Apple Develops Breakthrough Method for Running LLMs on iPhones

Apple GPT in your pocket? It could be a reality sooner than you think. Apple AI researchers say they have made a key breakthrough in deploying large language models (LLMs) on iPhones and other Apple devices with limited memory by inventing an innovative flash memory utilization technique.

siri symbol iphone

LLMs and Memory Constraints

LLM-based chatbots like ChatGPT and Claude are incredibly data and memory-intensive, typically requiring vast amounts of memory to function, which is a challenge for devices like iPhones that have limited memory capacity. To tackle this issue, Apple researchers have developed a novel technique that uses flash memory – the same memory where your apps and photos live – to store the AI model's data.

Storing AI on Flash Memory

In a new research paper titled "LLM in a flash: Efficient Large Language Model Inference with Limited Memory," the authors note that flash storage is more abundant in mobile devices than the RAM traditionally used for running LLMs. Their method cleverly bypasses the limitation using two key techniques that minimize data transfer and maximize flash memory throughput:

  1. Windowing: Think of this as a recycling method. Instead of loading new data every time, the AI model reuses some of the data it already processed. This reduces the need for constant memory fetching, making the process faster and smoother.
  2. Row-Column Bundling: This technique is like reading a book in larger chunks instead of one word at a time. By grouping data more efficiently, it can be read faster from the flash memory, speeding up the AI's ability to understand and generate language.

The combination of these methods allows AI models to run up to twice the size of the iPhone's available memory, according to the paper. This translates to a 4-5 times increase in speed on standard processors (CPUs) and an impressive 20-25 times faster on graphics processors (GPUs). "This breakthrough is particularly crucial for deploying advanced LLMs in resource-limited environments, thereby expanding their applicability and accessibility," write the authors.

Faster AI on iPhone

The breakthrough in AI efficiency opens new possibilities for future iPhones, such as more advanced Siri capabilities, real-time language translation, and sophisticated AI-driven features in photography and augmented reality. The technology also sets the stage for iPhones to run complex AI assistants and chatbots on-device, something Apple is already said to be working on.

Apple's work on generative AI could eventually be incorporated into its ‌Siri‌ voice assistant. Apple in February 2023 held an AI summit and briefed employees on its large language model work. According to Bloomberg, Apple is aiming for a smarter version of Siri that's deeply integrated with AI. Apple is planning to update the way that ‌Siri‌ interacts with the Messages app, allowing users to field complex questions and auto-complete sentences more effectively. Beyond that, Apple is rumored to be planning to add AI to as many Apple apps as possible.

Apple GPT

Apple is reportedly developing its own generative AI model called "Ajax". Designed to rival the likes of OpenAI's GPT-3 and GPT-4, Ajax operates on 200 billion parameters, suggesting a high level of complexity and capability in language understanding and generation. Internally known as "Apple GPT," Ajax aims to unify machine learning development across Apple, suggesting a broader strategy to integrate AI more deeply into Apple's ecosystem.

As of the latest reports, Ajax is considered more capable than the earlier generation ChatGPT 3.5. However, it's also suggested that OpenAI's newer models may have advanced beyond Ajax's capabilities as of September 2023​.

Both The Information and analyst Jeff Pu claim that Apple will have some kind of generative AI feature available on the ‌iPhone‌ and iPad around late 2024, which is when iOS 18 will be coming out. Pu said in October that Apple is building a few hundred AI servers in 2023, with more to come in 2024. Apple will reportedly offer a combination of cloud-based AI and AI with on-device processing.

Popular Stories

iPhone 15 Pro Cameras

iPhone 17 Pro Max Will Be First Model to Feature Three 48MP Cameras

Thursday July 11, 2024 12:20 am PDT by
Next year's iPhone 17 Pro Max will feature an upgraded 48-megapixel Tetraprism camera for enhanced photo quality and zoom functionality, according to Apple analyst Ming-Chi Kuo. In his n-iphone-tetraprism-upgrade-ca62dd37e364">latest investor note published to Medium, Kuo said the key specification change would be a 1/2.6" 48MP CIS sensor, up from the 1/3.1" 12MP sensor expected to be used...
Beyond iPhone 13 Better Blue Face ID Single Camera Hole

10 Reasons to Wait for Next Year's iPhone 17

Monday July 8, 2024 5:00 am PDT by
Apple's iPhone development roadmap runs several years into the future and the company is continually working with suppliers on several successive iPhone models simultaneously, which is why we sometimes get rumored feature leaks so far ahead of launch. The iPhone 17 series is no different – already we have some idea of what to expect from Apple's 2025 smartphone lineup. If you plan to skip...

Apple's AirPods Pro 2 vs. Samsung's Galaxy Buds3 Pro

Saturday July 13, 2024 8:00 am PDT by
Samsung this week introduced its latest earbuds, the Galaxy Buds3 Pro, which look quite a bit like Apple's AirPods Pro 2. Given the similarities, we thought we'd compare Samsung's new earbuds to the AirPods Pro. Subscribe to the MacRumors YouTube channel for more videos. Design wise, you could potentially mistake Samsung's Galaxy Buds3 Pro for the AirPods Pro. The Buds3 Pro have the same...
primeday2020 feature3

The Best Early Prime Day Deals on Apple Products

Saturday July 13, 2024 6:23 am PDT by
Amazon is soon to be back with its annual summertime Prime Day event, lasting for just two days from July 16-17. As it does every year, Prime Day offers shoppers a huge selection of deals across Amazon's storefront, and there are already many deals you can get on sale ahead of the event. Note: MacRumors is an affiliate partner with Amazon. When you click a link and make a purchase, we may...

Top Rated Comments

grantishere Avatar
8 months ago
“Siri, when’s Apple GPT going to be released?”

“Here are the neatest pizza restaurants. Would you like directions?”
Score: 66 Votes (Like | Disagree)
Infodataset Avatar
8 months ago
Apple is so clever to avoid adding more ram ?
Score: 54 Votes (Like | Disagree)
Abazigal Avatar
8 months ago

Does this mean Siri will finally be able to “call my wife on iPhone” again after losing that ability around 2-3 years ago?
At least it didn't respond with "which one?" ?
Score: 34 Votes (Like | Disagree)
wdfly Avatar
8 months ago
LOL innovative invention of swapping memory to storage…… maybe they can call it something cool like “cache”.
Score: 23 Votes (Like | Disagree)
subjonas Avatar
8 months ago
Sounds like some significant headway. I’d definitely like as much AI done on device as possible.

Apple will probably always be behind with their LLM as long as they prioritize privacy, which I’m very ok with. But just like with big phones, they will bend if there is enough market pressure, which I suspect may eventually be the case.
Score: 21 Votes (Like | Disagree)
contacos Avatar
8 months ago
I'll be impressed when Siri is finally multi-lingual. I cannot believe Siri was released with iPhone 4S and it still cannot do it. It could be as simple as being "aware" what language you usually type in with contact A and automatically use Siri in language X to communicate with this person while using language Y with contact Y. All they did was "this text is in another language, do you still want me to read [the gibberish nonsense]" with iOS 16
Score: 18 Votes (Like | Disagree)