Last week, the internet was blessed with some specifications for Sony's upcoming PlayStation 5 Pro, revealing a significant performance uplift compared to the standard PlayStation 5.
The original specifications leak began from Moore's Law is Dead, which was later backed up by Insider Gaming's Tom Henderson, a notable insider within the industry. Henderson wrote that the leaks were real and that Insider Gaming had seen leaked PS5 Pro documents sourced from a verifiable PlayStation developer portal. Now, Henderson has posted a new report revealing even more PS5 Pro specifications, such as System Memory, CPU, and Audio specs.
Insider Gaming has claimed to have learned more specifications of the upcoming console, and for comparison sake, the publication has included the specs of the standard PlayStation 5. The PS5 Pro system memory is rumored to have 576 GB/s (18GT/s), which is a 28% increase over the standard console. As for CPU, the PS5 Pro is rumored to have an identical CPU to the PS5 but includes a "High CPU Frequency Mode" that bumps the CPU to 3.85GHz, which is a 10% increase over the standard console.
Lastly, the PS5 Pro is rumored to have an ACV that has much higher clock speeds, providing 35% more performance in the ACM library compared to the standard PS5. These leaks pointed to fall 2024 release for the PS5 Pro.
System Memory
- Standard PlayStation 5 - 448 GB/s (14 GT/s)
- PlayStation 5 Pro - 576 GB/s (18GT/s) - A 28% increase over the standard console
CPU
- PS5 Pro has "High CPU Frequency Mode", boosting the CPU to 3.85GHz - A 10% increase over the standard PS5
Audio
- 35% performance in the ACM library
- More convolution reverbs can be processed
- More FFT or IFFT can be processed
GPU (Previously revealed)
- Rendering 45% faster than PS5
- 2-3x Ray-tracing (x4 in some cases)
- 33.5 Teraflops
- PSSR (PlayStation Spectral Super Resolution Upscaling) upscaling/antialiasing solution
- Support for resolutions up to 8K is planned for future SDK version
- Custom machine learning architecture
- AI Accelerator, supporting 300 TOPS of 8 bit computation / 67 TFLOPS of 16-bit floating point