AMD and Cerebras Team Up to Deliver Faster AI Inference Chips

AMD just dropped a stack of AI announcements, and the one that matters most for traders is the partnership with Cerebras. The two companies are combining AMD's Helios GPU rack with Cerebras' wafer-scale chip to build an ultra-low latency AI inference platform. Translation: faster AI responses for enterprise and hyperscale customers who can't afford to wait. Expect it on Cerebras Cloud in the second half of 2026.

But AMD didn't stop there. CEO Lisa Su also put the spotlight on the new EPYC "Venice" server CPU. AMD claims Venice outperforms Arm-based server chips by a significant margin and beats Nvidia's CPU offering by 20%. That's a bold claim in a market where Nvidia owns the mindshare. Venice is already in full production, with systems expected to ship in Q4. Demand is described as strong.

Read more TotalEnergies Logs Best Profit in Three Years on Iran War Oil Spike →

Here's the tradeable angle: AMD shares were down 3% on the day around $535, but that's the key detail — they were well off session lows near $525. More importantly, the price clawed back above its 100-hour and 200-hour moving averages near $532. Sellers had their shot below those levels and couldn't hold it. That's a bullish signal. Cerebras stock popped over 4% to around $218, having already reclaimed its own key moving averages earlier in the week. Next upside target for Cerebras sits near $245.

The bigger picture here is AMD's deliberate push to compete across the entire AI stack — from GPU training rigs to inference platforms to server CPUs. Every layer adds revenue and reduces dependence on any single product line. That's a long game, and partnerships like this one with Cerebras accelerate it without requiring AMD to build everything in-house.

Watch those moving averages on both tickers. As long as AMD holds above $532 and Cerebras holds above $202, the technical bias favors the bulls. A slip back below those levels flips the script fast. Continue reading at Forexlive.