
AMD
CPUs, GPUs and AI accelerators for data centres, PCs and gaming consoles.
Public · NASDAQ:AMD- Sector
- Semiconductors
- Based in
- Santa Clara, United States
- Founded
- 1969
- Invested
- Apr 2021
Advanced Micro Devices designs processors for servers (EPYC), PCs (Ryzen) and graphics (Radeon), plus the Instinct accelerators it sells for AI in data centres. Its custom chips power the PlayStation and Xbox consoles, and the adaptive chips from its acquisition of Xilinx are used in networking, cars and industrial equipment.
AMD is "fabless": it designs chips and has them manufactured by partners, mainly TSMC. It spun off its own factories as GlobalFoundries in 2009.
AMD was founded in 1969. Under Lisa Su, CEO since 2014, it went from near-bankruptcy to a leading supplier of data centre CPUs and the main challenger to Nvidia in AI accelerators.
As of 29 Sep 2026
Why I own it
AMD is the credible second source for data centre compute. Its EPYC chips took share from Intel in servers, and its Instinct accelerators give AI buyers the alternative to Nvidia they badly want.
As compute demand grows by orders of magnitude, no customer wants a single supplier. Being a strong number two in a market that big is a great place to be.
Then · Apr 2021
Since I invested
- Held for
- 5 years, 6 months
- Status
- Active
- Share price, past year
- +167%
- In the notebook
- 2 notes
Thesis check since 29 Sep 2026
7 developments support it, 0 challenge it
- SupportsBuilt for the AI era.
- SupportsOpen models need an open stack.
- SupportsTo meet the aggressive performance goals of AI, leaders need reliable solutions.
Checked against the thesis by Jev, an AI model. Can be wrong.
From the notebook
All notesComing up
TSMC's Q3 result lands on 15 October
TSMC reports third-quarter earnings on 15 October, Taiwan time, having guided revenue to $44.6bn-$45.8bn and flagged pressure on margins. That guidance reflects temporary pressure from 2-nanometer technology ramp-up costs, even as AI-driven orders keep volumes climbing. The call is the first real test of whether advanced-node demand from Nvidia, AMD and Apple is still outrunning TSMC's capacity to build it.
Since this note:TSMC −0.3%Nvidia −0.2%AMD +0.8%(S&P 500 +1.3%)
Connections
Xanadu and AMD's Backline wires quantum chips into EPYC servers
Xanadu and AMD this month released Backline, an open platform co-designed by the two companies for rapid communication between quantum and classical computing hardware. It targets end-to-end loop times under 3 microseconds between Xanadu's photonic processors and AMD's EPYC and Threadripper CPUs, Instinct GPUs and Versal FPGAs. For AMD it's a low-cost way into quantum-adjacent infrastructure; for Xanadu it ties its PennyLane software to chips already sitting in data centres.
Since this note:Xanadu −17.3%AMD −1.6%(S&P 500 +0.3%)
Written by AI from share prices, the companies' posts and the news. Can be wrong; not financial advice.
Latest on X
@AMD
AMD@AMD · 8 OctAgentic AI isn't just a GPU workload. CPUs play a big role.
Recent @mimiktech tests of agentic workflows on AMD Ryzen AI Embedded X100 processors found about 80% of operations were CPU-bound, supporting coordination, orchestration, scheduling and reporting.
Watch @Fayarjomandi, founder and CEO of mimik, as she explains why heterogeneous compute matters for deploying agentic AI at the edge and describes her vision for open, collaborative and scalable systems.
🎥 Full interview: https://youtu.be/8CFnJ5qwCrA
9990
AMD@AMD · 7 OctBuilt for the AI era.
Proud to power the next generation of @HPE ProLiant servers with 6th Gen AMD EPYC processors.28149
AMD@AMD · 7 OctOpen models need an open stack.
@ZyphraAI VP of AI Engineering Quentin Anthony shares how access to open software libraries and direct collaboration with AMD are helping the team train bigger models, keep growing compute resources working efficiently and pave new ground in AI.
Learn how Zyphra trained its ZAYA1-8B reasoning model from scratch on a full-stack AMD platform: https://bit.ly/4ixfdaB
1474
AMD@AMD · 7 OctTo meet the aggressive performance goals of AI, leaders need reliable solutions.
@OpenAI shared why their team selected AMD EPYC “Turin” CPUs to host its Jalapeño AI ASIC deployment, citing platform strength, partner experience and the importance of reducing unnecessary risk.
More from @TomsHardware: https://www.tomshardware.com/pc-components/cpus/openais-jalapeno-asics-are-deployed-alongside-amd-epyc-turin-cpus-as-hosts-hardware-vp-says-nvidias-vera-standalone-is-a-little-bit-behind-on-that-maturity-level1536424
AMD@AMD · 3 OctWe're enjoying our Saturday morning ☕, a bit of 📰 and seeing our AMD vLLM commits.
Open source FTW.1124260

