Apple Intelligence and iOS 18.1 should debut on October 28

Jump to First Reply
Posted:
in iOS

The public will finally get to try out Apple Intelligence at the end of October, with iOS 18.1 and Apple's AI features said to be exiting beta on October 28.

A hand holding a smartphone displaying various colorful app icons on its screen against a light background.
Apple Intelligence on an iPhone



Apple has been testing out various elements of Apple Intelligence since after the introduction at WWDC, via a second developer beta cycle. After multiple months, that cycle could finally come to an end with a public release.

On Sunday, a Bloomberg newsletter wrote that Apple had been clear about Apple Intelligence's arrival at some point in October. The release, the publication was informed, would occur on October 28.

Apple is reportedly taking its time to bring Apple Intelligence to the public, and with such a public launch, it has good reason to do so. It's making sure there are no major bugs for the release, which could severely harm the feature's reception.

It is also preparing its infrastructure, making sure that its AI cloud servers can handle the sudden influx of traffic.

The initial iOS 18.1, iPadOS 18.1 and macOS Sequoia 15.1 releases will only have part of the Apple Intelligence suite, with notification summaries and Writing Tools being the main inclusions.

Future updates will add more features. In iOS 18.2, there should be Genmoji custom emoji and ChatGPT integration with Siri. By iOS 18.4 in March, Siri will gain many updates, including contextual responses that rely on personal data points in files and communications with others.

The release on October 28 will start a few busy days for Apple. Earnings are on October 31, and new M4 Macs and an iPad mini are expected to ship on November 1.

Rumor Score: Likely

Read on AppleInsider

Comments

  • Reply 1 of 8
    22july201322july2013 Posts: 3,844member
    > It is also preparing its infrastructure, making sure that its AI cloud servers can handle the sudden influx of traffic.

    What hardware will Apple Private Cloud Compute be running? Xserve with Apple Silicon? I might want to buy one of those.
     0Likes 0Dislikes 0Informatives
  • Reply 2 of 8
    M68000m68000 Posts: 950member
    Going to wait on doing 18.1 after the somewhat rough start with 18.   Just tried changing my mesh wifi router to use different names for the SSID for the 2.4 and 5G networks.   Testing the outcome. Linksys app and WiFi itself has been sluggish or not working at all since 18
     0Likes 0Dislikes 0Informatives
  • Reply 3 of 8
    I am installing 18.1 as we speak.

    Oops, my bad: 18.0.1
    edited October 2024
     0Likes 0Dislikes 0Informatives
  • Reply 4 of 8
    gatorguygatorguy Posts: 24,769member
    Reading the current reviews from the beta users, the AI features coming end of this month range from useless, to good in theory but unreliable, to OK and a good start.
    edited October 2024
    grandact73
     1Like 0Dislikes 0Informatives
  • Reply 5 of 8
    mpantonempantone Posts: 2,498member
    > It is also preparing its infrastructure, making sure that its AI cloud servers can handle the sudden influx of traffic.

    What hardware will Apple Private Cloud Compute be running? Xserve with Apple Silicon? I might want to buy one of those.
    Apple did not mention a marketing name during WWDC 2024; they vaguely referred to some sort of server running Apple Silicon. It is likely that Apple will not market these servers so they likely will only be identified by internal codenames or model numbers.

    It is also unlikely that Apple will provide any details about the Apple Private Cloud Compute devices. Nothing new about that, Apple doesn't discuss their infrastructure. Whether they are binned samples of existing SoCs or something else was not detailed.

    My guess is that Apple currently has a mixture of hardware from third parties as well as their own devices in their AI datacenters. It is probable that the makeup of those servers will change over time and will lean toward devices that feature better performance-per-watt. If that is the case, it is likely that they will increase the percentage of Apple Silicon devices.

    If you want the closest thing to an Apple Private Cloud Compute device, you should probably aim for a Mac Studio with an Ultra SoC. Naturally the Apple Private Cloud Compute devices will be in a form factor better suited for datacenters but the Ultra chip will likely be the closest match to whatever is in their cloud servers.
    edited October 2024
    gatorguy22july2013
     2Likes 0Dislikes 0Informatives
  • Reply 6 of 8
    mpantone said:
    > It is also preparing its infrastructure, making sure that its AI cloud servers can handle the sudden influx of traffic.

    What hardware will Apple Private Cloud Compute be running? Xserve with Apple Silicon? I might want to buy one of those.
    Apple did not mention a marketing name during WWDC 2024; they vaguely referred to some sort of server running Apple Silicon. It is likely that Apple will not market these servers so they likely will only be identified by internal codenames or model numbers.

    It is also unlikely that Apple will provide any details about the Apple Private Cloud Compute devices. Nothing new about that, Apple doesn't discuss their infrastructure. Whether they are binned samples of existing SoCs or something else was not detailed.

    My guess is that Apple currently has a mixture of hardware from third parties as well as their own devices in their AI datacenters. It is probable that the makeup of those servers will change over time and will lean toward devices that feature better performance-per-watt. If that is the case, it is likely that they will increase the percentage of Apple Silicon devices.

    If you want the closest thing to an Apple Private Cloud Compute device, you should probably aim for a Mac Studio with an Ultra SoC. Naturally the Apple Private Cloud Compute devices will be in a form factor better suited for datacenters but the Ultra chip will likely be the closest match to whatever is in their cloud servers.
    I think an earlier article discussed that Apple made a large purchase that included server chips from NVidia for Apple’s AI Cloud Compute servers, and to be followed at some point with Apple silicon that are now in process of being designed in-house. 
    Was it Chip Loder?— or a repeat of some other news source?
     0Likes 0Dislikes 0Informatives
  • Reply 7 of 8
    danoxdanox Posts: 3,887member

     Apple maybe using Nvidia through a third-party when they contract out for something (AI training?), but I don’t think they’re getting anything from Nvidia directly that ship like Intel has sailed away long ago…..

    https://stratechery.com/stratechery-plus/ As far as Nvidia and Apple are concerned, I don’t think they’re using Nvidia for anything. which is the whole point of using Mac Studio M2 Ultras, and by this point, Apple is almost certainly testing getting ready to use in house  Mac Studio M4 Ultra's and beyond.

    https://www.macrumors.com/2024/05/06/apple-building-m2-ultra-and-m4-servers/

    https://www.macrumors.com/2024/05/09/apple-to-power-ai-features-with-m2-ultra-servers/



     0Likes 0Dislikes 0Informatives
  • Reply 8 of 8
    gatorguygatorguy Posts: 24,769member
    danox said:

     Apple maybe using Nvidia through a third-party when they contract out for something (AI training?), but I don’t think they’re getting anything from Nvidia directly that ship like Intel has sailed away long ago…..

    https://stratechery.com/stratechery-plus/ As far as Nvidia and Apple are concerned, I don’t think they’re using Nvidia for anything. which is the whole point of using Mac Studio M2 Ultras, and by this point, Apple is almost certainly testing getting ready to use in house  Mac Studio M4 Ultra's and beyond.

    https://www.macrumors.com/2024/05/06/apple-building-m2-ultra-and-m4-servers/

    https://www.macrumors.com/2024/05/09/apple-to-power-ai-features-with-m2-ultra-servers/



    I think an AppleInsider article a few weeks said Apple had purchased (or were loaned) Google's in-house designed TPU chipsets to install in Apple servers used for their AI training models. Presumably they were better for the task than Nvidia's.
    jellybelly
     0Likes 0Dislikes 1Informative
Sign In or Register to comment.