> What happened when Apple wanted to CNC machine a million MacBook bodies a year? They bought 10k CNC machines to do it.
That can’t be right. 1M / 10K = 100 MacBooks per CNC machine per year. Surely it’s at least an order of magnitude or two off?
That can’t be right. 1M / 10K = 100 MacBooks per CNC machine per year. Surely it’s at least an order of magnitude or two off?
I however recall that they claimed in a keynote it took 10 hours to machine a single unibody MacBook, which would make the capacity for a single machine about 700 cases a year.
2k machines to make 200k laptop bodies for an initial sales rush over a month is 3.3 laptops per day per machine but each laptop uses multiple parts front and backplate for top and bottom. Assuming 4 major components per laptop and 5 different CNC machines per component you might be at say 3.3 * 4 * 5 = 66 parts per CNC machine per day. Which doesn’t sound that far off when they are discarding such a large percentage of the initial blank, though it’s likely some stages are much faster than others.
The equipment would be extremely expensive, but it’s also going to last for a long time.
The real problem with the point in the article is that it’s easy to get CNC work done
Edit: The tact time per casing is somewhere around 30 seconds (assuming 24/7 operations). Which is absolutely doable, not just for apple but for everyone with a demand of 1 million units per year for the same part.
Presumably they wanted to create rumors to improve peoples perception of quality. On the other hand it’s several fairly large complicated parts and their removing a lot of material from the blanks.
Also, most likely they don’t run them for a full year. They need 1mm Macbooks in a window of a few months. In which case, 10,000 machines making 100 macbooks each makes sense.
Edit: also, this is assuming every machine is the same machine doing the same thing. Which most likely is not (unless the unibodies are made in one go on a 5-axis mill)