OpenAI's $20 Billion Cerebras Deal Signals Massive AI Infrastructure Push 
Chinese Robotics Stocks React as Humanoid Robot Marathon Sparks Competition Concerns 
Tesla Earnings Beat Expectations as EV Growth Holds Amid Robotics and AI Shift 
Microsoft Commits $18 Billion to Expand AI and Cloud Infrastructure in Australia 
John Ternus Signals Apple’s Future with Product-First AI Strategy 
Jeff Bezos Eyes $10 Billion Funding Round for AI Venture Project Prometheus 
Samsung Boosts DRAM Supply to Tesla as AI-Driven Memory Demand Surges 
Amazon Expands AI Bet with Up to $25 Billion Investment in Anthropic 
Meta Expands AI Training With Employee Activity Tracking Tools 
Elon Musk Faces French Probe Over X and Grok Amid Rising U.S.-EU Tensions 
Elon Musk Signals Intel 14A Chips for Tesla’s Terafab AI Semiconductor Venture 
SpaceX President Gwynne Shotwell Earns $85.8M as IPO Buzz Grows 
Nvidia Pushes 800V Data Center Power Systems to Boost Efficiency and Cut Costs 
Amazon Stock Rises as Meta Expands AWS Partnership for AI Infrastructure 
SK Hynix Launches 192GB SOCAMM2 Memory for Nvidia’s Next-Gen AI Chips 
Anthropic CEO Meets Trump Officials to Discuss Powerful New AI Model Mythos 
Tesla Q1 Earnings Preview: Robotaxi Delays and SpaceX Merger Speculation Grow 
AMD's strategic shift involves harnessing Samsung's 4nm process for low-end APUs and Radeon GPUs, as high demand saturates TSMC's fabs. Plans aim to bolster competitiveness in gaming markets.
AMD's Strategic Shift: Leveraging Samsung's 4nm Node for Low-End APUs and Radeon GPUs
AMD reportedly plans to use Samsung's 4nm process node to develop low-end APUs and Radeon GPUs. According to @Tech_Reve, AMD plans to use Samsung's 4nm node to power some of its client-side products, such as Ryzen APUs and Radeon GPUs.
Citing "reliable sources," AMD is expected to rely more on Samsung Foundries in the future, possibly because TSMC's fabs are completely booked due to high demand from other segments such as AI.
No specific products are mentioned, but the Samsung 4nm node will be used in low-end Ryzen APUs. AMD will release various APU options this year and next, including Strix Point (Monolithic and Chiplet), Kracken Point, and Fire Range enthusiast chips.
These 4nm chips could be designed for the handheld segment, which has become a popular gaming market and continues to see innovation. With the introduction of Intel's Core Ultra chips into this segment, we can expect AMD to make this market more competitive with future APUs.
Exploring AMD's Semiconductor Strategy: Samsung 4nm for APUs, RDNA 3+ on Strix Point, and RDNA 4 Rumors
Previous information has suggested that AMD will use Samsung 4nm for its Sonoma Valley APUs, which are intended for a next-generation Steam Deck variant.
There is also talk of next-generation Radeon GPUs based on Samsung's 4nm node, as per Wccftech. AMD has only confirmed that its optimized RDNA 3+ architecture will be available on Strix Point APU families.
Certain discrete GPU options based on the same RDNA 3+ architecture based on Samsung's 4nm node may appear, whereas the RDNA 4 lineup uses the TSMC 4nm process node. According to rumors, the AMD RDNA 4 lineup is aimed at the entry-level and mainstream segments, so no high-end GPU offerings are expected in the upcoming lineup.
Tech_Reve previously stated that AMD had planned to produce APUs for the Sony PS5 Pro console, but this was canceled. The last major lineup to use a Samsung node for gaming was NVIDIA's Ampere "GeForce RTX 30" family, which was built on the 8nm node and was distinct from the Ampere HPC lineup, which was based on TSMC's 7nm node.
Photo: Timothy Dykes/Unsplash
