Apple acquired WaveOne, a startup that used machine learning to compress video more efficiently than standard codecs, in a deal that closed in March 2023. WaveOne's website was taken down after the purchase, and its intellectual property and team were absorbed into Apple's video compression and streaming operations. The price was not disclosed.
WaveOne was founded in 2016 by Lubomir Bourdev and Oren Rippel. Before the deal, the company had raised venture funding from investors whose returns are not public. The startup's core innovation was a neural network that compressed video by analyzing the content of each frame, rather than applying a fixed algorithm to every scene.
Apple's interest in the capability is tied to its growing video business, which includes Apple TV+, FaceTime, and the camera systems on iPhones and iPads. More efficient compression means lower bandwidth usage for streaming and higher quality video at the same bitrate, both of which matter to Apple's bottom line and its users' experience.
What WaveOne's AI Compression Did Differently
Traditional video codecs such as H.264 and HEVC compress video by breaking each frame into blocks, then applying mathematical transforms to reduce the amount of data needed to represent those blocks. The same process is applied to every frame regardless of what the frame contains. A frame showing a blue sky and a frame showing a crowd of people are treated identically.
Content-aware allocation
WaveOne's approach was different. Its machine learning model was trained on millions of video frames and learned to recognize which parts of a frame could be heavily compressed without visible loss and which parts needed to retain more detail. A talking head against a plain background, for example, could have the background compressed aggressively while the speaker's face was preserved at higher quality. The model also learned to allocate bits across multiple frames in a sequence, so that a sudden cut to a complex scene could borrow bandwidth from a simpler scene that preceded it.
Claimed performance
The result was a codec that could match the visual quality of H.264 or HEVC at significantly lower bitrates, or produce higher quality video at the same bitrate. In benchmarks published before the acquisition, WaveOne claimed its system reduced file sizes by 40 to 60 percent compared to HEVC at equivalent perceptual quality. Those figures were the company's own and were not independently audited.
Why Standard Codecs Are Hard to Displace
H.264 and HEVC are not just algorithms. They are standards with hardware decoders built into every modern smartphone, television, and streaming device. A new codec that runs only in software cannot compete on power efficiency, because software decoding drains the battery far faster than a dedicated hardware block. That is why Apple, like every other device maker, includes hardware support for H.264 and HEVC in its chips.
Working alongside existing standards
WaveOne's system was designed to work alongside existing codecs, not to replace them. Its neural network could be used as a preprocessing step that optimized video for delivery over a network, with the output then encoded using a standard codec for playback on devices that lacked WaveOne's decoder. Alternatively, the network could be used as a decoder itself, but that required a device with enough compute capacity to run the model in real time.
Apple's silicon advantage
Apple's A-series and M-series chips already contain neural engines capable of running machine learning models at low power. That made Apple one of the few companies that could deploy WaveOne's approach as a full codec replacement, rather than just a preprocessing tool. The neural engine could decode WaveOne-compressed video without the battery penalty that a software-only decoder would incur on a device from another manufacturer.
The Strategic Fit Inside Apple's Video Ecosystem
Apple's video business spans three areas that benefit from better compression. The first is Apple TV+, where every gigabit of bandwidth saved reduces delivery costs for Apple and improves the viewing experience for subscribers on slow connections. The second is FaceTime, where lower bitrate requirements mean higher resolution calls over cellular networks. The third is the iPhone camera, where video files recorded in 4K or ProRes consume large amounts of storage. More efficient compression could let users record longer video at the same quality, or higher quality video at the same file size.
Exclusive intellectual property
The deal also gives Apple control over a compression capability that its competitors cannot license. Standard codecs are available to any company that pays the patent royalties. WaveOne's system, by contrast, is proprietary and covered by patents filed by Bourdev and Rippel. Apple can use it exclusively in its own products, or license it to others on its own terms.
Integration timeline
Apple has not announced which product will first include WaveOne's work. The integration is likely to take years, because compression algorithms must be tested thoroughly before they are deployed at scale. A bug that corrupts video on millions of iPhones is a different order of problem from a bug in a photo editing app.
What Happened to WaveOne's Team and Technology
WaveOne's website was shut down after the acquisition closed, which is typical for startups absorbed into a larger company. The domain now redirects to Apple's corporate site. The company's intellectual property and team were folded into Apple's video compression group, which is part of the broader software engineering organization.
Founders and staff
Whether Bourdev and Rippel remained at Apple after the deal is not known. Founders of acquired startups often stay for a retention period of one to four years, but Apple does not disclose the employment status of individual acqui-hire targets. The same uncertainty applies to the rest of WaveOne's employees. The company had a small team, typical for a startup that had raised only a modest sum over seven years, but the exact headcount was never made public.
Assets transferred
The transaction was not a talent grab alone. WaveOne had published research papers and filed patents on its compression techniques, and those intellectual property assets were part of the deal. Apple now owns the patents, the trained neural network models, and the training data and infrastructure used to create them.
The Financial Picture: What We Know and What We Do Not
The purchase price Apple paid for WaveOne was not disclosed. Acquisitions of this size, where the target had raised less than $10 million in venture funding, typically fall below the threshold that would require Apple to disclose the price in a regulatory filing. Estimates from outside analysts vary widely, from a low of $20 million to a high of $100 million, but those are guesses.
Investor returns
WaveOne's investors included a mix of venture capital firms and angel investors. The company had raised a seed round and a Series A. Their return on the deal depends on the acquisition price and on any liquidation preferences written into the investment terms. If the deal was structured as an acqui-hire with a low upfront payment and retention bonuses for employees, the investors may have recovered only their initial investment or less. If the deal was a traditional acquisition with a meaningful premium, the investors made a profit. Neither scenario is confirmable from public information.
Apple's pattern
Apple has a history of acquiring small technology companies without announcing the terms. The WaveOne deal fits that pattern. The company did not issue a press release about the purchase, and the news emerged only when WaveOne's domain changed and former employees updated their LinkedIn profiles.
Timeline and Status of the Deal
The acquisition was confirmed in March 2023, and the deal closed in the same month. The exact date of the closing is not public, but the shutdown of WaveOne's website and the migration of its team to Apple happened within weeks of the announcement.
Current status
As of April 2023, Apple has not disclosed any product plans related to WaveOne's intellectual property, and no device released since the deal has publicly credited WaveOne's compression. The capability could appear in a future iPhone, iPad, or Apple TV software update, or it could be used in Apple's server infrastructure for video delivery without ever being marketed to consumers.
Regulatory review
The deal did not require regulatory approval in any major jurisdiction. Apple's purchase of a small startup with no dominant market position does not trigger antitrust review in the United States or the European Union. That may change in the future if regulators begin to scrutinize larger companies' purchases of small AI startups, but as of the deal's closing date, the transaction faced no regulatory hurdles.
Key Facts
- Acquirer: Apple Inc.
- Target: WaveOne Inc.
- Founded: 2016 by Lubomir Bourdev and Oren Rippel
- Acquisition confirmed: March 2023
- Deal status: Closed successfully in March 2023
- Funding raised pre-acquisition: Approximately $9 million, per company disclosures and investor filings
- Purchase price: Not disclosed; Apple does not report terms for acquisitions below its materiality threshold
- Technology: AI-driven video compression using machine learning
- Post-acquisition fate: Website shut down; technology and team absorbed into Apple's video compression group
WaveOne vs. Standard Video Codecs
| Feature | H.264 (AVC) | HEVC (H.265) | WaveOne (pre-acquisition) |
|---|---|---|---|
| Year introduced | 2003 | 2013 | 2020 (research published) |
| Hardware decode support | Universal | Wide, but not universal | None (software only) |
| Patent licensing | Required, multiple pools | Required, complex pools | Proprietary to Apple post-acquisition |
| Compression vs. H.264 | Baseline | ~40% bitrate reduction | Claimed 40-60% reduction vs. HEVC |
| AI/ML based | No | No | Yes |
Frequently Asked Questions
Did Apple buy WaveOne for its technology or its team?
Both. WaveOne had patented AI compression technology and a small team of engineers with deep expertise in machine learning and video coding. Apple acquired the intellectual property and the people.
Will WaveOne's technology appear in the next iPhone?
Apple has not announced any product plans. The integration is likely to take years, and the technology could appear first in server-side video processing rather than in a device.
How much did Apple pay for WaveOne?
The purchase price was not disclosed. Estimates from outside analysts range from $20 million to $100 million, but those are not confirmed. For official financial disclosures, refer to Apple's investor relations page at investor.apple.com.
What happened to WaveOne's founders?
It is not known whether Lubomir Bourdev and Oren Rippel remained at Apple after the acquisition. Apple does not disclose the employment status of individual acqui-hire targets.
