4-mapd.txt

AWS Marketplace:
https://aws.amazon.com/marketplace/pp/B071H71L2Y

EC2 Launch p2.xlarge 


ssh -i XXX centos@ec2-XXX


sudo yum install screen htop R


R

set.seed(123)
n <- 100e6
m <- 1e6
d <- data.frame(x = sample(m, n, replace=TRUE), y = runif(n))
dm <- data.frame(x = sample(m))
write.table(d, file = "/tmp/d.csv", row.names = FALSE, col.names = FALSE, sep = ",")
write.table(dm, file = "/tmp/dm.csv", row.names = FALSE, col.names = FALSE, sep = ",")



[as root:]   ./cloud-init/per-once-bootstrap-mapd.sh


/raidStorage/installs/mapd-ce-3.0.0-20170507-7626e30-Linux-x86_64-render/bin/mapdql -p XXX


\timing

create table d(x int, y float);
create table dm(x int);

copy dm from '/tmp/dm.csv' with (header = 'false');
copy d from '/tmp/d.csv'  with (header = 'false');
--Loaded: 100000000 recs, Rejected: 0 recs in 52.022000 secs


select x, avg(y) as ym 
from d 
group by x
order by ym desc 
limit 5;
-- Execution time: 1342 ms, Total time: 1343 ms
-- Execution time: 1131 ms, Total time: 1132 ms

select count(*) as cnt 
from d
inner join dm on d.x = dm.x;
-- Execution time: 199 ms, Total time: 199 ms
-- Execution time: 83 ms, Total time: 83 ms






======================

On the above MapD is using CPU on the aggregation query and GPU only on the join.


According to MapD's @dwayneberry:

I ran your steps on my laptop (1 x 1060 GPU) and these are the times I got.
I have checked the code on a kepler machine (k80machine like you are running on AWS) and I notice the IR code is targeting the CPU for the AVG(float) query. My laptop is pascal so didn't see the CPU behaviour.
We have recently changed this so more operations will run on Float on kepler architecture. I will get that applied to this edition as soon as possible.

https://community.mapd.com/t/toy-benchmark-on-aws-cloud-gpu-usage/34/3


@dwayneberry's timing:

mapdql> select x, avg(y) as ym from d group by x order by ym desc limit 5;
Execution time: 464 ms, Total time: 465 ms
Execution time: 467 ms, Total time: 467 ms

mapdql> select count(*) as cnt 
..> from d
..> inner join dm on d.x = dm.x;

Execution time: 141 ms, Total time: 142 ms
Execution time: 67 ms, Total time: 67 ms
Execution time: 69 ms, Total time: 69 ms