Weird bumps in the speed from GPS

Are you using a speed sensor? If so, try disabling it to verify that it or isn't the cause and then retry with a new battery.

No speed sensor, using gps for speed and my power sensor is for power and cadence.

I also checked that it happens in totally open fields on the main road with no trees around. There seems to be a bump in altitude also when this happens even tho it should be flat so I suspect that's what is causing it.

Well, that eliminates the most obvious cause. Are there any hints (eg corners cut) in the track trace to suggest momentary GPS dropouts?

From what I can see nope... It's just a straight line perfectly along the side of the road. Looking at other spots it's also always quite accurate.

It might be correcting them of course or something strava though. It feels so weird to because it's like they are on a timer. Here's the long straight on the main road with time instead of distance.

The gaps between them are very even... a bit too much to be random, and it goes like this through the entire activity.

Unless it's interference from something else you are using, I'm out of ideas.

You might want to try adding a speed sensor. GPS can occasionally give odd transient data. It generally works well on flat open terrain, but even on those roads there can be glitches.

Speed sensor is always a "solution", the problem is that I have 3 others with me on the same rides a lot on and all of their speed graphs are straight using GPS while mine has gone absolute haywire.

it is an ancient bug in all Edge units. they (garmin) cant fix it, i reported it at least 5 times with various models (510, 820, 1000, 830, i had no more...)... (it is gps position duplication bug  which leads to distance value problem and the Strava is drawing the speed chart from the /change of/ distance data field from the FIT file.

yes, sometimes these spikes are very common, sometimes a bit rarely during an activity, due i couldnt read the source code i couldnt give exact info what can cause it. a theory is that a timing/priority issue between the cpu and gps part and if the cpu doesnt get back within 1s a new lat/lon positions from the gps chip then it is using the last known values which leads to gps data duplication , wrong distance data field values and so on... the Garmin was so lazy that they used lots of common (faulty) code parts from year 2013 model in the newer models and this is the end result... fck

yes, the speed sensor replaces the (gps) distance data field values with its own values , and this is why with speed  sensor the speed chart is not /so/ spiky (sometimes it could be even with speed sensor but that is an another bug, the missing data row from fit file, if a row is missing, there will be a  /weird square/ gap in the speed chart on Strava...)

Hmmm

Considering how evenly spaced out the things are on my device perhaps it's something running on the device slowing down the cpu then causing it. Unsure if I used the smartphone integration with bluetooth last year so I guess I could try turning that off.

It also seems to be worse if I start the activity too soon before the GPS is properly locked on. However that is just totally unverified.

Yesterdays ride was far better and only had the issue badly in one specific spot, interestingly enough the power sensor (+cadence) dropped a bit into the spiking before coming back.

So I'm guessing the sensor drops might be a bit related to this, I seem to get a dropped power sensor every 5 hours or so on average.