$Lumentum(LITE)$ Coherent is the picks-and-shovels play for data center optics, lasers, and transceivers moving data in and between DCs. It extended higher into the after-close print on Tuesday, riding a strong week driven by headlines around China transceiver restrictions. The structure should hold if it stays a leader here. I'm not touching it before earnings — but I'm expecting clean guidance. A weekly close above 900 opens the door to 1200+.

I added to my $Roundhill Memory ETF(DRAM)$  position recently. Still comfortable with the AI memory thesis and the demand outlook. It could see a shakeout and sweep lower, but that doesn't really bother me. I'd just treat any weakness as a chance to add more.

$Roundhill Memory ETF(DRAM)$ KOSPI, I know the pattern all too well. You'll probably let us down again, but honestly, we've been through enough pain already. This feels like the time to finally buy that massive dip.

$Roundhill Memory ETF(DRAM)$  If we aren't really going below 49 or 50, then where are we headed from here? Chip sector stocks don't move sideways for long. They tend to swing pretty wildly in both directions.

$Roundhill Memory ETF(DRAM)$  Options IV is collapsing, which suggests investors are positioning for upside right now. Looks like we could touch ATH again in the coming months.

$Lumentum(LITE)$ US-made products that are in strong demand set the stage for this earnings call to be a solid one. I fully expect them to meet or beat the projections from the last call, just like they have over the past year. Personally, selling isn't really an option for me — I've been in this for 20 years and have built up significant gains, so the tax hit would be rough. That keeps me on the long side for a while. I've added a few shares on the dips that pop up from time to time. My prediction hasn't changed: all-time highs around the call, then moving higher from there with the occasional pullback.

$Roundhill Memory ETF(DRAM)$  The amount of bullish memory-related news over the past day has been pretty intense, and yet some folks still seem determined to stay bearish.

$Lumentum(LITE)$  If any kind of US ban actually moves forward, I'd expect the hyperscalers to push back hard. It would hit their rapid data center buildouts directly, and they have heavy lobbying power to squash something like this. US companies simply can't match the demand right now, not in quantity and not in price.

AI infrastructure supply chains could be heading for a major shift. If the US moves to restrict Chinese-made components used in AI data centers, cloud giants might have to lean more on domestic and allied suppliers. Three names I'm keeping an eye on: $Coherent(COHR)$  is already getting attention as investors price in potential demand from hyperscalers like Amazon, Google, Microsoft and Meta. $Lumentum(LITE)$  is an established supplier to major cloud players — any supply shift away from China could open up a meaningful opportunity. $Applied Optoelectronics(AAOI)$  is building US manufacturing capacity in Texas and looks positioned as a direct AI d

$Roundhill Memory ETF(DRAM)$ TSLA mentioned their humanoid robots need twice the DRAM compared to AI chatbots. That basically locks in demand for the memory sector for the next decade.

$Lumentum(LITE)$ A potential U.S. restriction on new Chinese-made optical transceivers for AI data centers could reshape the competitive landscape and shift more attention toward domestic suppliers. I'm watching AAOI, LITE, and COHR closely. AAOI is a leading U.S. optical transceiver manufacturer with strong exposure to AI networking demand. LITE is a key supplier of high-speed optical components powering next-generation AI infrastructure and hyperscale data centers. COHR is a major player in photonics and optical communications, benefiting from expanding AI and data center investment. AI networking is becoming one of the hottest battlegrounds in tech. These optical leaders are worth keeping on the radar as the sector continues to evolve.

$Roundhill Memory ETF(DRAM)$ Memory stocks Micron and Sandisk jumped after Samsung warned of a memory crunch into 2028. Chip stocks have been a major beneficiary of the AI data center and infrastructure buildout.

$Roundhill Memory ETF(DRAM)$  CXMT as competition to Micron is honestly a fraud. They are stuck at 1x/1y/early‑1z DRAM nodes, and their most advanced DRAM is still built on legacy multi‑patterned DUV, which limits them to roughly 18–20 nm class for 1x, around 16–17 nm class for 1y, and about 15–16 nm class for early 1z. Those nodes are 5–7 years behind Micron's leading nodes. CXMT cannot shrink DRAM cells further because DRAM capacitor and bitline geometry hits a hard wall without EUV. They cannot produce 1α DRAM, 1β DRAM, 1γ DRAM, or HBM2E / HBM3 / HBM3E. They are permanently stuck in DDR4‑era physics with cheap inferior chips. Their DDR5 is fake — it is DDR4‑class DRAM with DDR5 branding, missing real DDR5 spe

$Roundhill Memory ETF(DRAM)$  ETFs like this really hurt MU by pulling in that Korean stuff. Otherwise, we might have been sailing toward $1500 by now.

$Roundhill Memory ETF(DRAM)$  I plan to open a Roth IRA for my kids and set up monthly buys of this ETF. By the time they graduate high school, they'll probably be in a pretty good position for the long run.

Bears might get some short-term relief from the Burry narrative, cyclical arguments, or any short-side play, but that's all temporary. The fundamentals and the underlying truth don't just disappear. $Roundhill Memory ETF(DRAM)$  $SanDisk Corp.(SNDK)$  $Western Digital(WDC)$  $Roundhill Memory ETF(DRAM)$  The truth will come through eventually, even if it takes a while.

$Micron Technology(MU)$ $NVIDIA(NVDA)$ $Roundhill Memory ETF(DRAM)$  $SanDisk Corp.(SNDK)$ It feels like there's only so long they can keep these down, with all the fundamentals supporting another 50% up move on stock prices. All the noise looks like nothing but FUD while they accumulate.

$Micron Technology(MU)$ $Roundhill Memory ETF(DRAM)$  I bought low on Micron, and now it's paying off as big tech ramps up AI spending. The confirmation is there. I've got 50% of my portfolio in it, and I'm looking toward 1200.

$Roundhill Memory ETF(DRAM)$ Saw a slight pullback earlier. I'm planning to reload and add some more heading into the weekend. For now I'm just taking the opportunity and accepting the volatility. I think higher highs are ahead in due time.

$Roundhill Memory ETF(DRAM)$ It's become pretty clear to me that South Korea as a whole hasn't really learned anything from the market crash. I'm still adding more DRAM but definitely selling calls and buying puts this time around. DRAM hitting 100+ one day feels inevitable. It is going to be a bumpy ride though.