Apple iWatch wafer demand estimate

[ippisl]

My estimate goes roughly like this:

The new linkit chip by mediatek that is aimed for watches is ~30mm2 @ 28nm. Apple will surely do it in 20nm/14nm because power is critical ,so there's no reason why it should be more than 30mm2.

The estimates sales for 2015 vary, but the come around 30million units. that's 900 million mm2 , and add less optimal yield, we'll get somewhere around 1.3 billion mm2. divide by wafer(77000 mm2) = ~16.5K wafers or ~1.4k/month. And with pricing(i'm guessing) around $6000/wafer , that comes around to $96million order. Nice , but nothing to write home about.

I'm curious what you guys think will be the demand for wafers for the new apple watch?

 

[lefty]

If power is critical, shouldn't they use 28ULP (ultra-low power)? Though I'm not sure if it was available on time for launch.

 

[ippisl]

I mentioned the mediatek process generally just to get an estimate density, it makes perfect sense for them to use 28ulp.

 

[Li Yisuo]

Compared with 28ulp, 28fdsoi was ready earlier. Price spike is tolerable compared with 20nm and 16/14 capacity limit. Besides Apple watch is considered to be high end wearable device. The operation cost will focus on fashion app design, conceptual events and so on. 50mil units ~ 30k/mon
So let us compare 28ulp vs fdsoi process. ULP is in general great. However I believe it needs higher Vdd and hard to boost performance. High end watch may needs some fancy power hungry functions sometimes. Otherwise, how do we differentiate it from the average.

I like this watch.

 

[Daniel Nenni]

My estimate goes roughly like this:

The new linkit chip by mediatek that is aimed for watches is ~30mm2 @ 28nm. Apple will surely do it in 20nm/14nm because power is critical ,so there's no reason why it should be more than 30mm2.

The estimates sales for 2015 vary, but the come around 30million units. that's 900 million mm2 , and add less optimal yield, we'll get somewhere around 1.3 billion mm2. divide by wafer(77000 mm2) = ~16.5K wafers or ~1.4k/month. And with pricing(i'm guessing) around $6000/wafer , that comes around to $96million order. Nice , but nothing to write home about.

I'm curious what you guys think will be the demand for wafers for the new apple watch?

My guess is 5k wafers per month @ Samsung 28nm LP. The same process used for the Apple A7. I'm interested in buying an iWatch but I'm not sure how wide of an appeal it will have. It really needs health applications to go mainstream.

 

[astilo]

My guess is 5k wafers per month @ Samsung 28nm LP. The same process used for the Apple A7. I'm interested in buying an iWatch but I'm not sure how wide of an appeal it will have. It really needs health applications to go mainstream.

Apple Watch may be powered by Samsung processor | ZDNet

Daniel, my fear is that the battery won't even last a full day, if you really want to use the watch as a smart device. If not, just buy a proper wristwatch and keep the iphone with you

 

[ippisl]

Daniel ,does it mean you think the chip will be around 105mm2 ? isn't that pretty huge for something like a watch , with a small screen , relatively low resolution, and simpler/fewer communication interfaces ?

 

[Daniel Nenni]

Battery is definitely an issue for me. I will charge my devices at night without a problem but I will not hug walls.

I didn't do the math on the die area. My wafer number came from a friend who should know. But again, there is no telling how many Apple will sell and I highly doubt Samsung will tell so all we can do is guesstimate.

 

[sophiw75]

You can catch all detail information of upcoming Apple Watch / iWatch Specifications, Price and Release date using this.