Get the newsletter
Google · AI / ML · Filed Apr 29, 2025 · USPTO published Aug 20, 2026 · Aug 20 release

Google Patents a Method to Store AI Model Data in Fewer Bits

Google is patenting a method that lets AI models train using number formats that don't follow the usual rules, potentially squeezing more performance out of the same hardware by picking exactly the right level of numerical precision.

Binary bit sequence layout illustrating exponent and mantissa bits for floating-point storage formats. Drawing from patent filing US 2026/0244987 A1.
Binary bit sequence layout illustrating exponent and mantissa bits for floating-point storage formats.All 17 drawings ↓
Publication number
US 2026/0244987 A1(US20260244987A1)
Applicant
Google LLC
Filed · Published
Apr 29, 2025 · Aug 20, 2026
Inventors
Aditya Binodkumar Agrawal, Gil Joseph Tabak, Narges Shahidi
Status when we published
Waiting for an examiner
Document
20 claims
US classification
706/12
Parent application
Claims priority from a provisional application 63758899 (filed Feb 14, 2025)

What Google's custom AI number format actually does

Every time a company trains a large AI model, thousands of chips crunch through billions of numbers, and those numbers have to be stored in a specific format. The format determines how precise each number is, and precision costs memory and power.

Right now, AI systems almost always store numbers in formats sized at standard powers of two: 8 bits, 16 bits, 32 bits. Google's patent covers a method that breaks that rule. It trains AI models using number formats with any combination of exponent and mantissa bits, even if the total doesn't land on a power of two. Think of it like choosing a ruler marked in thirds instead of halves, because thirds happen to be more useful for what you're measuring.

The idea is that for certain types of AI parameters, a non-standard format might capture exactly the right range and precision without wasting bits on detail the model doesn't need. That could mean faster training, less memory use, or both, without sacrificing how well the model performs.

From the filing · CLAIM 1
obtaining, by a computing system comprising one or more computing devices, a first machine learning model comprising a first plurality of parameters having a first binary format comprising X exponent bits and Y mantissa bits, wherein a precision of the first binary format is not an integer power of two; …

Translation: The system loads an AI model built on an unusual number format that does not follow standard powers of two.

How the non-power-of-two bit format works in training

Every number stored inside an AI model has a binary format: a fixed number of bits split between the exponent (which sets the range, how large or small the number can be) and the mantissa (which sets the precision, how finely the number can be expressed). Standard formats like FP16 or BF16 have bit counts chosen because they fit neatly into hardware registers sized at powers of two.

This patent claims a training method where the model's parameters use a binary format whose total precision is not an integer power of two. So instead of being forced into 8-bit or 16-bit buckets, a format might use, say, 5 exponent bits and 6 mantissa bits (11 bits total) if that happens to suit the data better.

The training loop itself works conventionally otherwise:

  • Input data is fed into the model each iteration.
  • The model generates an output using parameters stored in this custom format.
  • The parameters are updated based on the output, a process called backpropagation.

The patent also covers storing optimizer parameters (the extra numbers that guide how quickly and smoothly the model learns) in similarly non-standard formats. The core claim is that the format choice is decoupled from hardware register-size conventions, letting engineers pick whatever bit layout minimizes error for a given model type.

From the filing · THE ABSTRACT
… generating, by the computing system based at least in part on the input values and based at least in part on the parameters having the binary format, an output.

Translation: The system runs the model using these custom sized bits to produce a result.

What this means for AI chip efficiency and training costs

For people who actually use AI products, the payoff from research like this shows up as lower costs and faster model updates. When companies can train models more efficiently, they can iterate more often, which means the AI assistant or image generator you use gets updated more frequently and with less computing expense passed on to you.

The deeper stakes are about who controls the numerical standards that underpin AI hardware. Google designs its own AI chips (the TPU line), and controlling the number formats those chips support is a way to keep competitors from easily replicating the efficiency gains. This kind of infrastructure-level patent sits in the same stream of new Big Tech patents around AI chip efficiency that will shape which companies can afford to train the largest models over the next decade.

Google's tenth filing we've tracked since July in the AI chip wars adds to earlier work like the noise-cancellation chip and the power-adjusting chip.

Source: US 2026/0244987 A1 at the USPTO (PDF). Full patent text and figures are in the official publication.

Get our take in your Top Stories

Liked this breakdown? Add Patentlyze as a preferred source on Google, and our plain-English take shows up more often in your Top Stories the next time Google patent news breaks.

There are more where this came from

We read the new patents from Big Tech, carmakers and more and send you the ones worth knowing. Plain English, free, every week.

The drawings · 17 sheets

17 drawing sheets from US 2026/0244987 A1 · click any drawing to enlarge

Binary storage bit layout, drawing sheet 1 from US 2026/0244987 A1
Sheet 1 Binary storage bit layout
Computing system overview, drawing sheet 2 from US 2026/0244987 A1
Sheet 2 Computing system overview
Training parameter update flow, drawing sheet 3 from US 2026/0244987 A1
Sheet 3 Training parameter update flow
Optimizer parameter determination, drawing sheet 4 from US 2026/0244987 A1
Sheet 4 Optimizer parameter determination
Processor device data path, drawing sheet 5 from US 2026/0244987 A1
Sheet 5 Processor device data path
Machine learning model training steps, drawing sheet 6 from US 2026/0244987 A1
Sheet 6 Machine learning model training steps
Parameter format selection process, drawing sheet 7 from US 2026/0244987 A1
Sheet 7 Parameter format selection process
Training instance processing steps, drawing sheet 8 from US 2026/0244987 A1
Sheet 8 Training instance processing steps
Show the other 9 sheets
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 9
Sheet 9
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 10
Sheet 10
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 11
Sheet 11
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 12
Sheet 12
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 13
Sheet 13
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 14
Sheet 14
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 15
Sheet 15
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 16
Sheet 16
Google Patents a Method to Store AI Model Data in Fewer Bits patent drawing, sheet 17
Sheet 17
Patent filing page

More in AI training & infrastructure

All 324 AI training & infrastructure patents →

Watchlists this filing belongs to

All watchlists →