Damus
Open in Damus

Recent Notes

Pepe Maltese profile picture
AMD: What’s Actually Driving Its AI and Data Center Growth

There’s been a lot of noise around AMD vs. Nvidia, but underneath that is a clear trend: AMD is quietly building a real end-to-end stack for AI infrastructure. The execution is finally syncing across hardware, software, and systems.

Here are the key technical drivers — prioritized by near-term revenue impact:

→ MI350 (CDNA 4) GPUs
Launching mid-2025, 35× performance leap vs. MI300X. Plug-and-play compatible with current systems. Already in deployment w/ Oracle. This is AMD’s most credible shot at real inference market share.

→ EPYC Turin (Zen 5) CPUs
Live and scaling. 150+ server platforms. 30+ new cloud instances from AWS, Google, Oracle. Contributed to 57% YoY data center revenue growth in Q1. This is not a future story — it’s booked compute today.

→ MI300X + MI325X
Deployed and earning. Used in live LLM inference (e.g. LLaMA 405B). MI325X improves memory and smooths path to MI350. Transitional, but real.

→ Ryzen AI Series (Client AI PCs)
+50% QoQ notebook sell-through. +80% YoY commercial designs. Supported by top OEMs. May not be a long-term moat, but drives ASPs now.

→ ROCm Stack
Bi-weekly updates. 2M+ Hugging Face models supported. Day-0 support for LLaMA 4, Gemma, DeepSeek. No longer an adoption blocker.

→ ZT Systems Acquisition
Now AMD can sell rack-level, fully integrated systems (CPUs + GPUs + networking). Competing with Nvidia DGX on infrastructure, not just chips.

Why It Matters
AMD isn’t trying to be Nvidia — it’s building a full-stack alternative for a world that wants optionality.

They still have to execute cleanly — MI350 rollout is critical. But this isn’t a “wait and hope” story anymore. The pieces are live.
Pepe Maltese · 41w
9/ Still early. Nvidia’s CUDA + Blackwell + market capture is real. But AMD isn’t just playing catch-up. They’re building something that fits the post-monopoly stack.
Pepe Maltese · 41w
8/ What matters: → Turin is already delivering → MI325X is ramping → MI350 is sampling → ROCm is working → ZT is the enabler This is no longer “wait for next gen.” They’re in market. And in motion.
Pepe Maltese · 41w
7/ Acquisition of ZT Systems plugs the final gap: Full rack-scale systems — CPUs + GPUs + networking. Think Nvidia DGX competitor — but modular. ZT gives AMD the vertical muscle to sell full AI infrastructure to hyperscalers.
Pepe Maltese · 41w
6/ The missing piece? Software. ROCm isn’t CUDA. But it’s catching up. Fast. Bi-weekly updates. 2M+ Hugging Face models supported. LLaMA 4, Gemma, DeepSeek — all Day-0. This used to be AMD’s moat problem. Less so now.
Pepe Maltese · 41w
5/ AI PCs are noise in most headlines, but Ryzen AI is actually selling. +50% QoQ notebook sell-through +80% YoY commercial design wins HP, Lenovo, Asus — all in. Does it matter long-term? TBD. But it’s margin-positive now.
Pepe Maltese · 41w
4/ MI300X and MI325X are shipping now. They’re handling real-world inference, including Meta’s LLaMA 405B. MI325X builds on MI300X — better memory, better economics. Sets the stage for MI350 without disruption.
Pepe Maltese · 41w
1/ AMD isn’t Nvidia. It doesn’t need to be. What matters: they’re finally building a full-stack AI/data center offering that’s real — not theoretical. Execution is lining up. Revenue is foll...
 Pepe Maltese profile picture
Pepe Maltese @pepemaltese
2/
The upcoming MI350 (CDNA 4) GPU series is the big unlock.
Mid-2025 launch. 35× perf over MI300X.
No need to redesign racks — it drops into current MI300 systems.
Major deployment w/ Oracle already underway.
That’s not roadmap noise. It’s movement.
1
Pepe Maltese · 41w
3/ But the present is moving too. EPYC Turin (Zen 5) is in-market, shipping, and scaling. Used across 150+ server platforms. 30+ new cloud instances live from AWS, GCP, Oracle. This isn’t theoretical traction — it’s booked compute.
Pepe Maltese profile picture
1/
AMD isn’t Nvidia. It doesn’t need to be.
What matters: they’re finally building a full-stack AI/data center offering that’s real — not theoretical.
Execution is lining up. Revenue is following.

A few key reasons why:
8
Pepe Maltese · 41w
2/ The upcoming MI350 (CDNA 4) GPU series is the big unlock. Mid-2025 launch. 35× perf over MI300X. No need to redesign racks — it drops into current MI300 systems. Major deployment w/ Oracle already underway. That’s not roadmap noise. It’s movement.
note1h046v...
Pepe Maltese profile picture
Pepe Maltese @pepemaltese
Bear case watch:

ARPU growth slowing (6–7%)

FCF down YoY

Monetization still lags engagement
If Roku can’t close that TRC-to-revenue gap, the multiple stays stuck.
But if it does — re-rating potential is huge.
#Investing #Compounders #TechStocks