ipad speeds

I don’t think we’ve had the iPad Air 2 on here

6 seconds flat (vs 7.7 with Air 1)

Wow! @Ignatz, you won: the fastest to upgrade an iPad :wink:

Let us know what the 3D performance is like, the GPU is supposed to be very good

This is only a rough measure, but my Warships game on an iPad 3 could handle about 3 ships before FPS dropped into the 40’s.

The Air 2 can handle about 25 ships…

ipad2 55.59 secs

iPad Mini 62.46…
Time to upgrade to an iPad Mini 4!

Here is some code to test graphic speed. It adds mesh rectangles (to a single mesh) until speed drops below 60 FPS, then removes them until it gets back up again. Wait for the Count parameter to stablilise.

iPad Air 2 = 1410 approx

function setup()
    parameter.text("Count")
    m=mesh()
    m.texture="Small World:Icon"
    w,h=110,110
    FPS=60
end

function draw()
    background(200)
    FPS=FPS*0.9+0.1/DeltaTime
    if FPS>=59 then
        m:addRect(math.random(1,WIDTH),math.random(1,HEIGHT),w,h)
    elseif ElapsedTime>3 and FPS<59 then
        if m.size>6 then m:resize(m.size-6) end
    end
    m:draw()
    Count=m.size/6
end

@Ignatz on an Ipad mini 4 - I’m getting around 830.

As for the original test 6.11245 seconds.

After looking online it looks like the mini 4 has the Apple A9 processor in it where as the Air 2 has the A9X which includes a better class of graphics processor.

Either way it’s still a MASSIVE leap in performance from my old bulky 3rd gen retina iPad (and you can tell just by seeing how much more performant it is on normal day to day use)

I’m surprised there is so much difference in graphic speed, because the specs suggest they are pretty close.

only 640 on ipad air 128 go with iOS 8.3 -
with ipad pro it will be 2100 lol

Wouldn’t it be a better test if the rects were the same size?

On my iPad Air, the range seems to vary a lot to get an accurate value. I estimate around 660 in portrait mode. If I rotate my iPad to landscape, the value increases to around 760 then back to around 660 in portrait again.

@Ignatz from what I’ve read the Mini 4 has the same graphics processor as the iPhone 6 / 6+, the Air 2 has an upgraded one - I think one of the reasons is due to the battery size, the Air 2 has a bigger battery that can handle the extra drainage.

I noticed different values for portrait and landscape orientations that I mentioned above, but if I start the program in landscape mode my count is around 640. If I rotate the iPad to portrait, then the count goes to about 840 which is higher than my original values. Not sure why the values differ that much as a result of rotating and which position the iPad is in when the program is started in.

@Ignatz I added a Count_average to your program to make it easier to see the Count value. Tap the screen every few seconds to recalculate the average until it settles down.

function setup()
    print("Tap the screen to recalculate the Count_average.")
    parameter.text("Count")
    parameter.text("Count_average")
    m=mesh()
    m.texture="Small World:Icon"
    w,h=110,110
    FPS=60
    Count=0
    cnt=0
    tot=0
end

function draw()
    background(200)    
    FPS=FPS*0.9+0.1/DeltaTime
    if FPS>=59 then
        m:addRect(math.random(1,WIDTH),math.random(1,HEIGHT),w,h)
    elseif ElapsedTime>3 and FPS<59 then
        if m.size>6 then m:resize(m.size-6) end
    end
    m:draw()
    Count=m.size/6
    tot=tot+1
    cnt=cnt+Count
    Count_average=cnt/tot//1
end

function touched(t)
    cnt=0
    tot=0    
end

Would not be also interesting to check iPad’s capabilities by generating physics objects up to Codea’s explosion? ~O)

My iPad Mini got about 500, a lot higher than what I expected considering the last speed test

ipad3 61s while playing music ios9

iPad Pro about 3.3 seconds

Ignatz’s program runs about 3760 on iPad Pro