How long until new ADC chips make it into products?

[jerryfreak]

looking at some of the recent releases by ESS of the 9292pro and 9823pro with amazing specs, its great to see the SOTA advancing

ESS Technology Unveils the World’s Highest Performing Audio CODEC, the SABRE ES9292PRO. Designed for uncompromising performance, flexibility and precision audio signal processing. – ESS Technology, Inc.

www.esstech.com

as a reference, the Topping M62 (released in fall 2025) uses the ES9821Q, which was announced in 2022
The Adept ADC released in fall 2025 uses the ES9842PRO announced in 2020

so these are ~3 years out before we see them in production gear?

 

[Keith Flanker]

Your ~3 year figure matches the pattern: ESS announces the die, then reference designs, firmware, PSU/thermal work and measurement validation stretch it out before a product ships at a sane price - the M62 (9821Q, 2022 to 2025) and the Adept (9842PRO, 2020 to 2025) both fit. There is also little pressure to move quickly: ES9038/9822-class parts already measure beyond what most products need, so designers sit on them. Measurement-first outfits like E1DA and Topping tend to be first, so at that pace I would expect the ES9292PRO family in gear around 2028-29.

 

[jerryfreak]

so its kinda like storage where the flagship hard drives are nearly twice the price and the second-tier densities are the 'sweet spot'?

like the new chips are available in sample quantities but the unit cost makes the older designs which are plenty good enough more appealing

i see mouser has the 9023 and 9043 (announced 16 mos ago) as "new products" and priced considerably higher than the 9826. i know its just a few bucks but im sure that adds up as prob only a small segment of the market are really spec nerds

i also see on the 9023 datasheet that while it was released in May it had a major revision six months later in nov. 25. so certainly not ready for prime time in mass production at announce date

id like to see the field recording mfrs lose the 32bit dual adc hackery with its inherent distortion from switching and just do a design with single 9823PRO chips in mono mode for each channel at -131 dB SINAD. that would be far more practical than attempting to stretch dynamic range into the 140s with SINAD that cant hit -100 dB

FWIW the 9821Qs in my topping M62 purchased in april 2026 were manufactured week 32 of 2024

 

[lewdish]

Ive been waiting for their codec products to end up in dongle dac products for so long now, some of the dongles ive seen support an ADC via the Codec chip but for whatever reason the oems dont utilize it

 

[KSTR]

id like to see the field recording mfrs lose the 32bit dual adc hackery with its inherent distortion from switching and just do a design with single 9823PRO chips in mono mode for each channel at -131 dB SINAD. that would be far more practical than attempting to stretch dynamic range into the 140s with SINAD that cant hit -100 dB

For field recording dynamic range is far more important than distortion, and two mono'ed 9823's per stereo channel is not really cost efficient either. A well designed dual-path ADC based on a single 9843 would be way more rewarding. After all, how a dual path ADC behaves near the path transition is mostly software defined and with a chip as good as the 9843 that transition has ample level range to live in, crossfading very gradually.

As for the your OP question, roughly three years from chip release to mature product mass market availability is very reasonable for a number of reasons, not all of them technical.