Hacker News Reader

Comment by sudhirj | Hacker News Reader

It's interesting that development of the iMac Pro was already underway when Apple and pro users (mostly developers) had that big showdown after the touch bar came out. The touch bar seemed like the final slap in the face to the pro crowd, but this was already in the works then.

Every pro review I've seen so far has been that's it's excellent - not because they're fawning over this or that, but just that's it's a solid option with a great screen at what's actually a reasonable price for the components inside. If you think the price is unreasonable, you're probably not the sales target for this machine - the ones who really need this, like in the case of this reviewer, are buying three at a time - just because.

It's interesting because upgradability doesn't seem like a very big concern to anyone buying this - they're happy with that they're getting at the price they're paying. It almost seems like Apple needn't have promised everyone a new modular Mac Pro. If they'd just announced / launched this machine alongside the pointless MacBook Pro update, everyone would have been happy and with no accusations of taking the name of the Pro in vain.

View on HN

I'm not so sure the reviews would have been _quite_ so positive if they hadn't also announced they're releasing a 'modular' traditional Mac Pro.

I think the knowledge that a new 'modular' Mac Pro is coming gives a different context to this machine. I think people are framing this on the idea that: if you want a quiet, good looking machine with a great screen and the other 'benefits' of an all in one (and don't mind the trade offs), then it's good to have this option. If the trade offs don't work for you, then wait for the more traditional Mac Pro.

If it had been released without any mention of another option coming for pro users, then it would have been framed as "if you need workstation grade hardware running mac os, then your only choice is now an expensive, non-upgradeable all-in-one which has traded off performance for quietness and for other thermal constraints".

I'm definitely in the category you describe: owing to ancient personal prejudices, I consider the idea of having one's computer integrated with the screen to be an utter abomination. I have trouble suffering the trend towards laptops and notebooks precisely for this reason, let alone the rise of the integrated all-in-one.

Knowledge of the coming Mac Pro allows me to view this device with levity because I can view it as a herald of things to come in a more satisfying format; had I not known of the ‘imminent’ (end of 2018 or beginning of 2019) Mac Pro, I would have been utterly aghast at the idea of Apple pushing me towards the integrated solution.

I tend to cram my computers with cards (not graphics cards, but things like SDRs and FPGAs) but what really does it for me is having the monitor built-in: I'm used to multiple monitor desktops, and I like both screens to be absolutely identical make and model. That clearly isn't possible when you have an all-in-one computer with no available ‘identical’ screen.

Also, my computer belongs beneath my desk, and my monitors on top of it. It's just the preordained way. All else is heresy. ;)

Also, computers and monitors have vastly different life cycles; the former become obsolete much faster than the latter. Binding them together physically forces you to replace one or the other either too soon or too late for absolutely no benefit.

At least the displays Apple ships are very very good. Best in the industry.

I agree. I was the very happy owner of an Apple Cinema Display that served me well for many, many years, and was still in perfect working order when I had to abandon it in favor of an ASUS display because I needed more resolution and Apple wasn't selling standalone displays any more :-(

I want to be an Apple fan. I really do. But they are making it impossible for me nowadays.

Apple has mentioned that the next Mac Pro will also have a matching display.

I'm fine with a typical 4k display (or better). The panels of the better ones seem to come from LG anyway and are widely available on the market. I wonder if Apple would be selling a good quality 5k monitor with TB3, or if they go for a high-end 8k screen. Personally I'm not in the business for the highest end. LG has announced a bunch of TB3 screens, where the 4k 32" version would be sufficient for me. But LG lately has problems bringing their better screens in enough quantities to the market. Last years 32" 4k screen is still not widely available here in Germany.

I actually think that the 27" 5K display in the iMac Pro is going to age quicker than the internal components. I suspect that ~32" 8K displays will be mainstream in a year or two.

> I suspect that ~32" 8K displays will be mainstream in a year or two.

8K would require new HDMI cables and ports, would it not? I'm pretty sure that many people are still stuck with legacy hardware in this regard.

Current 8k displays use 2x DP connections just like early 4k displays.

Given your use case is a Mac, any Mac, running OSX even an option? I also use FPGA cards and what not but I do so in a separate Linux box because all the tools I need to for the setup only have native Linux versions. And if all you need is a screen running OSX which remotes into other things, isn’t that what the iMac is?

What SDR are you putting on a PCIe bus?

LimeSDR

I wish they would have left the Xeon and Pro GPU components to the new Mac Pro instead of putting them in this. It would’ve been nice just to have top of the line consumer grade hardware available in a Mac instead.

That and the amount of time they’re taking on this modular one is just nuts. All people want is a simple tower. They could just bring the damn cheese grater back.

I agree so, so much with what you wrote.

I only have one teeny tiny nit and I'm only mentioning it because I think it's a common misconception:

    if you want a **quiet,** good looking machine with a 
    great screen and the other 'benefits' of an all 
    in one (and don't mind the trade offs)

I don't think "quiet" needs to be the domain of all-in-ones; with a $60 CPU cooler from NewEgg one can have a ~4ghz i5/i7 PC that's nearly silent. The old cheese grater Mac Pros were noisy, but that really doesn't need to be the case, especially if Apple would just release a damn desktop i5/i7 based Mac Pro.

Which might have been okay, because iirc the complaints weren't really 'we want upgradable pro machines' - they were more 'just build a pro machine'. Other than upgradability of core components, I'm not sure what the Mac Pro can possibly bring to the table now. And it seems like there aren't that many folks who care about upgradability alone.

because iirc the complaints weren't really 'we want upgradable pro machines'

That is totally untrue. The places I was complaints were all about upgradability. That's why many people still get the previous generation of Mac Pro's and upgrade the processors, graphics cards and add SSDs. My main machine at work is one of these, and feels just as fast as my Mac Pro 6,1 (the trashcan version) at home.

> Apple and pro users (mostly developers) had that big showdown after the touch bar came out. Every pro review I've seen so far has been that's it's excellent

Almost every Touch Bar MBP review has also been excellent. Sales of the Touch Bar MBPs surpassed the sales of almost all other competitors combined, in less than 1 week [0] (annnnd here comes the expected downplaying of why it sold so much, offering every convoluted reason but the possibility that it is actually, in fact, a good product.)

"Pro" users and Apple did not have a "big showdown." It's just an obnoxiously loud minority, with their puerile 4chan'ish dismissals like "Emoji keyboard!!1", who had never even really used a Touch Bar MBP, that were clogging up a few forums. People who barely exhaust even 16 GB in their daily work fake-whining about the inability to upgrade to 32 GB, even though almost no viable competing laptop supported that either at the time.

[0] https://9to5mac.com/2016/11/09/2016-macbook-pro-sales/

[1] https://9to5mac.com/2016/11/01/2016-macbook-pro-ssd/

> Sales of the Touch Bar MBPs surpassed the sales

It would be fair to mention, that if you wanted the not-the-weakest-CPU or more than 2 USB ports, you had to get the touch bar version.

It would be fair to mention, if any consumers ( which far out number professionals ) that wanted a Macbook but dont want a silly 12" screen would need to get a Macbook Pro.

And just to Mention, the report does not spilt between Touch Bar and Non Touch Bar. And we already know most premium prices notebook goes to Apple. That means Apple would have sold just as much if they didn't had the TouchBar.

There is Data, and how you interpret that data.

Every other laptop manufacturer on earth had the opportunity to put together a computer with the not-weakest-cpu and more than USB ports without a touchbar.

isn't the only viable competition to a MacBook pro is a Lenovo Carbon X1...which is equally expensive ?

The XPS 15/13 is viable competition imho.

Not with MacOS.

I bought a Touch Bar MBP because it was finally time to retire my 2011 MBP and I think it's the best computer I've ever owned, period. I do video editing and media production (podcasts, company videos, etc.) and I feel like I put this thing though its paces regularly. On top of that, although I mainly have it hooked up to a separate display, keyboard, and mouse, the Touch Bar is amazing when I'm editing on the go. I kinda wish that my regular keyboard had it because I can scrub through all my media without having to jump back and forth between keyboard and mouse. On top of that, I installed Better Touch Tool and I can customize the Touch Bar to do pretty much anything I want and that's made it infinitely more usable. I also do app development and web development for our shop so having the terminal and being able to run node locally is wonderful.

I think the people that complained about it are people that either 1) didn't actually use one at all, 2) are 100% resistant to anything changing in their workflow, even if it might improve it in other ways (I'm looking at you ESC key nihilists in VIM), or 3) would never buy a Mac anyways because they like to tinker and/or upgrade their PCs regularly.

Frankly, I don't understand when these types of people get butthurt about a product that they will never buy and don't intend to buy just because others might get them. Just don't buy the damn thing and get on with your life. For those that do buy it because it fits their workflow and lifestyle, I think the damn thing is amazing.

The utility of Better Touch Tool in customizing the Touch Bar to a "pro" or personalized workflow can't be overstated:

https://www.boastr.net/bettertouchtool-touch-bar-customizati...

https://medium.com/productivity-freak/what-if-you-could-real...

It's a shame BTT won the mindshare, and Jitouch devs abandoned it. BTT still doesn't have the ergonomics Jitouch had ~2 or 3 OS versions ago, but it has the userbase.

Better touch tool is my favorite app purchase to date. I'm considering setting up a yearly donation because of how useful it has been for me.

no viable competing laptop supported that either at the time

But, so what? You shop at Apple if you want better than the competition.

I work with multi-Tera datasets, and my use cases are common enough these days. Anyone doing CAD, bio, finance, video, etc etc etc would easily find uses for 32G or even more.

but just that's it's a solid option with a great screen at what's actually a reasonable price for the components inside

PC Gamer did a price comparison that tried to match the components as closely as possible and found that the result is barely cheaper than the iMac Pro:

http://www.pcgamer.com/apples-new-imac-pro-costs-5000-but-is...

I am not in this market segment, but if you are doing something like high-end video editing, the iMac Pro is competitively priced and (probably) has a great build quality compared to systems in the competing in this segment.

I hope that they won't forget some other market segments. It would be a great to have a new Mac Mini in the 700-1000 Euro range.

Kind of suspect Apple will drop the dustbin Mac Pro form factor, instead release Mac mini Pro.

They already announced that the new Mac Pro is a ground up redesign and with serviceability in mind. So you don’t have to suspect anything.

I'd buy that. A sleeper Mac Mini would be a perfect machine in my books.

I'd be happy to see something in the middle.

Yes but what happens 2 years down the line? That’s an important question for professional devices.

The iMac Pro looks like a great machine. But it’s definitely not comparable to a Mac Pro (although the last Mac Pro also wasn’t really a Mac Pro).

A $650 motherboard? $242 PSU?

They obviously weren't trying very hard.

$5000 is a lot of money to pour into one PC. Even if you are spending $1300 of that on a monitor.

They under-balled it if anything. They used a $400 Broadwell Xeon, while the base iMac Pro has a Skylake Xeon W. The chip is a custom Apple SKU, but the closest is a $1,100 Xeon W 2145.

$650 is steep for a motherboard, but it has dual 10 gig-e ports and supports registered ECC. A SuperMicro motherboard with the Xeon W support will run about $350 (they’re not out yet) and an X550AT2 NIC is about $200 on eBay.

Slightly less: the iMac Pro uses the ~$100 AQC107 whereas the X99 motherboard chosen uses the more expensive X550AT2 because Aquantia only launched their chips this year.

But yeah, the Xeon they chose is quite a bit slower than the 8-core Xeon-W; the Broadwell equivalent was really the $1100 E5-1660 v4.

So -$200 for motherboard, +$700 for CPU, and -$100 for PSU if you're feeling generous. That's still $5k.

* Video Card: Zotac - GeForce GTX 1080 Ti 11GB Founders Edition ($678.99) whereas the actual gpu used in the mac is $399

* Power Supply: SeaSonic - PRIME Titanium 1000W 80+ Titanium Certified Fully-Modular ATX Power Supply ($242.89)whereas they could have gotten a fine psu for $99

* Operating System: Microsoft - Windows 7 Professional ($134.99) How about not and say we did.

* Monitor: LG - UltraFine 5K Display 60Hz Monitor ($1299.95) whereas one can probably get a really nice 4K monitor for 400-500. Some can spend the extra funds on another 4K monitor, others you know already have monitors.

* CPU Cooler: NZXT - Kraken X62 Liquid CPU Cooler ($158.99) Did I miss the part where the mac is liquid cooled why is this here?

* CPU: Intel - Xeon E5-2620 V4 2.1GHz 8-Core Processor ($408.99) This is an amazingly poor choice while relatively expensive its performance is decidedly mediocre being an older generation.

Motherboard: Asus - X99-E-10G WS SSI CEB LGA2011-3 Motherboard ($649.00) Expensive not primarily because its awesome but because its xeon.

The $1059 could have been better spent on this motherboard http://www.asrock.com/MB/AMD/X399%20Taichi/index.asp#Specifi... note supports dual gb lan and ecc, $340

this cpu

https://www.newegg.com/Product/Product.aspx?Item=N82E1681911... $535 and about $100 for a single port 10G lan

Total $975 between motherboard and cpu

All in all a machine could be built that is nearly equivalent for about 3544. Given the extra $1500 one could buy a second 27" 4k monitor upgrade to 16 cores and buy a second gpu.

Pretending mac is comparable in price to pc remains silly.

I can’t soeak specifics about the PSU but to get a server class intel motherboard (Xeon support, ECC ram), you really do have to spend quite a bit more than for the equivalent consumer board. Even if you reduce the cost of the motherboard to 200, and the psu to 150, you’re still not saving a huge amount off 5k, and you’re not getting an all in one

Now if only they also made a laptop with ECC RAM.

I'd gladly pay 1k more to have a stable system.

It's your system really that unstable?

If so, are you sure it's the RAM?

Systems without ECC do crash, have transient "weird issues" and corrupt data significantly more.

If stability matters to you, you should really demand ECC for any system you purchase. It's the best bang for buck to that end.

> ECC ram

I'm not sure how Apple has people convinced they need ECC. It's practically worthless outside of very specific workloads. A workstation user does not fill several gigabytes of RAM with terse important data, query it millions of times/day, and do so 24/7 all year long.

I've had ECC RAM once, it failed and ECC caught it before it could damage any documents. ECC memory also keeps a log of how many corrections it's made. In a long running machine that gets hot there _are_ bit flips and they _are_ significant. Especially if you're spent days working on a video file and don't have the technical skills to fix it in a hex editor.

I've also had memory issues on non-ECC machines, it's a lot harder to diagnose and much more frustrating. ECC memory should be standard in modern machines IMO.

> Especially if you're spent days working on a video file

And didn't keep a backup. Most serious software will make incremental copies to avoid that problem too.

Seriously, hardware failures are just as likely.

ECC is useful, but isn't worth it for the price unless you actually can't be fault tolerant in any way.

Backups and ECC address completely different problems. Backup systems will happily store corrupted files into your backup disks. ZFS-style filesystem level checksums won’t catch many problems either, because they only check to see if what you wrote is the same as what was read back. If the data was corrupted in memory, it’ll happily pass you back corrupted data.

it only takes one bit flip in the FS cache, which is what your 'free' RAM will be used for, to make your day miserable. ECC (maybe) isn't worth it for consumers of content, but for creators it should be standard. not to mention rowhammer-like attacks which ECC makes harder though not impossible.

Just because there are real reasons to have ECC does not mean you are going to experience them, let alone experience headache because of them.

Software is fault tolerant. Small memory failures can be trivially tolerated.

ECC simply isn't worth the cost, especially for workstations.

> Software is fault tolerant. Small memory failures can be trivially tolerated.

Most software is not at all fault tolerant in this respect. It just crashes. Fortunately, most users are somewhat fault-tolerant.

It only takes one bit wrong to fuck your file system.

From what I understand (very roughly) with ZFS or BTRFS at least it takes one flipped bit in the particular block of memory which has just been successfully checksummed and 'approved' by the filesystem as valid for a write operation. Those are quite a few hoops to jump through so let's not head towards melodrama just yet.

That’s fine but most of us are running on NTFS, HFS+ or ext4 where it matters.

And then there’s the same problem with MLC SSDs which are just arrays of capacitors encoding that state as different charge potentials. As a former EE, capacitors leak regardless of implementation.

In 20+ years that bit eluded me...

dram cells used to be bigger.

Yes. Exactly.

Also SSDs scare me more. MLC disks have so many states stored in a capacitor that is trusted not to leak over the lifetime of the device. All it takes is a few electrons to go walkies and it’s game over.

As far as you know.

You make it sound like filesystems are so vulnerable. They aren't.

The chances of memory failure ruining data are extremely slim; lower than for hardware failure. If you care about your data, back it up.

Backups don't do anything if the original material contains the flipped bits.

I ran a lab notebook in word for a number of years and the document got fried by a filesystem error. The backup was toasted as well. This was corruption which broke the stage machine or parser in word which acted as a damage amplification.

As for the source of the error, we can only speculate but the possibility of a flipped bit is higher than you think.

Try running memtest86 on a PC without ECC for a week. You will be surprised at the results. I’ve seen perhaps 3 or 4 failures of this in new hardware. DRAM is not perfect.

Yeah might as well get cheaper RAM, a slower SSD, and a Pentium, because everyone knows quality components are a ripoff, and the people at PC Gamer are just Apple fanboys and don't know what they're talking about.

That's a horrible system for $3700 dlls. You can easily shave $1k off if you're smart about shopping for the parts. Also, assuming you're not going for OSX (apples and oranges I know), a thread ripper would be a much better choice for a CPU.

ECC us unlikely to be worth the inflated price, too.

A slightly slower CPU can be had for a significantly lower price. GPUs are a lot more important to most workloads.

... But I didn't mention any of those.

The only reason for an expensive motherboard is to use multiple video cards.

>$200 is not a good price for a PSU. 1000watt is absolutely unnecessary unless you need to power 3-4 GPUs.

My point is that you can easily build a very similar system for half the price. This isn't the rattly '90s. $5000 is a lot for one PC.

If you want to shell out $3700 on PC components, you can build a much more powerful system, especially when you start putting in more GPUs.

> $200 is not a good price for a PSU. 1000watt is absolutely unnecessary unless you need to power 3-4 GPUs.

For sure. I had this argument a few weeks before the iMac Pro came out with someone who insisted that it'd come with an 850W, more likely 1KW+ PSU...

https://news.ycombinator.com/item?id=15922978

iFixit's tear down? 500W PSU.

Bit silly I reckon as well as you can get similar single threaded performance for much cheaper. If you need 18 cores maybe, but $5k? But Apple have always figured out how to upsell computers perfectly.

If you buy a workstation from HP/Supermicro, then you get something much more powerful for the same price. So, to say, they are very close is not very correct. The problem here is that the assembled PC is a customer PC assembled for workstation performance. Even then, they were not trying very hard.

If you buy a workstation from HP/Supermicro, then you get something much more powerful for the same price.

Not really. Go to HPs site and put together something like a z640 with as similar spec as possible and you're close to $4k before even adding a screen.

Point. Somehow, I have seen quotes from HP which were always much cheaper. May be it is because my employer had some sort of deal? Also, proliant servers were somehow cheaper than towers. Again I am not sure how, but yes, point taken.

What HP (and others) does give you is the opportunity do buy something almost as good (and quite possible more than good enough) for less money. They also let you buy something for the same price that is a lot better along the one axis you care most about (but worse along the axis you don't care so much about). But if you want to blindly match spec sheets you're going to pay about the same.

If you think the price is unreasonable, you're probably not the sales target for this machine - the ones who really need this, like in the case of this reviewer, are buying three at a time - just because.

I think it's pretty simple: People who want or need Macs and who have a workload that's highly parallelizable (sp?) will like this machine. Or people who need ultra-fast I/O. That group is basically video editors, (some) programmers who do a lot of local compilation for whatever reason, and a couple of others.

It's a pretty small-ish group but also probably a noisy, discerning one online.

For much of that group, the computer is a tax write-off anyway—something individuals buying for themselves often forget.

Many people forget that, in smaller businesses, the older machines are also quietly wiped and then sold for cash on Craigslist.

> the computer is a tax write-off anyway—something individuals buying for themselves often forget.

While this is true, the people who are unaware about what tax admitted expenses are, often just assume that it is "free money". It is not. You still have to earn whatever you spend on the machine, and it is up to you to consider, what is and isn't cost effective for your specific situation.

And not only that; in many jurisdictions, there are limits in what can be considered OPEX. If you go above that, your spending has to be accounted for as CAPEX and you have to depreciate the item. Depreciating a computer for several years, instead of expensing it in the current year, can quickly decrease its appeal.

In the us right now small businesses can depreciate capex fully on year of purchase for up to 500k. About to become 1 mill. So for many small businesses there's strong appeal.

Lucky you. Here I'm sitting in EU country, the cutoff is 1700 EUR (without VAT) and the shortest depreciation period is 4 years. Sigh.

> (some) programmers who do a lot of local compilation for whatever reason

How else do you run/test the code you're working on?

Upgradability and serviceability is a big concern for a very niche market. If I ran a smaller shop, the iMac Pro would be a clean solution to a complex problem. But once you hit edit-bays at scale the technical problems exponentially grow. You also can't purchase every machine maxed out when you have to fine-tune a budget. We've often purchased almost all of our Macs without RAM to shave costs.

I'm personally holding out for the new Mac Pro specs to become public. A modular system is ideal in a large post-production facility. I have enough external peripherals and connections for 10GbE, fiber and SDI that attaching them all to an iMac is as much of a headache as anything. Especially when you're doing live productions and you have to swap out parts frequently.

Price is the least of my concerns though. If I can shave a lot off my budgets by upgrading RAM and GPU over time vs replacing entire machines, that's a big bonus. But I have more concerns with overheating in the iMac Pro than anything.

And the touch bar is useless to most of my editors and graphics team. If we could opt for a model without it, we would, every time...

How much do companies who buy Xeon workstations actually normally upgrade, them, though?

They might care about having to round-trip them to a repair facility (even if that's the back of an Apple Store) though, and not (without complete disassembly) being able to take the storage out before handing them over.

Quick comparison, HP: 3 year Next Business Day onsite Warranty - they come to you, they make assertions about how quickly they'll be there. Can be extended to five years (for a Z840 for less than £200 or ~5% of the machine's purchase price)

Dell Precision: same. To add another 2 years to a Precision 7920 costs a bit over £200 or ~7%.

Apple: one year. £159 (~3% of base price) to extend to three years, and no option for more. You have to take it to an Apple Store (good luck with that), or Apple can arrange to have it shipped by courier. Hope whatever courier they use don't drop it along the way, and no sign of a guarantee when they'll get it fixed.

I expect the target market for the iMac Pro overlaps strongly with Apple's other machines (i.e. the iMac Pro's replacing an existing Apple computer) so this won't be a surprise but it's disappointing that Apple won't stand more behind their machines.

FWIW Apple does onsite repair service as a part of standard AppleCare, or you can get Enterprise/Premium for next day onsite (there's also mention of a 4 hour option, but im not sure what plan thats a part of)

I had a 2011(?) iMac at my last job that was having GPU issues common to all of those models. The machine had left warranty a couple years prior, but Apple knew of this issue and paid my local Apple dealer to replace the GPU.

For wide-spread issues at least, they do stand behind their machines.

> For wide-spread issues

Frequently because they're forced to, by the threat of litigation and/or a smackdown from consumer-protection organisations. There's too often a window where Apple won't acknowledge there's an issue but your warranty's expired so eh...maybe they'll make an exception for you? Or maybe not.

I'm looking for something a bit more systematic. On an iMac Pro that starts at a shade under £5000, it wouldn't hurt them to make the default warranty 3 years and offer an extension to 5 (I mean, who's throwing away these machines after 3 years, never mind one?).

Does AppleCare for Enterprise what you're looking for?

https://www.apple.com/support/enterprise/

In terms of the features on offer (on-site repair, SLA) sure. On pricing...

> Contact Apple, an Apple Authorized Reseller, or an Authorized Carrier Partner to receive a quote for AppleCare for Enterprise.

Not so much. I guess what I'm really expecting is that for a workstation-class machine this is the rule rather than the exception.

Exactly - I think by the time they decide to upgrade workstations every component in the stack has advanced enough to warrant a completely new machine.

That, and there are typically many other good reasons to just replace a machine:

- A new machine is in warranty.

- You don't need to keep track of what machines you have, how many memory slots they have, what DDR version and speed they require, etc. You don't need personnel to do the replacements.

- In large organizations, hardware is only a small cost compared to salary, etc. So, it is typically not a problem to replace machines every 2-3 years, if necessary.

Upgradability is more interesting to individuals and small businesses.

I'm not convinced.

> A new machine is in warranty.

Warranties are for individuals. Businesses will use them if they exist, but when a machine fails you replace it with a spare off the shelf. All a warranty does is save you the cost of buying another spare, which amortized over a fleet is nowhere near the cost of buying all new equipment.

> You don't need to keep track of what machines you have, how many memory slots they have, what DDR version and speed they require, etc. You don't need personnel to do the replacements.

There is inventory software to automate this. And upgrading a machine requires much less work than replacing it because an upgrade doesn't normally require transferring programs, settings and user data.

> In large organizations, hardware is only a small cost compared to salary, etc. So, it is typically not a problem to replace machines every 2-3 years, if necessary.

That other costs are larger is no excuse to waste money. An upgrade will often make a 2-3 year old machine equivalent to this year's machine, which is typically less expensive than replacing the entire machine. Especially since you don't have to replace any component that isn't actually a bottleneck in your applications.

But the main value of upgradability to large companies is as a hedge against unexpected requirements. Some vendor's new software requires twice the resources as the old version two months after you've bought equipment using the old spec. Either you can upgrade the machines or you have to replace them. Even if you don't always have to do it, you always want to be able to.

Nvidia-powered hackintoshes are a counterpoint mentioned in the article.

Yeah, but this came out at the last WWDC, I think - https://developer.apple.com/development-kit/external-graphic...

I believe Nvidia drivers are already in beta, this or any other eGPU enclosure should be able to remove that sticking point.

There is a slight performance penalty because fo the TB3 bus conversion bus, but it's only applicable for real time work like gaming - pro work should be able to use Nvidia cards just fine.

Similar to what the parent wrote, people who buy Nvidia-powered hackintoshes probably aren't the target market.

> It's interesting because upgradability doesn't seem like a very big concern to anyone buying this

From my discussions with an iMac hardware engineer, he explained that the component set that these machines are made with are at least 7-10 years away from being replaced because the software catches up. Whereas, traditionally, we're used to machines that last 3-5 years, but you can easily upgrade in the future.

If true, it appears to be a major paradigm shift, which rightfully the market is having a hard time trying to swallow.

I bought a max specced Macbook Pro and it's borderline unusable for me due to the touchbar. Apparently I rest my hands normally in the area where the touchbar is and it's very hard to undo habit. Since the touchbar doesn't need touch force I often randomly activate one of the buttons. I really wish they offered it without the touchbar as an option.

How do you rest your hands in the tiny space above the keyboard?

Reply on news.ycombinator.com