To The Moon

$AXT Inc(AXTI)$ $AXT Inc(AXTI)$ All three, $Lumentum(LITE)$ , $COHERENT(COHR)$ , and $Applied Optoelectronics(AAOI)$ , own their own foundries, achieving vertical integration of discrete InP electro-absorption modulated lasers or continuous-wave InP laser arrays to feed a silicon photonics engine. To fulfill this process, they need to purchase blank InP substrate wafers from AXTI. And the physical number of laser chips or array sizes from $SIVERS SEMICONDUCTORS AB(SIVEF)$  must scale up to deliver the nece

$Direxion Daily Semiconductors Bear 3x Shares(SOXS)$ First time in three months seeing some strength without a sell-off. Might be a decent play for the next few days.

$Direxion Daily Semiconductors Bear 3x Shares(SOXS)$ Pushing for a 40 percent gain today, I'm watching.

$SpaceX(SPCX)$ Just signed a $6.3B compute deal with Reflection AI, and the structure here is what really stands out. $150M/month starting July 1 through 2029 (CNBC). Access to NVIDIA GB300s for frontier model training. A 90-day exit clause after the first 3 months. What's interesting to me is how quickly these deals are stacking up: Anthropic compute deal Google compute deal Cursor compute + acquisition angle Reflection AI latest $6.3B deal It feels less like one-off headlines now and more like a steady pipeline of long-duration AI compute demand being booked in real time. From a trading perspective, this is where it shifts from "news flow" to "order flow visibility."

$SpaceX(SPCX)$ SPCX will be added to the NDX on July 6th. It should revisit $225. Shorting it before that doesn't seem smart.

$Applied Optoelectronics(AAOI)$  $Lumentum(LITE)$ $Cohen & Company Inc.(COHN)$ A 6.4T optical engine uses more Indium Phosphide (InP) material than a 3.2T optical engine. The main reason why 6.4T consumes more InP materials is the increase in total lane count and light source volume required to double the bandwidth from 3.2 Terabits per second to 6.4 Tbps. First, lane count multiplication is needed. To scale up multi-terabit architectures, optical engines rely on multiplying physical lanes or laser wavelengths. A 6.4T engine requires twice the data-carrying capacity of a 3.2T engine. Even when moving to next-gener

For years, AI infrastructure has been the leading theme, and there's no reason to think that has changed. It has only accelerated. Photonics-related names include: $Applied Optoelectronics(AAOI)$  $Silver Verde May Mining Co., Inc.(SIVE)$  $AXT Inc(AXTI)$  $Lumentum(LITE)$  $COHERENT(COHR)$  GLW. The AI infrastructure story is far from over, and photonics is becoming an increasingly important piece of the puzzle. As AI training and inference workloads continue to grow, data movement—not compute itself—is emerging as the next major bottleneck. AAOI, LITE,

$Direxion Daily Semiconductors Bear 3x Shares(SOXS)$ I think if I can get my average cost down below $6 by the end of July, there's a decent payout waiting.

The biggest returns often come from buying quality when others are looking away. My high-conviction buy zones are: - $ServiceNow(NOW)$  below $100 - $Ondas Holdings Inc.(ONDS)$  below $10 - $Applied Optoelectronics(AAOI)$  below $165 - $Hims & Hers Health Inc.(HIMS)$  below $30 - ZETA below $20 - OSCR below $20 - LMND below $50 - IREN below $50 These aren't predictions, just price levels where I think the risk/reward starts to look attractive. Which ones are already on your watchlist?

$ProShares UltraPro QQQ(TQQQ)$ It's a normal dip at the start of Fed meetings. Happens every time. Wish I had some cash to grab it.