PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteFacebook confirmed in 2018 that it was forming a silicon team and building a chip, but said the work was not then a primary focus. That early effort should not be confused with any specific later chip: Meta’s subsequent custom-silicon program includes MTIA accelerators and other workload-specific hardware, while continuing to involve outside silicon partners.
What Facebook announced in 2018
At an @Scale event, Facebook vice president of infrastructure Jason Taylor said the company was bringing up a silicon team, working with silicon providers, and building a chip. He also qualified the announcement: chip work was “not our primary focus” at the time. The statement described an early, bounded engineering effort—not a shift away from Facebook’s core business. EE Times reported the announcement in 2018.
The report placed the chip effort alongside Facebook’s support for Glow, its open-source deep-learning compiler. Five chip companies were reported as supporting Glow, illustrating that the early work involved both internal chip development and relationships with external silicon providers.
How Meta’s custom silicon effort developed
In 2023, Meta engineering lead Olivia Wu described a cross-functional organization responsible for designing and developing an in-house machine-learning accelerator. Its work reached beyond chip design to co-design, architecture, verification, implementation, emulation, validation, systems, firmware, and software. Wu recalled seeing a 2018 social-media post from Meta chief AI scientist Yann LeCun seeking someone to help build AI silicon in-house. Engineering at Meta’s account shows how the effort matured into a broad engineering program.
#1 Best Overall
- ✅Powered by 26 Tera-Operations Per Second (TOPS) Hailo-8 AI Processor. 2.5W typical power consumption
- ✅Scalable, enabling simultaneous processing of multi-streams & multi-models
- ✅Enabling real-time, low latency and high-efficiency AI inferencing on the edge devices
- ✅Supports TensorFlow, TensorFlow Lite, ONNX, Keras, Pytorch frameworks
- ✅Supports Linux and Windows. Supports the temperature range of -40°C to 85°C
Meta’s custom hardware has served different workloads over time. The company has discussed custom ASIC designs for AI inference and video transcoding, and its 2023 infrastructure announcement described both a custom AI chip and the in-house-developed MSVP ASIC for video workloads. Those examples should not be collapsed into one product: they represent distinct workload contexts. Meta’s 2019 overview of application-specific accelerators and its 2023 infrastructure announcement provide that background.
What MTIA does, and what Meta reported in 2026
Meta describes MTIA as a family of custom-built chips for its own AI workloads. In a March 11, 2026 update, the company said it had developed and deployed hundreds of thousands of MTIA chips for inference across organic content and ads on its apps. That is Meta’s reported figure, not an independently audited count.
Rank #2
- High-Performance Dual-Core with Ample Memory--- Equipped with a 360MHz dual-core RISC-V processor, 32MB of onboard PSRAM, and 32MB of Flash memory, providing powerful processing capabilities and ample runtime for complex multimedia applications and edge computing.
- Powerful Multimedia Processing Center--- Integrated with a dedicated image processor (ISP), H.264 video encoder, and JPEG codec, perfectly supporting camera input and video processing, making it an ideal choice for developing smart displays, video surveillance, and other projects.
- Hardware-Level Security Protection--- Built-in digital signature, encryption accelerator, and key management unit, providing a one-stop hardware-level security solution from secure boot and data encryption to access control management, ensuring the security of your products and data.
- Full Connectivity Coverage: Wi-Fi 6, Bluetooth, PoE Power Supply--- Onboard with an ESP32-C6 chip, supporting the latest Wi-Fi 6 and Bluetooth 5.0; it also integrates an Ethernet port with PoE functionality, providing high-speed, flexible, and stable network connectivity, and can be powered directly via Ethernet cable, simplifying deployment.
- Rich interfaces and strong expandability--- It provides a MIPI camera/display interface, high-speed USB, SD card slot, microphone/speaker interface and a large number of programmable GPIOs, which greatly facilitates the expansion of external devices and meets the needs of various human-computer interaction and Internet of Things applications. Supports AI Speech Interaction: Allows access to online large model platforms such as ChatGPT, DeepSeek, Doubao, etc.
The same update said Meta was developing and deploying four new generations of chips within the next two years for ranking, recommendations, and generative AI workloads. Meta also said MTIA 300, designed for ranking and recommendation training, was already in production. These are company roadmap and status statements as of March 11, 2026; roadmaps and production status can change. Meta’s update on expanding MTIA also characterizes its custom full-stack solution as more compute-efficient and cost-efficient than general-use chips for its intended purposes. Those efficiency claims are Meta’s assessment, not a published independent benchmark.
Custom chips complement outside silicon
Meta’s custom silicon program does not establish that the company has stopped using chips from other suppliers. Its stated approach is a portfolio: matching accelerators to workloads while sourcing silicon from industry leaders. On April 14, 2026, Meta announced an expanded Broadcom partnership to co-develop multiple generations of MTIA chips. The announcement indicates that outside partnerships remain part of its custom-chip strategy, rather than evidence of an all-internal supply chain. Meta’s Broadcom announcement describes the partnership.
Rank #3
- Powered by 26 Tera-Operations Per Second (TOPS) Hailo-8 AI Processor.
- 2.5W typical power consumption
- Enabling real-time low latency and high-efficiency AI inferencing on the edge devices
- Supports TensorFlow TensorFlow Lite, ONNX, Keras, Pytorch frameworks
- Supports Linux and Windows.
What is—and is not—established about the original chip
The available accounts establish continuity in Meta’s custom-silicon activity, but they do not identify the chip Taylor mentioned in 2018 as MTIA, MSVP, or any particular later design. The original report supports a narrower conclusion: Facebook was forming a silicon team, building at least one chip, and working with silicon providers. The later program is broader and more mature, with deployed accelerators, a multi-generation roadmap, and external co-development.
MTIA is custom infrastructure silicon for Meta’s internal workloads, not a retail chip that consumers can buy. Its significance is in how Meta tailors hardware and supporting software to its own ranking, recommendation, inference, and generative-AI systems.
Quick Recap
Rank #4
- High-Performance AI Processing: The MX3 is designed to handle the most demanding AI computer vision workloads, delivering exceptional performance and efficiency.
- Flexible Integration: The MX3 can be easily integrated into your existing systems via its M.2 M-key form factor and support for Linux operating systems.
- Energy Efficient: The MX3 is designed to provide high performance while minimizing power consumption.
- Comprehensive Software Development Kit (SDK): The MX3 is supported by a comprehensive SDK that simplifies development and deployment.
- Hardware compatability: The MX3 is compatible with the PCI-SIG M.2 M-key 2280 Specification. It can be used with the Raspberry Pi 5 with a M-key 2280 HAT.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
