Bloomberg News‘ Mark Gurman is reporting that there is a strong possibility that the chip in the new iPad Pro will be the M4, not the M3 and it will be Apple’s first major AI-driven device.
Mark Gurman for Bloomberg News:
I believe Apple will position the tablet as its first truly AI-powered device — and that it will tout each new product from then on as an AI device. This, of course, is all in response to the AI craze that has swept the tech industry over the last couple years.
By introducing the new iPad Pro ahead of its Worldwide Developers Conference in June, Apple could lay out its AI chip strategy without distraction. Then, at WWDC, it could focus on how the M4 chip and new iPad Pros will take advantage of the AI software and services coming as part of iPadOS 18 later this year. I fully expect Apple to position the A18 chip in the iPhone 16 line as built around AI as well.
To be fair, though, these new products aren’t engineered and developed entirely around AI. This is partly about marketing…
The company is holding a “Let Loose” event on May 7 to introduce the new iPad Pro, iPad Air, Magic Keyboard and Apple Pencil.
Support MacDailyNews at no extra cost to you by using this link to shop at Amazon.
MacDailyNews Take: Yup.
It will take years and billions of dollars to catch up just to where GenAI leaders (OpenAI, Microsoft, Alphabet, etc.) are today.
So, the only solution is to partner with [OpenAI, Microsoft, Alphabet, etc.] for the real GenAI stuff while pretending (marketing) really hard that some on-device AI Apple has whipped up in a few months is “insanely great Apple innovation” that’s at the heart of Apple’s 2024’s AI announcements when it’s really just an adjunct. Apple will tout their homegrown on-device AI and act like it’s powering everything when, in reality, it’s Google (or whichever is licensed) that’s powering most of it. Watch Apple make a big show of its on-device AI at WWDC and run many ads touting it from June onwards.
Apple hopes to buy time for the data center buildouts and investments that will be required for them to someday own their own AI technology and not have to license it. – MacDailyNews, April 1, 2024
The potential for an iPhone 16 super cycle depends on how well Apple marketing sells its GenAI bandaid (licensing from Google, Baidu, etc.) as it works (for years, likely) to catch up with its own generative AI solution. – MacDailyNews, April 11, 2024
Please help support MacDailyNews. Click or tap here to support our independent tech blog. Thank you!
Support MacDailyNews at no extra cost to you by using this link to shop at Amazon.
Can someone educate us here? If Apple processes on device, then do they need to overhaul their data centers with AI chips that need to be cooled with massive amounts of energy? Especially if they are licensing someone else’s AI? While others are investing billions upon billions of dollars in infrastructure, will Apple be able to leverage the commoditization of these services? I’m not sure MDN’s negative perspective is correct. Apple doesn’t own Search and they seem to do just fine. Isn’t this just Search on steroids? Sorry for my naïveté.
2 separate issues:
Limited on-device Apple AI will be baked into the M4 / Neural Engine.
Really capable GenAI needs the cloud. Apple missed AI (screwing around with electric cars and headsets) and will have to build out necessary infrastructure so they can do their own. They will have to license it in the meantime. Apple prefers to own their own primary tech.
The only reason why Apple doesn’t do its own search engine is because Alphabet (Google) pays them not to, but the U.S. gov’t may kill that deal.
I’m aware that Google pays Apple to not do search and that they like to own fundamental/essential IP tech. But they use TSMC for wafer and chip fabrication, so sometimes it’s not cost effective or strategic to do such a thing. They focus on designing the chip for their own use. You can leverage others work more cost effectively while focusing on the piece that helps you the most. I mean does NVIDIA have a generative ai solution. I don’t think so. They are focusing on the chips and servers that enable others to do it and laughing all the way to the bank with the hype and hysteria. I’ve used chatbots and honestly can’t say they have ever been useful to me as a consumer. Maybe they will.
Again, pardon my naïveté. Just such a fascinating time to be living in.
Model M4 chip. Model T software.
m4 yawn, M series chips are mobile designed chips focused on power draw and mobile performance (integrated cpu, gpu) Apple needs to design true workstation data center chips.
Nvidia Blackwell design is a decade ahead of apple.
Here is what the bleeding edge looks like. Shipping today and ready for Nvidia B series. Apple still with the AI smoke and mirrors.
Nice ovens!
Yep, in the near term Apple may be using them directly or indirectly for the AI tasks too heavy to do on-device.
The Nvidia Blackwell processor is primarily intended for data center and cloud computing applications, especially for tasks that require substantial computational power and memory for AI processing. It is not designed for desktop or laptop environments. The Blackwell platform, including the GB200 Superchip and various server configurations like the GB200 NVL72, is tailored for large-scale, high-performance computing environments typical of data centers. These configurations support the demanding needs of generative AI supercomputing and are aimed at enterprise-level deployments.
It is not particularly useful to compare Nvidia’s Blackwell processors with Apple’s M series chips, such as the M3 or M4. These two families of chips serve very different purposes and are designed for entirely different computing environments.
You might be interested in Qualcomm’s new Snapdragon chip though.
Qualcomm has introduced a new Snapdragon chip that aims to compete with Apple’s M3 series in the realm of AI and overall performance. This new chip, called the Snapdragon X Elite, is designed to be a high-performance ARM processor for PCs, boasting significant improvements in multi-core performance. Qualcomm claims that the Snapdragon X Elite offers a 21% faster multi-core performance compared to Apple’s M3 chip and also competes well with Intel’s latest processors (MacRumors).
Qualcomm’s Snapdragon chips, including the latest Snapdragon X Elite, feature a Neural Processing Unit (NPU) similar to Apple’s Neural Engine in the M series chips. The Snapdragon X Elite incorporates Qualcomm’s Hexagon NPU, which is significantly optimized for AI tasks. This NPU supports running large AI models directly on the device, enhancing performance for AI-driven features without needing to rely on cloud processing.
https://www.nvidia.com/en-us/data-center/technologies/blackwell-architecture/
Good reading. There are more and more articles like this exposing the fraud of the ai industry. Sure some of it will succeed but VCs get paid a fee on the funds they raise regardless of the success or failure of their investments. And as this article points out, MS and Amazon get their investments back as income for cloud services.
https://apple.news/AtKj3W2wNRCOq502Rb6QFLA
Apple has all the tech necessary to produce this 1980s fantasy now…
So why are they holding back? Nothing i’ve seen out of Siri or any software Apple has offered has even come close to this. AI doesn’t fully capture it. It’s also about making connections with information already inside your own devices (notes, files, contacts, photos, etc.). Machine learning of specific commands should allow for this.
The most basic thing I’ve wanted from Siri is to have it read the contents of a specific note or text document. “Hey Siri, read my To-Do list note”, “Hey Siri, read my Russia-Ukraine War Pages document back to me”. I STILL don’t think this will be anywhere close to happening, let alone sequencing a string of requests into a coherent whole.