$Micron Technology(MU)$ The #2 ranked analyst from UBS raised the price target to $1625, and the #13 ranked analyst (out of 12,304 analysts) from TD Cowen raised theirs to $1500. Plus, it's one of the cheapest stocks around. That's good enough for me.

$Roundhill Memory ETF(DRAM)$  Chips keep using more and more HBM with every new release.

Retail investors are piling into SpaceX stock. According to data from Vanda Research, over the last two trading sessions, retail bought nearly as much SpaceX ($SpaceX(SPCX)$ ) as every other US single stock combined during the entire prior week. On Monday alone, retail purchased around $93.8 million worth of $SpaceX(SPCX)$  shares, accounting for roughly 73% of all retail single-stock purchases that day. In the first ten minutes of Monday's session, retail turnover in the stock exceeded $7 million. Individual investors also bought some semiconductor stocks like Marvell ($Marvell Technology(MRVL)$ ), Micron ($Mic

$Intel(INTC)$  Just kicked off production on 18A-P, its most advanced node yet. The turnaround story finally has a manufacturing milestone behind it. The chart agrees, with stacked moving averages and the uptrend still intact. Execution from here is everything.

Adding more $CEVA Inc(CEVA)$  here on weakness. This is one of those names that tends to move with the broader semiconductor sector, especially $Micron Technology(MU)$  and memory-related sentiment. When the DRAM/NAND market stabilizes, high-beta names like CEVA usually re-rate faster than expected. In that context, dips like these aren't breakdowns—they're more like liquidity resets within a larger sector move. From a positioning standpoint, I see this as an opportunity rather than a warning signal. Recent analyst commentary from Needham also reinforces the view that the underlying demand in semis remains intact, not deteriorating. If the semiconductor cycle continues to grind higher, CEVA has ro

These are the four stocks I'm buying on every dip. I'm building my exposure here, not chasing strength. $Micron Technology(MU)$  | The backbone of memory chips powering everything from phones to AI servers. Without it, nothing computes. $Roundhill Memory ETF(DRAM)$  | A pure-play memory basket riding the AI storage and bandwidth supercycle. Demand is structural, not cyclical. $Advanced Micro Devices(AMD)$  | Chips for gaming, data centers, and high-performance compute. Still one of the key AI beneficiaries outside NVDA. $NEBIUS(NBIS)$  | A European AI cloud challenger. Building infrastructure for compani

$Roundhill Memory ETF(DRAM)$ I'm quite skeptical that the Fed will cut rates on Wednesday, hoping this might give us a brief dip to add more. This ETF seems about ready to move.

$Roundhill Memory ETF(DRAM)$  SK Hynix is planning a massive 100 trillion won shareholder return program, which will officially launch after its ADR listing. The program includes share buybacks and dividends.

$Micron Technology(MU)$ $Roundhill Memory ETF(DRAM)$  SK Hynix got a price target update, up 125%. And it will be listed in the US market in July, even bigger than SpaceX.

$Micron Technology(MU)$ $Roundhill Memory ETF(DRAM)$ Korean media reports indicate that SK Hynix's ADR listing process is in its final stages, with only SEC approval remaining. Citing sources familiar with the matter, it's now expected to take place after mid-July, rather than in early August.

$Intel(INTC)$  Global Equity Research just set a near-term price target of $200 for Intel. They expect the long-term market cap to reach $5 trillion. There's a "sweeping thesis" section in the report for those who want to dig deeper. Basically, they anticipate foundry engagements across the board, while CPUs move in as the main driver of spending in data centers. They project earnings per share to hit $5 by 2030.

$Applied Materials(AMAT)$ Applied Materials just launched two new semiconductor manufacturing systems, aiming to advance next-gen 3D scaling for AI-focused chips, targeting higher performance, better energy efficiency, and improved yield. This directly targets the backbone of AI hardware evolution: GAA transistors Advanced DRAM 3D NAND scaling And importantly, this isn't just theoretical – adoption is already flowing through key players in the ecosystem. The supply chains of Micron, TSMC, and Nvidia are all directly exposed to this upgrade cycle. What this really means is simple: the AI trade isn't just about chips anymore; it's about the manufacturing technology enabling the next decade of compute scaling. Applied Materials sits right in the

Everything in $Roundhill Memory ETF(DRAM)$  is the absolute best. Every single aspect. The stock could hit $200, and even $300 is possible.

$Micron Technology(MU)$ $Roundhill Memory ETF(DRAM)$ $SanDisk Corp.(SNDK)$ This shift in the memory sector is getting harder to ignore. Micron has locked in a first-ever 5-year strategic customer agreement, and now Samsung and SK Hynix are also moving toward 3–5 year deals. That's a significant change from how memory used to be traded. Less spot market chaos, more visibility, more structure. If that continues, the "pure cyclical" discount could start to break down. Feels like the re-rating story is still in its early stages.

The AI data center value chain – the real enablers. AI doesn't run without the underlying infrastructure layer: $Advanced Micro Devices(AMD)$  – CPU compute for training and inference workloads $Micron Technology(MU)$  – HBM memory feeding GPUs at extreme speed $SanDisk Corp.(SNDK)$  – NAND storage for datasets and model checkpoints $Arista Networks(ANET)$  – High-speed networking between servers and GPUs $Marvell Technology(MRVL)$  – Data movement silicon inside AI data centers MXL – Connectivity chips for broadband, fiber, and wireless infrastructure

$Roundhill Memory ETF(DRAM)$ The chart is quietly forming a cup pattern near the highs. There's not a lot of noise here, but the structure is getting interesting. The top holdings tell you where the real weight is: SK Hynix Samsung Electronics Micron $Micron Technology(MU)$  Sandisk $SanDisk Corp.(SNDK)$  I'm watching this because, in this AI run, memory isn't just a cyclical trade anymore. It's become one of the strongest structural themes underpinning everything else. You don't usually see this kind of price stability and leadership clustering at highs unless there's real underlying demand supporting it. This feels less like a bounce setup and more

$Micron Technology(MU)$ TD Cowen just made a pretty aggressive move, raising Micron's price target from $660 to $1,500 while maintaining a Buy rating. That's not a small adjustment; it's essentially a different long-term model. What's driving it: increased DRAM content per 1GW even after SOCAMM de-specing, a ~$150 CY27E EPS framework, and CPU demand pushing expectations for stronger, longer pricing power. But the bigger shift is in the underlying assumption, not just the numbers. They're now modeling pricing strength extending into the second half of CY27, instead of a digestion phase in the first half. That's the key change. Because once the market stops thinking "cycle peak → correction" and starts thinking "extended pricing power," everythin

The core names in the data center server chain are the hidden backbone that AI actually depends on to run. $Advanced Micro Devices(AMD)$  CPU layer – still central for both training and inference compute. $Micron Technology(MU)$  HBM memory – feeds GPUs with the bandwidth they actually need. $SanDisk Corp.(SNDK)$  NAND storage – for datasets, checkpoints, and the model persistence layer. $Arista Networks(ANET)$  Network backbone – handles high-speed switching between racks and clusters. $Marvell Technology(MRVL)$  Data movement silicon – optimizes internal

$Everspin Technologies Inc(MRAM)$ Astro Digital relies heavily on the SpaceX Transporter program to deploy its hardware. Multiple Astro Digital spacecraft, built with Everspin's PERSYST MRAM, have successfully reached orbit on missions such as Transporter-10, Transporter-12, Transporter-13, and Transporter-14.

$Intel(INTC)$ Looking good on Hyperdash for Marvell, it's over 134. Broke through a nice resistance level. Marvell looks like it could go even higher, maybe 350+. The stock is apparently 5x what $NVIDIA(NVDA)$  said, not me! A trillion-dollar business.