Microsoft’s Maia 200 AI Chip Aims to Outpace Nvidia Globally

Microsoft launches Maia 200 AI accelerator — 140B transistors, 216 GB HBM3e, up to 10 petaFLOPS and 30% better performance-per-dollar to power GPT‑5.2 and Microsoft services.

  • Microsoft introduced the Maia 200 AI accelerator, built to improve inference performance across its cloud and enterprise services.
  • The chip packs 140 billion transistors, 216 GB of HBM3e memory and delivers up to 10 petaFLOPS; it is manufactured on TSMC’s 3-nanometer process.
  • Scott Guthrie said Maia 200 offers “30% better performance per dollar”, making it the company’s most economical AI processor to date.
  • Maia 200 will run OpenAI’s GPT-5.2 models and Microsoft services such as Microsoft 365 Copilot and Foundry apps, and Microsoft says it is cheaper and more efficient than rivals’ custom chips.
  • The processor is already deployed in Microsoft data centers; shares rose about 1% and Raymond James reaffirmed an Outperform rating with a $600 price target (about 34% above a roughly $470 price).

On Jan. 26, 2026, Microsoft launched the Maia 200, a next-generation AI accelerator designed to speed inference across its cloud and enterprise platforms. The chip includes 140 billion transistors, 216 GB of HBM3e memory and delivers up to 10 petaFLOPS, and it is built on TSMC’s 3-nanometer process. Microsoft says the design aims to lower power use and operating costs for high-volume AI workloads.

- Advertisement -

Scott Guthrie, the company’s Executive Vice President, described the processor as more economical, saying it has “30% better performance per dollar” than current-generation technology. Microsoft plans to use Maia 200 to run OpenAI’s GPT-5.2 models and its own services, including Microsoft 365 Copilot and Foundry apps.

The move follows industry trends of major cloud providers building custom silicon to reduce reliance on third-party chips. Microsoft said Maia 200 is cheaper to run and more efficient than the custom AI chips used by some cloud rivals, helping reduce dependence on NVIDIA technology. The company emphasized cost and energy savings for high-volume inference across Azure.

Investors reacted to the chip’s deployment in Microsoft data centers, with shares rising about 1% at the close. Raymond James reaffirmed an Outperform rating and a $600 price target, which implies roughly a 34% upside from a trading price near $470.

✅ Follow BITNEWSBOT on Telegram, Facebook, LinkedIn, X.com, and Google News for instant updates.

- Advertisement -

Previous Articles:

- Advertisement -

Latest News

$50M AAVE Swap Yields $36K Despite Warning

A trader lost nearly $50 million on Thursday after swapping that amount of USDT...

Teamsters Threaten to Block Paramount-WBD Merger

The International Brotherhood of Teamsters opposes the Paramount Skydance-Warner Bros. Discovery merger without enforceable...

STRC Sales Surge, Eye Record Single-Day Bitcoin Buy

A community dashboard tracking Strategy's STRC sales suggests March 12, 2026 could see the...

SEC’s Peirce Urges Simpler Rules Amid Tokenization Talks

SEC Commissioner Hester Peirce argues regulators should avoid micromanaging markets and consider simplifying disclosure...

Rust VENON Malware Targets Brazilian Banking Apps

A new Rust-based banking Trojan named VENON is targeting Brazilian users, departing from the...

Must Read

Top 9 VPNs That Accept Bitcoin And Crypto

CyberGhost | FastVPN | TorGuard | Private Internet Access | ExpressVPN | NordVPN | Private VPN | SurfShark | AirVPN | Why Buy VPN...