Hello, Thanks for signing up for MarketBeat Daily Ratings—we’re excited to have you on board. Every weekday, you’ll get a curated summary of new “Buy” and “Sell” ratings from Wall Street’s top-rated analysts, the latest stock news, and bonus investing content—all delivered straight to your inbox. You’re just two quick steps away from completing your sign-up: 1. Make sure our emails go to your inbox Gmail users:
Mobile: Tap the three dots (…) in the top right and select Move to Inbox or Move to Primary
Desktop: Click the folder icon at the top and select Move to Inbox or Primary Apple Mail users:
Tap our email address at the top (next to From: on mobile), then select Add to VIP Other providers:
Reply to this message and add newsletters@analystratings.net to your contacts 2. Confirm your subscription Click this link to confirm your subscription. This verifies your account and ensures you receive your newsletters without interruption instead of getting stuck in your spam filter. Confirm your subscription here. After you confirm, feel free to download our popular free report, "7 Stocks to Buy and Hold Forever" with this link. Thanks again for subscribing—we look forward to being part of your investing journey. 
Matthew Paulson
Founder and CEO, MarketBeat. P.S. If you didn’t mean to subscribe, no problem—you can unsubscribe here.
Tuesday's Featured Story Why Meta's AI Chip Announcement Has Broadcom Investors Paying AttentionReported by Leo Miller. First Published: 3/18/2026. 
Key Points - Meta publicly confirmed Broadcom as its custom chip partner for the first time, removing lingering doubts about one of Broadcom's most important AI relationships.
- The MTIA chip roadmap is expanding from ranking and recommendation into generative AI inference—a workload many expect to dominate AI compute by decade's end.
- One notable gap in Meta's announcement: no generative AI training chip, lending weight to reports that its most ambitious custom silicon project has been shelved for now.
- Special Report: Have $500? Invest in Elon's AI Masterplan
In a recent announcement, the Magnificent Seven tech giant Meta Platforms (NASDAQ: META) unveiled four custom artificial intelligence (AI) chips. This follows semiconductor design behemoth Broadcom's (NASDAQ: AVGO) earnings report, where CEO Hock Tan directly addressed Meta. Meta is now giving a wink back to Broadcom in a development that has clear positive implications for AVGO. There are, however, some negatives worth noting. What does this mean for Broadcom going forward? META and AVGO Confirm MTIA Partnership For some time, market watchers and analysts have believed Meta is one of Broadcom's custom AI processor customers. Broadcom itself had not publicly named Meta until now. On Broadcom's Q1 2026 call, Tan said, “Contrary to recent analyst reports, Meta's custom accelerator MTIA road map is alive and well. We're shipping now.” MTIA, which stands for Meta Training and Inference Accelerator, is a custom chip family developed in collaboration with Broadcom. Tan's comment followed reports that Meta halted development of its most advanced custom AI training chip, codenamed Olympus. Meta explicitly named Broadcom in its MTIA press release: “Meta Training and Inference Accelerator (MTIA), our family of homegrown AI chips developed in close partnership with Broadcom, has remained and will continue to be an important part of Meta's AI infrastructure strategy.” Markets had generally accepted the partnership; the mutual acknowledgment from both companies removes lingering doubt. The Good: META-AVGO Partnership Expands Into GenAI The title of Meta's post, “Four MTIA Chips in Two Years: Scaling AI Experiences for Billions,” reinforces the bullish case Broadcom outlined on its earnings call. Hock Tan noted that many customers are developing two custom chips with Broadcom per year—the same pace Meta described—lending credibility to Broadcom's claims about deepening customer relationships. Meta is deploying MTIA across a range of uses, including training and inference for its ranking and recommendation (R&R) models. Training develops more capable models; inference deploys those models to answer queries and perform tasks. R&R training and inference help Meta deliver increasingly engaging content and more targeted ads to its users. Meta has about 3.5 billion users across its apps—more than 40% of the world's population—so its needs for these workloads are immense, and it is relying on Broadcom to meet them. The MTIA series extends beyond R&R. Meta plans to use MTIA 450 and MTIA 500 to enable GenAI inference, with mass deployments expected in 2027. GenAI inference likely refers to chatbot queries, video and image generation, and AI business agents in WhatsApp. Experts do not consider Meta's LLaMA models state-of-the-art compared with ChatGPT, Claude, or Gemini. That said, they can still be useful and monetizeable. Meta AI already has over 1 billion users, which creates a significant opportunity. For Broadcom, MTIA's expansion from core R&R workloads into GenAI inference is a clear positive. As Broadcom supports both Meta's established and emerging workloads, the logical outcome is increased chip demand. The Bad: META-AVGO GenAI Training Chip Takes a Back Seat Meta's announcement did not include a GenAI training chip, which lends weight to reports that Meta has scaled back Olympus development. Meta CFO Susan Li recently told the Morgan Stanley Technology Conference that Meta “expects” and is “hopeful” it can expand its use of custom silicon to train AI models “eventually.” That's a negative for Broadcom, which likely would have co-developed Olympus with Meta. Meta hasn't abandoned training-chip ambitions entirely, but any revenue Broadcom might have expected from that project could be delayed significantly. AVGO and META: Powering the Growth in AI Inference Overall, Meta's relationship with Broadcom is now fully substantiated and appears to be expanding outside of GenAI training. Importantly, many expect inference to overtake training as the dominant AI workload in the coming years. McKinsey predicts that inference will grow at a compound annual growth rate of about 35% over the next five years, accounting for more than half of AI compute by 2030. That outlook supports Broadcom's case: its deepening relationship with Meta—especially around inference workloads—should be an important driver of future growth. | 
Post a Comment
Post a Comment