllers on the same die. 2013 release version APUs contained 20 GCN compute units on die, two of which are thought to be present to provide redundancy to improve manufacturing yield. CPUs plus CPU caches make up approximately 15% of the chip area, and the GPU compute units take up approximately 33% of the 348 mm2 (0.539 sq in) die area.
Central processing units
The central processing unit (CPU) consists of two x86-64 quad-core modules for a total of eight cores, which are based on the Jaguar CPU architecture from AMD. Each core has 32 kB L1 instruction and data caches, with one shared 2 MB L2 cache per four core module. The CPU’s base clock speed is said to be 1.6 GHz. That produces a theoretical peak performance of 102.4 SP GFLOPS.
Graphics processing unit
1152 Stream Processors
72 Texture mapping units
32 Raster operators
18 Compute units
8 Asynchronous compute units (64 queues)
The graphics processing unit (GPU) is AMD’s GPGPU-capable Radeon GCN architecture, consisting of 18 compute units (CUs) for a total of 1,152 cores (64 cores per CU), that produces a theoretical peak performance of 1.84 TFLOPS. This processing power can be used for graphics, physics simulation, or a combination of the two, or any other tasks suited for general purpose compute. GPU is mostly based on the Bonaire architecture using GCN 1.1 technology.
Though based on AMD’s GCN architecture, there are several known differentiating factors between the PS4’s GPU and current-gen PC graphics cards featuring first-gen GCN architecture:
An additional dedicated 20 GB/s bus that bypasses L1 and L2 GPU cache for direct system memory access, reducing synchronisation challenges when performing fine-grain GPGPU compute tasks.
L2 cache support for simultaneous graphical and asynchronous compute tasks through the addition of a ‘volatile’ bit tag, providing control over cache invalidation, reducing the impact of simultaneous graphical and general purpose compute operations.
An upgrade from 2 to 64 sources for compute commands, improving compute parallelism and execution priority control. This enables finer-grain control over load-balancing of compute commands including superior game-engine integration.