tht
About
- Username
- tht
- Joined
- Visits
- 195
- Last Active
- Roles
- member
- Points
- 8,024
- Badges
- 1
- Posts
- 6,023
Reactions
-
13-inch iPad Pro review: hardware of the future running software of the past
hexclock said:FTA:
”Since I can't have more than one audio app active on iPad, I can be listening to music and have a video ad in Safari stop playback. System audio sounds like notifications interrupt audio, too.”
it depends on the App. I can run Audiokit’s SynthOne, their FM synth, and the 909 drum machine concurrently, either playing loops or played live. Maybe System audio is handled differently. I’m sure someone else here can shed more light on that.
For this particular issue, I've always attributed it the ad injector, where I think it is sending a stop/pause command to the background audio stream, so that the user can hear the audio in the ad. Yup, they audio-jack. It gets worse than that. They ad injector can also scroll-jack. It stops registering touch input in the area of the ad-injection so you can't scroll quickly away. You have to find an a non ad area to register your touch to scroll away. This is one of those hypocritical things about reviewers hemming and hawing about this or that product having such poor design while their own website was a user hostile design nightmare. They always say they can't do anything about it because the ad-side and editorial side don't mix.
Apple might have designed it this way on purpose so that the user knows a video ad is playing and burning CPU cycles. Who knows. On the desktop, browsers have features where background browser tabs are auto muted, UI buttons/symbols have to be added to the tab to indicate something is playing, etc. So, I expect ads will be a usability problem one way or another on whatever platform you use.
I tested audio playback on iPP10.5 just now. I can play Apple Music in a SlideOver sheet hidden away, and have 2 FileBrowser.app windows in SplitView, each with a video playing. The 3 sources are muxed together and I hear all 3 sources of audio at the same time. When I switch to Safari, one of the videos goes into pop-up video mode, I still hear Apple Music. Hard to tell if the 2nd video was still going with Safari in fullscreen, a pop-up video playing, and Apple Music playing.
My question with Stage Manager is how much can you load it up. Each Stage can have 4 windows. Is every window in each stage running? Only the front-most stage are active? What are the multitasking characteristics? If I run Pythonista script that takes 10 minutes to complete, will it keep running if it is in Stage while I'm using Safari?
-
WWDC unlikely to see debuts of any new hardware at all
Fred257 said:Apple isn’t doing itself any favors by not updating the Mac Studio and pro this month. This rumor makes no sense unless Apple isn’t serious anymore about their desktop pro lines
One the one hand, it speaks to desktop computers being only 10% of their Mac units, but on the other hand, you'd think the effort to update the iMac, Mac mini and Mac Studio on a yearly basis to be minimal. It should be worth that effort, but sigh. Same with everything that is not iPhone, Watch or Macbook. iPads could go 18 months or more. Apple TV is every 2 years or more. Some story with the software too.
For the M Ultra, they need to find a way to put it into the MBP16 and for use as a server, to spread the development costs around so that they can have yearly Mac Studio and Mac Pro updates. -
Microsoft's Copilot PC and the M3 Mac killer myth
AppleZulu said:
It boggles the mind that still no other major manufacturer has sought to replicate Apple's closed-system design paradigm, with operating system and hardware conceived and implemented as a single, unified thing.
Even when the cost of the platform is free, like with Linux and ChromeOS, on hardware that is half the cost of average PC hardware, they can't really make inroads. A lot of the value in the PC platform market is locked up in MS Office and its satellite software and services.
Even Google, who burns up billions of dollars per quarter without accomplishing much, can't even make many inroads with ChromeOS. It is a little curious why they can't. Like, they won't make a deal with Valve for Steam, can't get developers to make good webapps, etc. Undercutting MS and PCs seems to have plateaued in the education market. -
AI computer showdown - MacBook Air vs. Microsoft Surface Laptop Copilot+ PC
brianm said:KITA said:<...>
The 23 W version is what's in most devices.
From Anandtech (Oct. 2023)
"Thin and Light" version has a 23W TDP:
I'm curious how the X Plus and the X Elite "Thin and Light" are both listed with 23W according to that Anandtech article. the X Elite has 2 more cores than the Plus - so either it must be running at lower clock, or it is actually using a little more power (each core is likely using somewhere around 1W if they are anything like the Apple Performance Cores)
These are all one SoC chip, binned into 4 SKUs. I would guess the X1E-80, X1E-78 and X1P-64 are the "23W" SKUs. "23W" in quotes because it seems Qualcomm has just dropped that distinction, or stopped using TDP as a SKU distinction all together. Will have to wait for measurements.
And "thin and light" isn't much of a category these days either? Like, these Surface Laptops are as thick as the MBP14 or MBP16. Maybe the categories these days are consumer laptops, business laptops, and gaming/workstation laptops? The MBA is basically in a unique category. -
AI computer showdown - MacBook Air vs. Microsoft Surface Laptop Copilot+ PC
saarek said:KITA said:saarek said:“AI computer showdown”.So which one of them is actually better at processing ML/AI? Surely that should be the primary focus based on the title?
Snapdragon X Plus - 45 TOPS (INT8)
Snapdragon X Elite - 45 TOPS (INT8)
Apple M3 - 18 TOPS (FP16)
Apple M4 - 38 TOPS (INT8)
The NPU was meant to be a big selling point for this particular chip. Looks like they actually delivered on that claim.
For those confused by this whole new world. A rating of 40 TOPS or higher is generally seen as the benchmark where AI (basic AI) can be processed on device as opposed to needing cloud assistance. Obviously there is a lot more to it, but the marketing seems to match the hype on this occasion.
MS doesn't have to have a trained AI model that only can work at 40 TOPS or more. You will soon see that Co-Pilot can run on a CPU from 4 years ago. How much slower will it be? Nobody knows just yet. What's to stop them from developing an AI model that works on 30 TOPS, or 20 TOPS, or 10 TOPS? Then, doing what? It's all a design choice. Heck, the cycle time of doing a network query to computer on the network somewhere and having the results come back will be quite variable in itself. There's going to be some kind of model that will say this or that question should be done locally or through the network.
Obviously, if it is running on a discrete GPU, it will likely be faster than the Snapdragon NPU. Microsoft or the GPU OEMs will enable it to run on GPUs. It will be interesting to see how Snapdragon versus Intel/Nvidia/AMD system war plays out.
Lastly, it's not clear to me that the M3 TOPS number of 18 INT16 TOPS doesn't equate to 36 INT8 TOPS. If it can't pack 2 INT8s into a INT16 and operate on them simultaneous on the M3, that's a curious design decision. That's usually a cheap freebie.