When will nRF52 throughput be same as nRF51 or better

[Chris Barratt]

We are planning to use nRF52 for its superior computing power in an application. Unfortunately it appears that the throughput is more limited than on nRF51. Since we need to down load data this is a severe handicap.

Is there a planned solution and if so when and what?

Chris

[RK]

The throughput on nRF52 with the production softdevice shouldn't be any different than the nRF51. The softdevices support the same number of packets per connection, the same connection intervals etc and that's basically all there is to throughput.

So ..

1. What softdevices (versions and revision) are you comparing between nRF51 and nRF52?
2. Are you in peripheral or central mode
3. What does a sniffer trace of each tell you about the connection parameters negotiated and the number of packets transmitted per connection event?

[Christopher]

You can use the Softdevice S130-SD-v2 of the nRF51 or the S132-SD-v2 for the nRF52. Both of them allow you to configurable the BLE bandwidth.

The BLE throughput is the same for the nRF51 or 52, but you will consume about 50% power less with the nRF52 when transmitting data. And of course the computing power of the M4 is much higher than the M0.

[endnode]

I guess Chris is complaining about S110 V6-8 vs. S132 V2 (which is of course equal to S130 V2). Are they able to achieve the same number of MTUs (is it 6 or 8?) at connection interval of 7.5ms - of course in GAP Peripheral mode?

[FormerMember]

S130 v.2.0 supports 6 packets per connection interval, the softdevice specification has not yet been updated for S130 v.2.0.