Mag 7 Is Dead? Meet DIAMANS! Did You Get a Seat on New AI Hardware Basket?
Recent spotlights shine on these stocks: $SanDisk Corp.(SNDK)$ +16.6%, $Micron Technology(MU)$ +15.5%, $Intel(INTC)$ +14.0%, $Dell Technologies Inc.(DELL)$ +13.1%, $Advanced Micro Devices(AMD)$ +11.4%, $NVIDIA(NVDA)$ +1.7%, $Apple(AAPL)$ +2.0% — Wall Street has a new concept: the Magnificent 7 era is over. Welcome to DIAMANS, the AI hardware chain that just went vertical today.
What Is DIAMANS?
Dell + Intel + AMD + Micron + Apple + NVIDIA + SanDisk.
Not a random name mashup — a complete AI infrastructure chain:
-
$Dell Technologies Inc.(DELL)$ : Servers — the physical delivery layer for all compute
-
$Intel(INTC)$ : CPU + manufacturing
-
$Advanced Micro Devices(AMD)$ : Alternative compute, inference + enterprise data center
-
$Micron Technology(MU)$ : Memory — the fuel for every AI inference call
-
$Apple(AAPL)$ : Edge device — AI moving from cloud to every pocket
-
$NVIDIA(NVDA)$ : GPU — training and large-scale inference
-
$SanDisk Corp.(SNDK)$ : NAND + SSD — storing the world's unstructured data
Mag 7 was about platform monopoly and traffic dominance. DIAMANS is about supply bottlenecks and delivery capacity.
Why hardware has more certainty than software right now?
The core tension in AI investing: consumer AI monetization is the opposite of the internet flywheel.
Internet era: more users → marginal cost → zero → higher margins. That's how Google and Meta were built.
LLM era: more users → every query burns tokens → bigger GPU bills. More traffic can mean more losses.
That's why "who makes money in consumer AI" is still an open question. But hardware cash flows have already landed:
Hyperscalers building data centers; Servers shipping; HBM + NAND undersupply; AI PC rolloutIn the AI capex cycle, hardware is where capex converts to cash flow first.
Does that mean software is uninvestable?
$iShares Expanded Tech-Software Sector ETF(IGV)$ is up ~5% this week. $$ (semiconductors) is up ~11%. Software is recovering — it's just not leading.
But this is exactly where FOMO destroys returns: your software position is up 5%, the DIAMANS names are making new highs every day, and the pressure to chase and rotate builds. The moment you can't take the underperformance is usually the top. Don't use someone else's returns as your trade signal.
Chase or wait for the rotation?
Even within DIAMANS today, divergence is appearing: $SNDK$ +16.6%, $MU$ +15.5%, $INTC$ +14% — while $NVDA$ lagged at just +1.7%. Funds are clearly rotating toward storage and CPU names that had fallen behind. $NVDA$ is the temporary underperformer inside its own basket.
If AI capex doesn't roll over, the hardest supply constraints to fix — NAND, DRAM, CPU production — are where premium stays longest.
🎯 What's Your Take on DIAMANS?
Within the seven, which do you think has the most room from here?
Is the rotation a short-term blip, or is the market repricing which part of the AI stack is the real bottleneck asset?
Are you positioned hardware-heavy, software-heavy, or balanced?
Leave comments to win tiger coins~
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.
While NVIDIA is the king, I believe that $Intel(INTC)$ has the most upside potential. If Intel successfully executes its Foundry strategy & becomes a possible competitor to TSMC, the upside is generational. Intel can possibly be the ultimate patriotic trade backed by billions in US government subsidies.
Intel is no longer a turnaround story. It is morphing into a dual engine powerhouse that combines a legacy CPU recovery with a futuristic power play.
While NVIDIA and AMD battle for AI design supremacy, Intel is positioning itself as the only US company that can actually build the future of silicon on a massive scale.
The biggest catalyst for Intel now is the successful high volume manufacturing of its 18A process node.
Exciting times are ahead for Intel.
@Tiger_comments @Tiger_SG @TigerStars
Among the seven, I still think Micron Technology has one of the strongest setups because every AI workload needs more HBM and DRAM. I also don’t see $NVIDIA(NVDA)$ smaller move today as bearish — it feels more like funds rotating into lagging AI hardware names instead of leaving the AI theme.
I don’t think software is dead either, but chasing every hot rotation usually destroys returns. For me, staying balanced makes more sense: hardware can continue leading if AI capex stays strong, while software could catch up later once monetization improves.
@Tiger_comments @TigerStars @TigerClub
This rotation is just the market pricing in the bottleneck which is the supply chain of the hardware. Hardware always take time to catch up as building it requires both expertise, facilities and raw materials. All these cannot be quickly set up or acquired. As long as demand outstrip supply, the companies have pricing power and the buyers have no choice but to pay for them. However, this is not sustainable. Supply and demand will eventually even out in all open markets.
Software depends on hardware to be able to drive further capabilities. Yet, hardware is useless without the software to perform the intended task. So, I prefer to be balanced for the longer term investment horizon.
A few sessions ago the talk was that names like MU and SNDK had already run too hard.
Now one supply shift later and people are back to asking whether the real move is only just starting.
That’s why this part of the market is so hard to trade.
When supply tightens, pricing power can change fast, and suddenly what looked “too expensive” starts getting re-rated again.
I still think memory is one of the most important pieces of the AI hardware story.
The real question now is whether this is the start of a bigger leg higher, or just another sentiment spike that gets sold.
Are you treating this as a real breakout in the memory theme, or just hype getting overheated again?
Most upside from here may still be in Advanced Micro Devices and Micron Technology because the market is repricing the actual bottlenecks of AI scaling:
HBM memory,
advanced packaging,
networking,
storage throughput, not just GPUs alone.
NVIDIA remains dominant, but expectations are already enormous. Meanwhile, SanDisk could still have strong upside if AI storage demand becomes structurally persistent rather than cyclical.
I do not think this rotation is just a short-term blip. Markets are shifting from “who has AI exposure?” to “who controls constrained infrastructure capacity?”
That said, after such violent rallies, risk management matters more:
trim parabolic moves,
keep core winners,
avoid low-quality AI hype names.
Hardware still looks like the clearer monetisation layer today, while much of software AI monetisation remains crowded and harder to differentiate.
The Strategy is shifting from "Hardware Heavy" to "Balanced," favoring companies with proven software monetization channels.
2. Is the rotation a short term blip or the market repricing which part of the ai stack is the real bottleneck. The bottle net for ai for $Microsoft(MSFT)$ is their ability to invest in computer hardware for their azure data centres
2. Positioning hardware software of balanced. $Microsoft(MSFT)$ has both a hardware division with surface and a software division with windows 11
Internet era: more users → marginal cost → zero → higher margins. That's how Google and Meta were built.
LLM era: more users → every query burns tokens → bigger GPU bills. More traffic can mean more losses.
That's why "who makes money in consumer AI" is still an open question. But hardware cash flows have already landed:
Hyperscalers building data centers; Servers shipping; HBM + NAND undersupply; AI PC rolloutIn the AI capex cycle, hardware is where capex converts to cash flow first.