danwells

This card is essentially the Radeon Vega 64 that shows up in the higher-end iMac Pro (there may be modest differences in clock speed, etc.). The iMac Pro version has 16 GB of RAM, just like the WX 9100. Unfortunately, Vega 64s available separately have 8 GB. Unless you really need the 16 GB for a very RAM hungry task, or you have an application that uses the WX series workstation drivers, a standard $600 Vega 64 in the same eGPU box should perform similarly. I'm not sure that the certified drivers issue even exists on Macs? On the PC side, the GPU manufacturers write more stable drives that enable specific features in CAD software and a few other pro applications (not including Photoshop and friends) that work only with the "workstation" variants of cards. Many of the Radeon WX and NVidia Quadro cards are just the gaming GPU, sometimes actually downclocked a bit, sometimes with more RAM or some other minor variation, sold for 2-3x the price with a minor firmware tweak that enables the workstation driver. At the high end, there are workstation cards that use different GPUs.

Unless you know you need the certified workstation driver for some particular application, just use a Vega 64...

The good news is that almost any USB-C cable should charge an iPhone. I don't know if pure data (no power at all) cables are even USB standard legal. Every cable I've ever seen will carry at least 15 watts. I think even a Thunderbolt cable in a non-Thunderbolt port does 15 watts.  Of course, there are probably some non-certified cables that won't carry power (but I've never seen one), and I think some USB-A to USB-C adapter cables might be 5 watts instead of 15.

The one obvious exception is so expensive that nobody is likely to try and use it as a phone charger. Thunderbolt supports optical cables with optical-to-electrical transceivers in  in the connector ends. You do see them as extra-long Thunderbolt cables, and that's obviously not going to carry power unless somebody's run a couple of copper wires alongside the optical fibers. I don't know if hybrid cables like that exist, but I do know that at least some optical Thunderbolt cables are pure optical and won't carry power by definition. The good news is that an optical Thunderbolt cable is at least $200, and most are more like $300-$500 so it's not the kind of thing that's going to be lying around in a drawer waiting to bite unwary people looking to charge their phone - it probably runs from your Mac Studio to a big noisy RAID in a closet...

One important caveat on products like this (I found this out the hard way on an earlier generation) is that any cloud backup solution that will back up a RAID, but not a NAS (Backblaze is one, I suspect the same restriction applies to many others) considers these things to be NASes. So does any piece of software that doesn't like network drives - even when you're connected to the Thunderbolt port, it's considered a network drive. 

It's not really a Thunderbolt/NAS hybrid drive - it's a straight-up NAS that issuing Thunderbolt's ability to serve as a fast network interface - but it's a NAS from a management and software standpoint, and that's important to realize...

The headline is misleading, looking at the graph... Since 2020, we've seen the same pattern each year. Samsung outsells Apple three quarters every year, then Apple takes a big lead in calendar Q4 each year - that's the iPhone intro quarter. and the holiday quarter in a significant part of the world. The real news here is that Apple dang near caught Samsung in Q1 2023. If you look at Q1 2020, 2021 or 2022, the gap between Samsung and Apple is much greater. Apple always has a huge sales bump in Q4, and Samsung's lowest sales percentage is always in Q4 (I'm guessing they move more units in Q4 than any other quarter, which this graph doesn't comment on) , but their PERCENTAGE is at its lowest - because they're competing against brand-new iPhones with 6 month old Galaxies. Q4 has huge overall sales because of the holiday quarter in North America, Europe and some other regions...

The other thing to remember about any Apple/Samsung unit sales comparison is that pretty much every iPhone competes with the TOP of Samsung's line. Everything except the SE is clearly competing with the S series Galaxies (and the less common Z Flip). Even the SE is competitive with the top of the Galaxy A line, which extends to price points significantly below the iPhone SE. Samsung makes even lower-end phones (well below the iPhone SE) in the Galaxy A0xx, F and M series that don't commonly appear in the US market, but are responsible for a lot of sales volume elsewhere in the world. This is slightly complicated by the VERY expensive Z Fold, but my best guess is that sales of that phone are minimal compared to the iPhone Pro Max (which competes with the S23 Ultra). 

At first, I thought "this makes no sense" - two high end laptops that almost end up in the mobile workstation market and a consumer desktop that sells a fraction of what the consumer laptops do. The iMac is the next machine up for an update, but it pretty much shares its innards with the 13" MBP or the Mini (ignoring the M2 Pro Mini) - it just didn't get those innards updated the last time around (why?).

Then I realized that it makes sense in one specific situation... What if all three chips are ready, but supply is constrained by yield? The pro laptops have very, very high profit per machine (their percentage margin is typical of an Apple product, or even slightly low - but their high selling prices mean a lot of dollars per machine). Building an M3 Max and sticking it in a $3000-$6000 workstation is a lot more profit per chip than selling a base M3 in an iPad or a MacBook Air. The M3 Max is also larger, and I don't know how profit per wafer compares... 

The iMac is old, and it has been attracting attention in the Mac press for how old it is, so it's a good machine to update just for the PR. Why not update its innards-mates with it? If there are supply constraints, it shares its innards with the most popular Macs of them all...

 If Apple updated the Airs, it's possible that the limited chip supply would simply shoot availability dates into March right away. Update the slower-selling iMac and people might actually get one in a reasonable time frame. Update the 14" and 16" Pro and you've made some very visible creative users happy, and sold a bunch of high profit machines

A digital ad tax is about the best tax I've ever heard of.

1.) Taxes reduce consumption of whatever they're applied to (this is why cigarette taxes are high - they're equal parts anti-smoking measure and revenue generator). An ad is a nearly unique product of generally negative value to the consumer (the consumer would prefer to have FEWER of them). Special cases like Super Bowl commercials which are funny enough that consumers seek them out notwithstanding, no recipient will pay for ads, and many people will pay to get rid of them. Netflix WITHOUT ads costs more than Netflix WITH ads, ad-blockers cost money, websites offer paid subscriptions to get rid of ads. Maryland has made the brilliant decision to tax something PEOPLE DESPISE.

2.) The tax falls on big companies that can easily afford it, and they don't have an easy way to pass it on. They can't charge end users more for ads, because end users don't pay for ads, they sometimes pay to avoid them. Any tax that sticks to big companies is useful - if you add a tax to most products, even if the producer is responsible for the tax, they pass it on (if you charge a value-added tax that manufacturers have to pay, they add it to the price of the product, passing it to consumers). The best way to keep a tax applied to the intended target is to target a product that makes passing the tax on nearly impossible - and taxing a product with a negative value is the best possible way to do that.

Moof! I'm wondering if this has to do with Apple spinning out in house software to avoid anti-trust issues? Many of the potential (and some actual?) lawsuits around the App Store have to do with Apple both running the store asnd offering products in it. Would just taking the products external (everything from iWork (Pages, Numbers, Keynote) to Final Cut Pro) solve the problem, or would they have to spin off the services (Apple Music) as well?

 I can't imagine the internal software has any great impact on revenue - they wouldn't mind losing Final Cut, FileMaker is already semi-external, and most of the rest are free (or bundled with the OS). iWork could even be a modest new revenue source to Claris if unbundled. They don't have to unbundle Mail or Safari, because Microsoft bundles Outlook Express and Edge, and Google bundles Chrome and a Gmail app - browsers and mail are accepted parts of the OS.

If they managed to spin off the software to avoid challenges to the much more valuable services, it's almost certainly worth it to them...

Apple is historically better than most about letting you turn off some of the targeting. Hopefully this can also be turned off. Of course the problem is that gambling companies (or other deep-pocketed spammers) will get around everything by simply buying so many slots that everyone gets it. It's like the horrible flyers pitching cheap cigars and really bad idea gold investments that spam every physical mailbox. They get the cost to print and deliver down so low that they can afford to annoy nearly everybody to catch a few suckers.

At least some part of the problem is Apple consistently taking every improvement as "thinner and lighter, without losing too much performance", rather than sometimes keeping size and weight similar with a major performance increase. How many people would like a Skylake/Retina MBA better than the MacBook? Yes, they kept the MBA around, but using old chips and the only non Retina display in the Mac line.

I'm afraid this is where they're going with GPUs - sacrificing performance for battery life (even, astonishingly, on the iMac, which doesn't run on batteries :-) ). There are certainly lower-end machines that don't need discrete GPUs, but I think Apple needs four total discrete GPU machines (three of which exist), and two of them need to take gaming into account (one argument people are making is "Apple uses AMD for compute, which is great for photo and video editing, but not so much for games"). External graphics are a real option if Apple supports them well especially with increasing Thunderbolt speeds. 

1.) Mac Pro - expensive, generally not purchased for gaming - GPUs need to be selected for very high performance in creative pro applications. As long as it's kept updated, the choice of workstation GPUs is fine... The biggest problem with the "trashcan" Mac Pro is that it's too elegant! An internal HD capacity of 0 in a machine like that? It should have a couple of 3.5" bays (as well as the SSD). 

2.)iMac - 27" needs a discrete GPU (it's a 5k screen!!!) 21" should probably have the option (there's a very good reason to make a no discrete GPU model, as a reasonably priced home/student desktop). These ARE gamer-friendly machines, so nVidia options probably make a lot of sense. Why not use desktop GPUs for better price/performance - who cares about shaving a bit of weight off a desktop? A thicker machine could provide the cooling without being much louder... Do laptop parts actually offer better power/performance, or are they just drawing less power for less performance? I can see not offering the multi-hundred watt super-performance gaming GPUs for a mixture of environmental and cooling concerns, but it should have a lot of GPU options, not all of which are laptop parts. Could an external GPU theoretically drive the INTERNAL display? If so, what would it take to make that work? A Thunderbolt 3 port would connect the GPU (almost certain to exist on the next model) - could it feed the display signal back in on the same port? If not, what would it need for a second attachment? 

3.) MacBook Pro - one discrete model (15" only) is fine, laptop GPU is (of course) fine. In my opinion, this is mostly a creative pro machine, not a gaming machine. Maybe it wants one higher-end GPU option, but what I primarily care about is that it doesn't LOSE its discrete GPU option. Thunderbolt 3 opens the possibility of external GPUs with excellent performance, and the question of the internal display is somewhat less important - the internal isn't the incredible screen the 5K iMac has, and I'd imagine either a serious gamer or a creative pro would want an external display when they went to the trouble of plugging in external graphics. One really interesting accessory Apple could make for the MacBook Pros is a 5K Thunderbolt Display with its own GPU? Is there any reason the GPU couldn't be in the display with a fast Thunderbolt connection?

4.) Midline desktop Mac. Apple NEEDS to make a screen less Mac in between the Mini and the Pro, especially with the Mini moving down the line.  It could be a quad core Mini with either internal discrete graphics or enough Thunderbolt ports (on multiple buses - you don't want a GPU contending with storage)  to make external graphics a good option. I'd like to see a "Mac Midi" with space for a PCI-e card and a couple of drives, but it'll never pass muster with Jony Ive. Barring a Mac Midi, a quad core Mini with plenty of Thunderbolt would serve much the same purpose - some enterprising company might even make a housing that holds a GPU and some drives and matches the Mini. 

I hope that, by the time Sierra is released, this will be extended to the iPhone as well, rather than used as a ploy to sell the Apple Watch. It would be perfectly easy to do the same thing with an iPhone, which most Mac owners have (there are so many more iPhones out there than Macs that I'd assume the Mac/Android combination is somewhat rare (or owning a relatively expensive recent Mac but no smartphone at all)). Of course, you can still unlock your Mac the old way if you prefer Android or don't have a smartphone - but MANY more people could use auto-unlock if it worked with the iPhone.