Matt Casebolt and Chris Lattner does not sound good for Macs.https://9to5mac.com/2017/01/11/matt-casebolt-touchbar-macbook/
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Surely Apple has employees who read these forums...
- Thread starter aaronhead14
- Start date
- Sort by reaction score
They probably left a dead end job for one with a future.
Considering how much stick the current Mac Pro case and Touch Bar get on this site, the guy responsible for them leaving Apple should be seen as a good sign.
It isn't a declining industry. It is a steady industry, founded upon people who like to connect to great computers. The dilletantes are going after the gadgets; the computer people want great computers. Just a theory ...
There are dozens of industries built around a powerful desktop computer as a workstation + hub for peripherals and interfaces. These are long standing markets that wouldn't move to another ecosphere just because some other vendor releases a shinny thinner model of a machine. This sort of market is rock steady and it would take minimum effort to satisfy just with incremental updates.
But Apple chose to not do even that because like some said, on a purely capitalistic point of view, the success of mobile platform is too much to not focus the company's resources on. I think Apple's biggest problem is its reluctance or sluggishness in merging their macOS and OS ecosphere. Refusing to implement touch onto OS X essentially puts themselves in the current position where priority is given to iOS with macOS undermined, then disappointing macOS users as a compromise. If they had a touch desktop OS in place similar to Win 10, both spheres could then move forward concurrently.
I would argue the only thing Apple lacks are good graphics cards. For most people there are plenty of options to expand IO, storage, and ram (16-32GB being pretty good).
But it's like why make it easy for Mac Pro graphics card upgrades if you only release cards that are slow and ancient?
Why is it easy for PC laptop makers to put two 1080s in their machines that cost less than a Mac Pro? It's ridiculous.
But it's like why make it easy for Mac Pro graphics card upgrades if you only release cards that are slow and ancient?
Why is it easy for PC laptop makers to put two 1080s in their machines that cost less than a Mac Pro? It's ridiculous.
I think Microsoft is again experiencing growth, particularly as hardware offerings expand to fill in the hole neglected by Apple and the Mac line. They missed the boat with mobile phones, but who knows? If Apple fades in its iPhone / iPad products and sales MS might rise again in that area (though Android makes that unlikely). MS is particularly expanding in hybrid offerings, as are several PC makers on the Windows 10 platform.
Hybrids and super niche markets fill quite small holes. The Surface Studio's base model and maybe mid tier Studio can't even run Adobes Lightroom adequately without problems. ( See the TWIT networks Windows Weekly #494 )
https://twit.tv/shows/windows-weekly/episodes/494
While I'd obviously prefer a Mac with up to 128 or 256 GB of RAM, PCI-E slots, expandability, etc., I also agree with what you're saying here! Out of all the problems, the graphics card issue is the one that needs fixing the most. Seriously, why can't Apple just use some GTX 1080's in their machines? It seriously can't be that difficult...
I've used Touch on Windows for many years (going back to Windows XP on the Panasonic Toughbook) and while it can have a function on portables, for desktops... Look at how hard Windows 8 flopped in the marketplace because it demanded users use touch as their input, even in scenarios (like desktops) where it was at best superfluous and at worst actively hostile.
As for merging iOS and macOS, Apple is only accelerating that - but they're doing it in the background, leveraging their common foundation. And even then, we see a fair bit of backlash in these forums about the usability of a new bit of functionality added to macOS that was previously an iOS-exclusive.
Just an FYI for those that might not know: today I discovered that not only does Apple have iTunes for Windows but it also has iCloud for Windows so that Outlook can see your email, contacts, calendar and reminders...
From what I have seen, the OSX apps get butchered so they can fit within the capabilities of iOS.
Usual stuff - browsers, IDEs, remote desktops, SSH clients, image and video editors...
And VMs running databases and other VMs running test environments and other VMs running different IDE versions.
Looking at expansion options - either upgrade to 256 GiB (and eBay the 128 GiB), or get a new system with possibly two sockets and lots more DIMM slots.
And this is my home PC - for work I use serious hardware.
And RAM disks are slower than my spinners and SSDs.
Easy.
Disk controllers use DMA so that the CPU isn't involved in data transfer.
RAMdisks use "memcpy" CPU instructions to move data - so data transfers need 100% of a thread.
Disk controllers use NCQ and other parallel transfer tactics - most RAMdisks are single-threaded.
______
Therefore, if your application is multi-threaded and CPU-bound, and you have fast disks - the CPU hit from a RAMdisk can mean that your application runs slower from a RAMdisk than from real disks.
Yeah. We spent a bunch of time redoing a workflow to use RAMdisks - only to find out that the new version was slower.
It turned out that it was CPU-bound and highly parallel - and we had fast RAID-5 and RAID-0 disks. The CPU hit from the RAMdisk made it slower.
Perhaps a better RAMdisk that had better parallelism would have won - but we licked our wounds and went back to real disks.
Support computing operations for a research lab, including a handful of systems with 72/144 C/T, 1 TiB RAM, 12K CUDA cores. Most systems 12 to 32 cores, 128 GiB to 768 GiB RAM, 2K to 6K CUDA cores. No ATI GPUs anywhere. Upwards of 100 10GbE ports and an uncountable number of 1GbE ports. 16 Gbps FC for storage. Typical disk purchases are in increments of 96TB 2U shelves, about a $1M capex budget per year. Mix of FC and DAS depending on the project. The usual midrange stuff.
Seeing all that sexy hardware must be fun. Although its definitely a hard job. Thanks for the reply.