Re: Query: Protocol & Module Recommendations for 500k-Point Waveform Data Integration into EPICS

["Sukhanov, Andrei via Tech-talk" <tech-talk at aps.anl.gov>]

Hi,
FYI, I just released a module which simplifies python development of EPICS PVAccess servers using p4p package:

 https://github.com/ASukhanov/epicsdev
Tested it by generating one million of F32 points at 10 Hz and it occupies only 8% of CPU.

Concerning standard industrial protocols, the pyvisa-py is my choice for many off the shelf devices.

From other hand using direct python socket package in python is pretty simple, you may gain some performance with it.

Regards, 

Andrey

---

From: Tech-talk <tech-talk-bounces at aps.anl.gov> on behalf of lvhh--- via Tech-talk <tech-talk at aps.anl.gov>
Sent: Friday, January 16, 2026 4:39 AM
To: tech-talk <tech-talk at aps.anl.gov>
Subject: Query: Protocol & Module Recommendations for 500k-Point Waveform Data Integration into EPICS

 

Dear all, We have a custom device generating 500k-point waveform data (float32/float64 arrays) and need to convert it to EPICS PVData for system integration. We seek guidance on two key points: 1. What communication protocol should the custom

ZjQcmQRYFpfptBannerStart

This Message Is From an External Sender

This message came from outside your organization.

 

ZjQcmQRYFpfptBannerEnd

Dear all,

We have a custom device generating 500k-point waveform data (float32/float64 arrays) and need to convert it to EPICS PVData for system integration. We seek guidance on two key points:

1.What communication protocol should the custom device support? Given the large waveform size, we prioritize high throughput, low latency, and structured data (to bundle waveform + metadata like timestamps/units). We wonder which standard industrial protocols are suitable? (e.g. VISA, GigE)

2. Which EPICS modules/tools simplify communication?

The data does not require real-time monitoring. Our local device has a cache, and data can be retrieved from the cache to be stored as EPICS PVs.

Any  documentation references are greatly appreciated!

Best regards,

Huihui

---

lvhh at sari.ac.cn