SQL Azure query aggregate performance issue - sql

Question

I'm trying to improve our SQL Azure database performamce, trying to change the use of CURSOR while this is (as everybody told me) something to avoid.
Our table is about GPS information, rows with a id clustered index and secondary indexes on device, timestamp and geography index on location.
I'm trying to compute some statistic such minimum speed (doppler and computed), total distance, average speed, ... along period for a specific device.
I have NO choice on the stat and CAN'T change the table or output because of production.
I have a clear performance issue when running this inline tbl function on my SQL Azure DB.
ALTER FUNCTION [dbo].[fn_logMetrics_3]
(
#p_device smallint,
#p_from dateTime,
#p_to dateTime,
#p_moveThresold int = 1
)
RETURNS TABLE
AS
RETURN
(
WITH CTE AS
(
SELECT
ROW_NUMBER() OVER(ORDER BY timestamp) AS RowNum,
Timestamp,
Location,
Alt,
Speed
FROM
LogEvents
WHERE
Device = #p_device
AND Timestamp >= #p_from
AND Timestamp <= #p_to),
CTE1 AS
(
SELECT
t1.Speed as Speed,
t1.Alt as Alt,
t2.Alt - t1.Alt as DeltaElevation,
t1.Timestamp as Time0,
t2.Timestamp as Time1,
DATEDIFF(second, t2.Timestamp, t1.Timestamp) as Duration,
t1.Location.STDistance(t2.Location) as Distance
FROM
CTE t1
INNER JOIN
CTE t2 ON t1.RowNum = t2.RowNum + 1),
CTE2 AS
(
SELECT
Speed, Alt,
DeltaElevation,
Time0, Time1,
Duration,
Distance,
CASE
WHEN Duration <> 0
THEN (Distance / Duration) * 3.6
ELSE NULL
END AS CSpeed,
CASE
WHEN DeltaElevation > 0
THEN DeltaElevation
ELSE NULL
END As PositiveAscent,
CASE
WHEN DeltaElevation < 0
THEN DeltaElevation
ELSE NULL
END As NegativeAscent,
CASE
WHEN Distance < #p_moveThresold
THEN Duration
ELSE NULL
END As StopTime,
CASE
WHEN Distance > #p_moveThresold
THEN Duration
ELSE NULL
END As MoveTime
FROM
CTE1 t1
)
SELECT
COUNT(*) as Count,
MIN(Speed) as HSpeedMin, MAX(Speed) as HSpeedMax,
AVG(Speed) as HSpeedAverage,
MIN(CSpeed) as CHSpeedMin, MAX(CSpeed) as CHSpeedMax,
AVG(CSpeed) as CHSpeedAverage,
SUM(Distance) as CumulativeDistance,
MAX(Alt) as AltMin, MIN(Alt) as AltMax,
SUM(PositiveAscent) as PositiveAscent,
SUM(NegativeAscent) as NegativeAscent,
SUM(StopTime) as StopTime,
SUM(MoveTime) as MoveTime
FROM
CTE2 t1
)
The broad idea is
CTE is selecting the correponding rows, following the parameters
CTE1 perform aggregation within two consecutive row, in order to get Duration and Distance
then CTE2 perform operation on these Distance and Duration
Finally the last select is doing aggregation such sum and average over each columns
Everything working pretty well, until the last SELECT call where the agregate function (which are only few sum and average) killed the performance.
This query selecting 1500 rows against table with 4M rows is taking 1500ms.
when replacing the last select with
SELECT ÇOUNT(*) as count FROM CTE2 t1
then it's take only few ms.. (down to 2ms according to SQL Studio statistics).
with
SELECT
COUNT(*) as Count,
SUM(MoveTime) as MoveTime
it's about 125ms
with
SELECT
COUNT(*) as Count,
SUM(StopTime) as StopTime,
SUM(MoveTime) as MoveTime
it's about 250ms
like each aggregate are running on consecutive loop operation over all the row, within the same thread and without beeing parallelized
For information, the CURSOR version (I wrote couple of year ago) of this function is running actually at least twice fast...
What is wrong with this aggregate? How to optimize it?
UPDATE :
The query plans for
SELECT COUNT(*) as Count
The query plans for the full Select with agregate
According the answer of Joe C, I introduce a #tmp table in the plans and perform the aggregate on it. The result is about twice as fast, which is an interesting fact.