Re: NOP optimises Group By?

[Rick James]

The answer probably lies in the indexes; please provide
* SHOW CREATE TABLE tbl\G -- engine, indexes
* SHOW TABLE STATUS LIKE 'tbl'\G -- sizes
* EXPLAIN SELECT ...\G -- clues of inefficiencies
* SHOW VARIABLES LIKE '%buffer%'; -- cache size
and surround them with [ code ] and [ / code ]

What is "[PercentExtend]" ??

Your queries, cleaned up a bit:

SELECT  v.PointId,
        Sum((100-[PercentExtent]))/523  AS Weight,
        Count(*)  AS SppCount
    FROM  sppmodelvalues v
    INNER JOIN  spphistograms h
           ON     (v.SppId = h.SppId)
             AND  (v.Prob  = h.Prob)
    WHERE  v.SppId > 0         -- the No-op
    GROUP BY  v.PointId;

I would hope for indexes on both tables of (SspId, Prob) or (Prob, SspId).
sppmodelvalues might benefit from INDEX(SspId, Prob, PointId)