dewme

appleinsideruser said:

dewme said:

I’m confused by the article’s reference to product failure rates following a bell curve. In all of my years working with electrical and electronics based product development, including working closely with reliability engineers and component engineers, the failure rates of such products and systems have followed a bathtub curve model, not a bell curve. 

What is your source of this bell curve failure rate model? It seems intuitively incorrect simply from a standpoint that nothing that I can think of would cause a reduction in failure rate as a product gets older and accumulates more runtime and cycles. 

The bathtub failure rate curve reflects the fact that electronic components and systems built using these components tend to have an early higher failure rate due to “infant mortality” failures, including manufacturing defects, then dip to a much lower failure rate for a very long period of time, but then increase again as components wear out over extended run times power cycles, component degradation, environmental stress, heating and cooling effects, etc. 

If the batteries in smartphones degrade with an increasing number of charge-discharge cycles, exposure to thermals, etc., I would expect the failure rate for batteries to be monotonically increasing with time. What would cause a battery’s failure rate to decrease at some point in the future?

The only thing I can think of is that the author is trying to describe the total number of reported failures over a product’s lifetime. As the product is removed from service later in its lifetime of course the number of failures that get reported would decrease because there are far fewer of the products still in use. But that’s not the definition of failure rate within the scope of reliability engineering. 

I feel the distinction is that bath-tub curve is for component failures (high early failures, then a great mid-life, followed by old-age problems). The Normal distribution is useful for variations if the performance across a population — battery aging isn't really a failure issue, it's that different samples/batches work better than others.

Okay. You’re reiterating what I said in your first sentence. Battery aging, I.e., wear-out, is already considered in the bathtub shaped failure rate curve. That’s all part of what you call old-age problems and it is a failure mode. Failures are failures. Keep in mind that even during the flatter part of the bathtub failure curve failures are still occurring, but at a lower rate than the higher parts of the curve. Product failures that occur during the flatter part of the curve are potentially due to what you call bad-batch and not-quite as good as others issues.

Keep in mind that every unit has its own failure curve. The failure rate curves that are produced by reliability analysis are done with large sample sizes and by applying statistical methodology. If you wanted to talk about bell curves and normal distributions you could probably say that the population of products that conform to a product’s predicted “bathtub curve” failure rate model is based on a bell curve model, where the majority tend to follow the predicted failure rate curves and there is a distribution and outliers on both sides that don’t follow the predictions at all, both better and worse.

Getting back to iPhone 14 battery drain, my iPhone 14 Pro Max is still at 100% after about 9 months of use. I don’t do anything special to affect battery life and leave it all up to the iPhone’s charging circuitry. I’ve never tried to babysit my products and rely on them to do their jobs as designed. I very rarely unplug or turn off any of my computers, PCS and Macs, but just let them sleep. Some of them have essentially been running for more than a decade with occasional cleaning and making sure airflow to cooling fans and ports is okay.

If I was investigating why an iPhone or Mac was losing battery capacity in an accelerated manner I’d take a look at what’s running and the CPU load, number of active threads, etc., kind of things. Apple is always adding more background processes for all of the new features, e.g., crash detection, so you’d have to at least consider that the device is being asked to do more when you think it’s idle. Same thing with notifications and constant updates for social media apps like TikTok, Facebook, Twitter, etc. I’d want to know what the heck is going on to consume more CPU cycles and battery capacity, if indeed that’s the case.

I have noticed an infrequent issue with the latest iOS release version on my iPhone 14 Pro Max. What I’ve noticed a few times is my phone will go into a state where it jacks up the screen brightness beyond my selected level and run the CPU at a rate that causes the phone to heat up considerably. If it’s in my pocket will notice it immediately, but if it’s just sitting on a desk or counter I may not notice it for a while. It seems like it is in a thermal runaway state because the normally cool iPhone is unusually hot. If I restart the phone everything is fine.

I think there is a bug in the latest iOS version that is causing the CPU to get into a hot spin cycle. If you don’t notice it because you don’t have your phone on you where you can feel it, this would very likely drain the battery much faster than normal and heat up the phone unnecessarily, which is also bad for batter life.

Unfortunately I can’t get the phone into this state in a repeatable manner or else I’d submit a bug report. It has happened when running Maps at least two times, but it’s also happened outside of Maps. It’s only appeared in the last few point iOS updates. 

I don’t know why so many people are getting all bent out of shape about Apple potentially including a charging-only low data rate UBC-C cable in the box. Does anyone expect Apple to include an expensive TB or DP compatible/certified USB-C without passing along the extra cost to consumers?

Take a look at the comments section of 4K or 5K monitors that have USB-C with Display Port protocol. The display vendor, say Dell, will usually include a compatible USB-C cable in the box. Problem is, it’s often too short for certain mounting scenarios. So buyers go out and buy a longer inexpensive USB-C cable and are shocked to see that it doesn’t work. And they complain about it. 

anome said:

darkvader said:

Just a reminder that VMware Fusion can be had for free:  https://customerconnect.vmware.com/en/evalcenter?p=fusion-player-personal-13

That's the Fusion Player. Does that let you create Windows VMs?

My renewal just went through. I have to make sure I've got the update.

Yes, VMware Fusion lets you create Windows VMs. On Intel Macs you can also create Linux VMs. On M1/M2 Macs you are limited to ARM64 versions of Windows, which is also true of Parallels on Apple Silicon.

The “Player” version of VMware Fusion, unlike the Windows “Player” product version, has the same feature set as the standard version of Fusion, including snapshots. The only limitation is that it’s limited to non-commercial use only. There is also a paid Fusion Pro version that has additional features. 

If you prefer Parallels to VMware Fusion, and if you qualify for the free version (honor system), it costs you nothing other than your time to give VMware Fusion Player a look. I previously paid the upgrade price for the inevitable Fusion yearly upgrade linked to every new macOS version, but since I was using it for personal use I was able to move to VMware Fusion Player with zero loss of functionality and no longer have to pay the upgrade price.

So why use Parallels? Well, Fusion Player worked out great for me because I was already invested in VMware on both the Mac and PC. If you are invested in Parallels and if it’s working out for you the yearly upgrade cost is probably still well worth it, especially if you rely on it for your business or personal productivity. It’s a great product with a great support team behind it. You know that you’re getting a product that has a dedicated team behind it.

VMware Fusion is an outstanding product, but if they are able to give it away for free to a lot of users they are likely making the majority of their money on other parts of their business to be able to afford to offer it for free. While I have never had a situation that made me think VMware wasn’t completely dedicated to supporting and continuing support for Fusion or Fusion Player, common sense tells me that if their business ever has to cut back to reduce costs, their products that have no direct contribution to sales numbers are most at risk of being abandoned. This may be overly simplistic, but it’s reasonable.

The transition to USB-C for device charging should not be significantly different or any more painful than the transition from the 30-pin connector to Lightning. It’s not a big deal.