Qualcomm has announced upgrades to some of its sound platform chips, with the S3 gen 3 and S5 gen 3. Here we look at the major changes made.
The S3 is aimed at mid-tier audio devices. Looking at the specs for gen 2 and gen 3 the major upgrades on the previous generation include:
- 2 x compute power
- An additional Kalimba DSP (2 instead of 1)
- Improved DAC delivering better SNR and a reduced noise floor
- Connectivity to external SRAM or 2nd Flash Device
- Addition of stereo audio playback interface
- Availability in VFBGA package (WLCSP too).
The S3 also gives customers access to 3rd party solutions from the Qualcomm Voice & Music Extension Program. There are more details on the S3 here.
The S5 is aimed at higher end devices and has seen a much more major overhaul. Some of the major upgrades include:
- 3x increase in compute power
- 2x increase in DSP power (Move from Qualcomm Kalimba to Tensilica HiFi DSP)
- 50x more on-device AI (Dedicated AI core)
- 1.5x more memory (RAM)
- Addition of “4th Generation Qualcomm ANC”
- Improved DAC delivering better SNR and a reduced noise floor
The move to using the Tensilica HiFi DSP has significantly improved the performance of the latest S5 chip. Qualcomm states on the S5, Gen 3 product page that the DSP clock speed is now 350 MHz, up from 240 MHz in the S5, Gen 2. This, alongside a new CPU delivering a clock speed of 200 MHz, up from 80MHz, and dedicated NPU, has delivered a much more powerful platform for premium devices.
