Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
put that thing at $1799-1999 and literally will disappear from web and stores, remove the external OLED, i dont care a sh... :cool:
 
I find it hard to understand why the Vision Pro comes with an M2 Chip, instead of an M3. M3 was already launched in 2023, consumes less power, and has much more GPU performance, something critical in a VR/AR product. Games are one big reason for buying this, and choosing the M2 seems weird.
Weirder yet because the M2 lacks M3"s hardware decoding of AV1, GPU lacks Mesh Shading and Ray Tracing that the M3 offers.
 
I’m really curious what compromises Apple will choose to implement on the non-Pro Apple Vision in the future.

128GB storage?
8GB RAM?

I’m not sure what else to reduce.

Reducing the screen quality would pretty much eliminate interest in that version completely.

Removing the outer screen would eliminate the trademark look of having your simulated eyes visible to people around you.

Maybe no battery, requiring it to be tethered at all times? But then you can’t show it off in public unless you’re near an outlet.

I feel like it’ll be difficult to make a version that is $1,999, but that price point would definitely attract more buyers. Not that they’ll need more buyers just yet. They’ll sell out of the limited units they’ve manufactured in its first year.
 
M2 Chip? Why are they not using M3, isn’t it supposed to be a little bit faster and more efficient?
 
I am utterly disappointed to see that they will nickel and dime us on this device too.

The sheen is wearing off at an increasing rate.
 
I’m heading out now to sell blood so I can get the upgrade to 512, I didn’t budget for it over the last 6 months because i thought it would surely come with 1 tb for $3499
 
Is it a hint Apple might put 16 GB of RAM in the next Macs ? I certainly hope so. They'll say whatever they want with the RAM performance on their SoCs, at the end of the day, we all know 8 GB is not enough, and so much that it's even been proven to be THE bottleneck of Apple Silicon.
Any amount of memory is the bottleneck of ANY silicon. Devise a test large enough such that no processor can maintain it in memory and see the RAM be the bottleneck. :)
 
Everyone is making fun of this for having 16gb of ram when the M3 MacBook Pro starts with 8gb. But remember, this device has 2 chips: The M2 chip and the R1. Since it has unified memory, they're both sharing the RAM. I doubt the R1 needs an entire 8gb of RAM, but it still needs a few gigabytes. So you're not going to be able the have the same experience as a 16gb MacBook Air.
My guess is Apple realized that 8gb wasn't enough to run the M2 chip and still have some left for the R1 chip, so they increased it to the next amount that the M2 chip could support, which was 16gb.
Unified memory provides access to CPU and GPU. Have you seen any evidence that R1 has access to it too? That seems highly unlikely. It's a separate chip/package. It can't be physically connected to the unified memory.
 
Except, the motion co-processor didn’t have to deal with video which will require substantially larger buffers for the video pipelines.
From Apple:
“R1 chip is specifically dedicated to process input from the cameras, sensors, and microphones, streaming images to the displays within 12 milliseconds”
As I read this, R1 is not capturing images for storing in a buffer. Or, if it IS a buffer, it’s a tiny and very fast one that is only pushing to the screens. There’s no need to hold or retain any images past the 12 ms it takes to stream them. Any added buffer would just increase the lag over 12 ms. I think I remember this being described as a part of the security/privacy as the R1’s images of the world go straight to the internal screens. So, when using it as a non-capture device, the M2 is simply overlaying graphics on the R1 stream and not actually recording the faces around you.

IF this is true (we’ll find out soon) it would mean that, if someone is doing a demo and recording what they see, the external screens should really be doing that thing that indicates it’s recording the surroundings (just like it would when holding down the capture button). Will be good to get a deeper dive soon (and peruse the documentation).
 
M2 Chip? Why are they not using M3, isn’t it supposed to be a little bit faster and more efficient?
Availability would be my first guess.
Also, the Vision Pro was most likely designed when the M1 was available. So the M2 was an upgrade and fixed most of the GPU issues M1 had. It does lack AV1 "decode" and Raytracing. But, my first guess is what I'll stick with for the answer. M2 is more plentiful and M3 is not. We don't have Ultra M3's. And they didn't offer them in Studios and or Mac Pro's. Most likely due to the same reason. There is still a supply chain issue when it comes to CPU's.

Plus, Apple sticks with whatever plans they have till the product is "ready". M2 is ready, and so was the rest of the system. Better to get it out the door on time knowing it will work as expected. Then to "wait" and maybe short the available chips for iMac's and MacBooks. We don't even have an M3 mini now that I think about it. So, tight supply.
 
I find storage configurations and prices by Apple quite disappointing . I had a iMac 21" bought about 15 years ago with a 200 Gb HD. Today it is still 256 Gb aber SSD. Much faster indeed but same size. Many computers end up having similar storage than iPhones or iPads. I am not even discussing prices. Just sizes.
The new storage world today, network storage. No need for large on system storage. It works!
 
From Apple:
“R1 chip is specifically dedicated to process input from the cameras, sensors, and microphones, streaming images to the displays within 12 milliseconds”
As I read this, R1 is not capturing images for storing in a buffer. Or, if it IS a buffer, it’s a tiny and very fast one that is only pushing to the screens. There’s no need to hold or retain any images past the 12 ms it takes to stream them. Any added buffer would just increase the lag over 12 ms. I think I remember this being described as a part of the security/privacy as the R1’s images of the world go straight to the internal screens. So, when using it as a non-capture device, the M2 is simply overlaying graphics on the R1 stream and not actually recording the faces around you.

IF this is true (we’ll find out soon) it would mean that, if someone is doing a demo and recording what they see, the external screens should really be doing that thing that indicates it’s recording the surroundings (just like it would when holding down the capture button). Will be good to get a deeper dive soon (and peruse the documentation).
You are probably right, I forgot about the 12ms latency. That doesn’t leave much time for doing much processing. Maybe this is just the processing that an ISP would normally do which doesn’t need full frame buffers.
 
  • Like
Reactions: Unregistered 4U
We were gifted two Meta Quest 2 units and they mostly sit in a drawer. Without a killer app the Vision Pro is a very expensive paperweight. With the amount of attention paid to iPhone, MacBook, and Apple Watch battery life, what is so compelling about a $3500 headset that you can only use for two hours at a time?

Nothing Apple presented in the Vision Pro reveal made me want one and I don't expect the launch reviews will change that. I already gave Apple close to $3500 for my Space Black M3 Max MBP. There's no way I'm spending an equal amount for a Vision Pro with limited use. I'd rather put that money towards a family vacation or a very large screen TV/ultra short throw projector that everyone can enjoy together.

I will easily splurge on a future foldable iPhone but Tim & Co will have a very tough time convincing me to help Apple recoup their R&D money on some fancy spectacles.
 
Last edited:
My first Mac was a PowerBook 180 with an upgrade to 12MB RAM in the early 90s, when it graced the covers of all the Mac and PC magazines. I vaguely recall that state-of-the-art Mac with its 16-grayscale active-matrix display cost some $4200, and it was worth every penny as I wrote my dissertation. So the whiners complaining about $3500 today as outrageous have lost perspective on just how cheap our tech has become. And that's before Apple's 0% financing!

QFT.

People also complained about Macs having small screens, not realising Apple stated the VIS (viewable image size), not measuring the bezels when stating the diagonal.

The size of a screen is usually described by the length of its diagonal, which is the distance between opposite corners, usually in inches. It is also sometimes called the physical image size to distinguish it from the "logical image size," which describes a screen's display resolution and is measured in pixels.

The method of measuring screen size by its diagonal was inherited from the method used for the first generation of CRT television, when picture tubes with circular faces were in common use. Being circular, the external diameter of the bulb was used to describe their size. Since these circular tubes were used to display rectangular images, the diagonal measurement of the visible rectangle was smaller than the diameter of the tube due to the thickness of the glass surrounding the phosphor screen (which was hidden from the viewer by the casing and bezel). This method continued even when cathode ray tubes were manufactured as rounded rectangles; it had the advantage of being a single number specifying the size, and was not confusing when the aspect ratio was universally 4:3. In the US, when virtually all TV tubes were 4:3, the size of the screen was given as the true screen diagonal with a V following it (this was a requirement in the US market but not elsewhere). In virtually all other markets, the size of the outer diameter of the tube was given. What was a 27V in the US could be a 28" elsewhere. However the V terminology was frequently dropped in US advertising referring to a 27V as a 27". This was not misleading for the consumer as the seller had to give the actual screen size by law. Flat panel displays by contrast use the actual diagonal of their visible display size, thus the size is the actual size presented to the viewer in all markets. This means that a similarly specified size of display will be larger as a flat panel display compared with a cathode ray tube display.



 
The higher storage variants, if launched might have a price difference of $500. $3999 for 512 GB and $4499 for 1TB variant. Expecting costlier than Mac storage upgrade prices.
 
  • Like
Reactions: Flowstates
Why do you need the device you’re using to post on this forum?
The device I use to post on this forum has nothing to do with the functionality the vision pro has. These 2 things are not comparable at all. Your logic is very foolish. It’s like when your daddy asks you why you need $5 for, and you asks him why he needs $5 in his pocket…
 
Last edited:
The higher storage variants, if launched might have a price difference of $500. $3999 for 512 GB and $4499 for 1TB variant. Expecting costlier than Mac storage upgrade prices.
Well sure.... 512GB might cost $500 if you ignore the consistent $200 upgrade cost for 512GB on every apple silicon device. Honest question, when you were fabricating that upgrade cost, why did you pick $500? Did you think that was the best scary number or did you simply copy someone else's scary number?
 


Apple's upcoming Vision Pro headset is equipped with 16GB of unified memory, according to files related to the device in Xcode 15.2.

apple-vision-pro-chips.jpg

Our finding in the latest version of Apple's app development tool confirms a June 2023 report from Bloomberg's Mark Gurman that said the Vision Pro would be equipped with 16GB of RAM, and that is also the same amount of memory that was included in Vision Pro development kits distributed by Apple last year.

Xcode 15.2 officially adds support for developing visionOS apps, and Apple is now accepting these apps via App Store Connect.

In a press release this week, Apple reiterated that the Vision Pro will be powered by its M2 chip for "powerful standalone performance," along with an all-new R1 chip that "processes input from 12 cameras, five sensors, and six microphones to ensure that content feels like it is appearing right in front of the user's eyes."

Apple's press release also confirmed that the Vision Pro will start at $3,499 in the U.S. with 256GB of storage, and this wording seems to imply that higher storage capacity options will also be available. It was reported last year that at least some of the Vision Pro developer kits were equipped with 1TB of storage.

Apple has yet to share full tech specs for the Vision Pro, which will be available to pre-order in the U.S. starting Friday, January 19 at 5 a.m. Pacific Time. The headset will launch there a few weeks later, on Friday, February 2. When it first announced the Vision Pro, Apple said the headset would launch in additional countries later this year.

Thanks to Dimitris Sartzetakis, Michael Burkhardt, and Steve Troughton-Smith for assistance.

Article Link: Apple Vision Pro Features 16GB of RAM and Likely Up to 1TB of Storage
So it is technically feasible to put 16gb ram on a m2 chip. The innovation. 🤯
 
So, I'm not impressed by a $3,499 device with the same amount of RAM as my smartphone. Especially with how magical some people think this device will be.
So a top-tier camera is bad device? Because those have even less than 16Gb of RAM and are way more expensive than 3500 dollar.
 
I’m really curious what compromises Apple will choose to implement on the non-Pro Apple Vision in the future.

128GB storage?
8GB RAM?

I’m not sure what else to reduce.

Reducing the screen quality would pretty much eliminate interest in that version completely.

Removing the outer screen would eliminate the trademark look of having your simulated eyes visible to people around you.

Maybe no battery, requiring it to be tethered at all times? But then you can’t show it off in public unless you’re near an outlet.

I feel like it’ll be difficult to make a version that is $1,999, but that price point would definitely attract more buyers. Not that they’ll need more buyers just yet. They’ll sell out of the limited units they’ve manufactured in its first year.
Removing the outer screen as that significantly increases cost on the manufacturing side. But it is also possible they just improve the device and sell the old model along side it discounted.
 
Last edited:
So a top-tier camera is bad device? Because those have even less than 16Gb of RAM and are way more expensive than 3500 dollar.
Are you seriously comparing the amount of RAM in a top-tier camera to the amount of RAM in an AR/VR headset and a foldable smartphone? If so, that doesn't make sense.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.