AMD
AMD CEO Lisa Su just hiked her agentic CPU market forecast to $220 billion in three months, proving the chipmaker is capturing insane underlying demand.
AMD just secured a massive 5 billion dollar deal with Anthropic to deploy two gigawatts of chips and use AI models to break its software bottleneck.
Hyperscalers like Google are torching over $200 billion on AI infrastructure, so buy the hardware makers directly catching that massive cash firehose.
Nvidia delaying its flagship AI server rack by 12 months into 2028 hands AMD a massive opening to steal market share with its own upcoming hardware.
NVIDIA coerces neoclouds away from AMD GPUs, blocking AMD's datacenter ramp at the operator level
NVIDIA's alleged suppression of AMD GPU adoption at neoclouds is anticompetitive; antitrust action could unlock AMD's data center market share
AMD's acquisition of MEXT, which makes flash behave like DRAM, positions it to win AI inference by routing around the memory tax that is crippling competitors
Low-latency AI inference chips are the critical bottleneck for computer-use agents; AMD captures the CPU demand surge as agentic AI scales
AI-driven agentic workloads are tripling the server CPU TAM by 2030, making AMD the prime beneficiary of the CPU supercycle.
CPU TAM forecasts keep doubling as agentic AI shifts compute ratios; AMD is directly named as a CPU demand beneficiary reporting actual demand spikes
CPU server TAM forecasts being rapidly doubled by AMD, Nvidia, and Bernstein as agentic AI shifts CPU:GPU ratios and creates a new structural demand wave
CUDA software gains widen NVIDIA's moat, making AMD's ROCm-based AI inference stack increasingly uncompetitive on cost-per-token.
MangoBoost's turnkey DPU racks squeeze more tokens from AMD hardware, making it an acquisition target
Agentic AI shifts compute demand back toward CPUs, triggering a re-rating for AMD as the market rotates from pure GPU focus
Agentic AI shifting compute from GPU-centric to CPU-balanced architectures could re-rate AMD as the market's best GPU-and-CPU play.
Bernstein sees a CPU renaissance across a $223bn TAM; AMD is best-positioned to take share as the leading challenger CPU designer.
Neo-clouds desperate for a non-Nvidia GPU alternative create structural demand tailwind for AMD's AI chip lineup.
Flash-as-DRAM is the next big theme as DRAM price and capacity cap agentic AI workloads
AMD's hardware plus a fast-improving ROCm stack is closing CUDA's moat and winning real AI workloads
OpenAI's CFO confirming AMD chips are live in the pipeline is a concrete demand signal for AMD's data-center accelerators from the single largest AI compute buyer. Among the publicly tradeable names she lists, AMD is the cleanest single long.
AMD is good enough now on rack scale-up; Lisa Su figures out MI450 and may ship a fast SRAM offload inference chip within 12 months
AMD is a strong AI chip play, bought a year ago and just printed fantastic earnings
AMD inference software is closing the gap on NVIDIA, turning Instinct GPUs into a credible second source and lifting datacenter share.
CPUs completely sold out as RL training environments and AI-generated app sprawl drive demand.
CUDA's software moat is cracking as porting tools mature; the named challenger AMD wins share as labs diversify off NVIDIA.
Even free chips lose if they can't match Nvidia's pace, a direct knock on AMD's cheaper alternative
CPU makers benefit as agentic AI drives a CPU and edge inference demand spike, away from GPU-centric
AMD is ahead of everyone at the advanced chip-stacking that lets it pack extra fast memory onto its processors, a manufacturing edge rivals haven't matched.
AMD pulled off a decade-long rebuild and now competes at the top of CPUs, GPUs, and AI systems at once, a structural edge the author frames admiringly.
OpenAI's multi-vendor inference push lands huge deals at AMD as it diversifies off Nvidia
AMD is mid: too many things to do and not enough of a leap to challenge NVIDIA's accelerator dominance.