𝗠𝗲𝗱𝗶𝗮𝗧𝗲𝗸 𝗤𝟯’𝟮𝟲: 𝗥𝗲𝘃𝗲𝗻𝘂𝗲 𝗚𝗿𝗲𝘄 𝟭𝟭% 𝗬𝗼𝗬, 𝗘𝘅𝗰𝗲𝗲𝗱𝗲𝗱 𝗛𝗶𝗴𝗵 𝗘𝗻𝗱 𝗼𝗳 𝗚𝘂𝗶𝗱𝗮𝗻𝗰𝗲; 𝗦𝗺𝗮𝗿𝘁 𝗘𝗱𝗴𝗲 𝗢𝗳𝗳𝘀𝗲𝘁 𝗠𝗼𝗯𝗶𝗹𝗲 𝗣𝗵𝗼𝗻𝗲 𝗗𝗲𝗰𝗹𝗶𝗻𝗲 @MediaTek'𝗌 𝖰𝟥 𝟤𝟢𝟤𝟨 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝖾𝗑𝖼𝖾𝖾𝖽𝖾𝖽 𝗍𝗁𝖾 𝗁𝗂𝗀𝗁 𝖾𝗇𝖽 𝗈𝖿 𝗂𝗍𝗌 𝗀𝗎𝗂𝖽𝖺𝗇𝖼𝖾, 𝗋𝖾𝖺𝖼𝗁𝗂𝗇𝗀 𝖴𝖲$𝟧.𝟥 𝖻𝗂𝗅𝗅𝗂𝗈𝗇, 𝗎𝗉 𝟣𝟣% 𝖸𝗈𝖸. 𝖬𝗈𝖻𝗂𝗅𝖾 𝖯𝗁𝗈𝗇𝖾 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝖽𝖾𝖼𝗅𝗂𝗇𝖾𝖽 𝟤𝟥% 𝖸𝗈𝖸, 𝖺𝗅𝗍𝗁𝗈𝗎𝗀𝗁 𝗍𝗁𝗂𝗌 𝗐𝖺𝗌 𝗉𝖺𝗋𝗍𝗅𝗒 𝗈𝖿𝖿𝗌𝖾𝗍 𝖻𝗒 𝗀𝗋𝗈𝗐𝗍𝗁 𝗂𝗇 𝗍𝗁𝖾 𝖲𝗆𝖺𝗋𝗍 𝖤𝖽𝗀𝖾 𝗌𝖾𝗀𝗆𝖾𝗇𝗍. 𝗞𝗲𝘆 𝗧𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀 • 𝖪𝖾𝗒 𝗀𝗋𝗈𝗐𝗍𝗁 𝖽𝗋𝗂𝗏𝖾𝗋𝗌 𝖽𝗎𝗋𝗂𝗇𝗀 𝗍𝗁𝖾 𝗊𝗎𝖺𝗋𝗍𝖾𝗋 𝗂𝗇𝖼𝗅𝗎𝖽𝖾𝖽 𝖼𝗈𝗇𝗍𝗂𝗇𝗎𝖾𝖽 𝗌𝗍𝗋𝖾𝗇𝗀𝗍𝗁 𝗂𝗇 𝖲𝗆𝖺𝗋𝗍 𝖤𝖽𝗀𝖾 𝗉𝗅𝖺𝗍𝖿𝗈𝗋𝗆𝗌 𝖺𝗇𝖽 𝗍𝗁𝖾 𝗋𝖺𝗆𝗉-𝗎𝗉 𝗈𝖿 𝖠𝖲𝖨𝖢 𝗉𝗋𝗈𝖽𝗎𝖼𝗍𝗂𝗈𝗇, 𝗐𝗁𝗂𝖼𝗁 𝗉𝖺𝗋𝗍𝗅𝗒 𝗈𝖿𝖿𝗌𝖾𝗍 𝗐𝖾𝖺𝗄𝗇𝖾𝗌𝗌 𝗂𝗇 𝖬𝗈𝖻𝗂𝗅𝖾 𝖯𝗁𝗈𝗇𝖾 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝖺𝗆𝗂𝖽 𝗋𝗂𝗌𝗂𝗇𝗀 𝗆𝖾𝗆𝗈𝗋𝗒 𝗉𝗋𝗂𝖼𝖾𝗌 𝖺𝗇𝖽 𝗌𝗈𝖿𝗍𝖾𝗋 𝖽𝖾𝗆𝖺𝗇𝖽. • 𝖲𝗆𝖺𝗋𝗍 𝖤𝖽𝗀𝖾 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝗀𝗋𝖾𝗐 𝟧𝟥% 𝖸𝗈𝖸, 𝖽𝗋𝗂𝗏𝖾𝗇 𝖻𝗒 𝗏𝗈𝗅𝗎𝗆𝖾 𝗉𝗋𝗈𝖽𝗎𝖼𝗍𝗂𝗈𝗇 𝖺𝖼𝗋𝗈𝗌𝗌 𝟧𝖦 𝗆𝗈𝖽𝖾𝗆𝗌, 𝖠𝖨 𝖯𝖢𝗌, 𝗍𝖺𝖻𝗅𝖾𝗍𝗌 𝖺𝗇𝖽 𝖺𝗎𝗍𝗈𝗆𝗈𝗍𝗂𝗏𝖾, 𝖺𝗅𝗈𝗇𝗀𝗌𝗂𝖽𝖾 𝗁𝗂𝗀𝗁𝖾𝗋 𝖳𝖵 𝖲𝗈𝖢 𝗋𝖾𝗏𝖾𝗇𝗎𝖾. • 𝖬𝗈𝖻𝗂𝗅𝖾 𝖯𝗁𝗈𝗇𝖾 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝖽𝖾𝖼𝗅𝗂𝗇𝖾𝖽 𝟤𝟥% 𝖸𝗈𝖸 𝗂𝗇 𝖰𝟥, 𝗂𝗆𝗉𝖺𝖼𝗍𝖾𝖽 𝖻𝗒 𝗆𝖾𝗆𝗈𝗋𝗒 𝗌𝗎𝗉𝗉𝗅𝗒 𝖼𝗈𝗇𝗌𝗍𝗋𝖺𝗂𝗇𝗍𝗌, 𝗁𝗂𝗀𝗁𝖾𝗋 𝖽𝖾𝗏𝗂𝖼𝖾 𝖼𝗈𝗌𝗍𝗌 𝖺𝗇𝖽 𝗌𝗈𝖿𝗍𝖾𝗋 𝖾𝗇𝖽-𝗆𝖺𝗋𝗄𝖾𝗍 𝖽𝖾𝗆𝖺𝗇𝖽, 𝗉𝖺𝗋𝗍𝗂𝖼𝗎𝗅𝖺𝗋𝗅𝗒 𝗂𝗇 𝗍𝗁𝖾 𝗆𝗂𝖽- 𝖺𝗇𝖽 𝗅𝗈𝗐-𝖾𝗇𝖽 𝗌𝖾𝗀𝗆𝖾𝗇𝗍𝗌. • 𝖯𝗈𝗐𝖾𝗋 𝖨𝖢 𝖼𝗈𝗇𝗍𝗋𝗂𝖻𝗎𝗍𝖾𝖽 𝟧% 𝗈𝖿 𝗍𝗈𝗍𝖺𝗅 𝗋𝖾𝗏𝖾𝗇𝗎𝖾, 𝗀𝗋𝗈𝗐𝗂𝗇𝗀 𝟣𝟪% 𝖸𝗈𝖸, 𝖽𝗋𝗂𝗏𝖾𝗇 𝖻𝗒 𝗆𝖺𝗋𝗄𝖾𝗍-𝗌𝗁𝖺𝗋𝖾 𝗀𝖺𝗂𝗇𝗌 𝗂𝗇 𝖽𝖺𝗍𝖺 𝖼𝖾𝗇𝗍𝗋𝖾-𝗋𝖾𝗅𝖺𝗍𝖾𝖽 𝖻𝗎𝗌𝗂𝗇𝖾𝗌𝗌. • 𝖠𝖨 𝖠𝖲𝖨𝖢 𝗌𝗁𝗂𝗉𝗆𝖾𝗇𝗍𝗌 𝗌𝗍𝖺𝗋𝗍𝖾𝖽 𝗂𝗇 𝖲𝖾𝗉𝗍𝖾𝗆𝖻𝖾𝗋, 𝗐𝗂𝗍𝗁 𝗌𝗍𝗋𝗈𝗇𝗀 𝖼𝗎𝗌𝗍𝗈𝗆𝖾𝗋 𝖽𝖾𝗆𝖺𝗇𝖽 𝗌𝗎𝗉𝗉𝗈𝗋𝗍𝗂𝗇𝗀 𝖾𝗑𝗉𝖾𝖼𝗍𝖾𝖽 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝗈𝖿 𝖴𝖲$𝟤.𝟧 𝖻𝗂𝗅𝗅𝗂𝗈𝗇 𝗂𝗇 𝖰𝟦 𝟤𝟢𝟤𝟨.   • 𝖰𝟦 𝟤𝟢𝟤𝟨 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝗂𝗌 𝖾𝗑𝗉𝖾𝖼𝗍𝖾𝖽 𝗍𝗈 𝖻𝖾 𝗂𝗇 𝗍𝗁𝖾 𝗋𝖺𝗇𝗀𝖾 𝗈𝖿 𝖴𝖲$𝟨.𝟨–𝟨.𝟫 𝖻𝗂𝗅𝗅𝗂𝗈𝗇, 𝗋𝖾𝗉𝗋𝖾𝗌𝖾𝗇𝗍𝗂𝗇𝗀 𝟥𝟩–𝟦𝟥% 𝖸𝗈𝖸 𝗀𝗋𝗈𝗐𝗍𝗁, 𝗉𝗋𝗂𝗆𝖺𝗋𝗂𝗅𝗒 𝖽𝗋𝗂𝗏𝖾𝗇 𝖻𝗒 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝗋𝖾𝖼𝗈𝗀𝗇𝗂𝗍𝗂𝗈𝗇 𝖿𝗋𝗈𝗆 𝗍𝗁𝖾 𝖠𝖲𝖨𝖢 𝖻𝗎𝗌𝗂𝗇𝖾𝗌𝗌 𝖺𝗇𝖽 𝖼𝗈𝗇𝗍𝗂𝗇𝗎𝖾𝖽 𝗀𝗋𝗈𝗐𝗍𝗁 𝗆𝗈𝗆𝖾𝗇𝗍𝗎𝗆 𝗂𝗇 𝗍𝗁𝖾 𝖲𝗆𝖺𝗋𝗍 𝖤𝖽𝗀𝖾 𝗌𝖾𝗀𝗆𝖾𝗇𝗍. 𝗥𝗲𝗹𝗮𝘁𝗲𝗱 𝗜𝗻𝘀𝗶𝗴𝗵𝘁𝘀 𝗠𝗲𝗱𝗶𝗮𝗧𝗲𝗸 𝗗𝗶𝗺𝗲𝗻𝘀𝗶𝘁𝘆 𝟵𝟲𝟬𝟬 𝗣𝗿𝗼: 𝗠𝗼𝘃𝗶𝗻𝗴 𝘁𝗼 𝟮𝗻𝗺 𝘄𝗶𝘁𝗵 𝗮 𝗙𝗼𝗰𝘂𝘀 𝗼𝗻 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗮𝗻𝗱 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆: 𝗠𝗲𝗱𝗶𝗮𝗧𝗲𝗸 𝘁𝗼 𝗖𝗼𝗻𝘁𝗿𝗶𝗯𝘂𝘁𝗲 𝟭 𝗶𝗻 𝟰 𝗔𝗜 𝗔𝗦𝗜𝗖 𝗦𝗲𝗿𝘃𝗲𝗿 𝗖𝗼𝗺𝗽𝘂𝘁𝗲 𝗦𝗵𝗶𝗽𝗺𝗲𝗻𝘁𝘀 𝗶𝗻 𝟮𝟬𝟮𝟴: 𝗗𝗮𝘁𝗮 𝗖𝗲𝗻𝘁𝗲𝗿 𝗢𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝗶𝗲𝘀 𝗔𝗰𝗰𝗲𝗹𝗲𝗿𝗮𝘁𝗲 𝗠𝗲𝗱𝗶𝗮𝗧𝗲𝗸’𝘀 𝗗𝗶𝘃𝗲𝗿𝘀𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆 #MediaTek# #Earnings# #Semiconductors# #SmartEdge# #AIASIC# #ASIC# #EdgeAI# #MobileSoC#

👀 #MediaTek# confirms its AI ASIC collaboration with #Intel’s# EMIB-T advanced packaging. The chipmaker is now pushing 400G SerDes IP to target next-gen ASIC opportunities, including reportedly #Google’s# TPU v10 and Meta’s custom AI chips.💡More: 🔗

$AVGO already controls ~75% of AI ASIC market with six core customers including $GOOGL, OpenAI and Anthropic which Hock Tan says becomes its largest custom chip client in FY27 as agentic inference drives more workloads toward custom silicon. The part I like most is shipments are expected to double to ~10M in 2027 and reach 24M by 2030 while ASPs rise from ~$10K to ~$18K as chip complexity increases. So if AI revenue really moves above $100B in FY27 then ~20x earnings looks ridiculous for a business still scaling custom compute and the networking around every cluster.

Stock Ratings [June 7th]: On current AI sector crash. Explanations below. Strong Buy: $GOOGL $MU $SNDK SK Hynix Buy: $AMZN $AEHR $AAOI $CIEN $COHR $CRDO $DELL $FN $FORM $GLW $JBL $LITE $MDB $MRVL $MSFT $NBIS $NOW $NVDA $RDDT $RKLB $SIVE Hold: $ARM $ASML $AVGO $AXTI $BE $META $MTSI $PLTR $SOFI Avoid: $CBRS $CRWV $ETH $HIMS $IBIT / $BTC $IREN $MELI $SNAP $TSLA $SPCX (SpaceX) IPO --- Thoughts: Strong Buy: GOOGL - $85B raise is dilutive but they actually have ROI on their capex. Tbh, they'll probably always be a Strong Buy for me. Just the cleanest AI ROI among all the megacaps. MU / SNDK / SK Hynix - If you're not bullish on memory, then idk for you. Buy: AMZN - Mainly for AWS reacceleration + Trainium. But some tension comparing AWS growth (+17%) vs Azure (+31%). Feel like custom silicon + distribution combo is durable even if growth rate lags a bit. AEHR - H2 ramp in WLBI/PLBI systems coming, anchored by "significant" follow-on Sonoma order from lead hyperscale customer. Just need to wait a bit esp. for rev to inflect. But AI ASIC burb in is mandatory as device power goes up. AAOI - Q3 capacity ramp (via facility expansion in Texas) toward 650k+ 800G/1.6T units/mth. Capacity coming online is the catalyst imo along w/ already known laser bottleneck + Made in US premiums. CIEN - Just a high quality biz that got pounded last week (-22%). Beat + raise earnings, but stock dropping this much is an overreaction. CEO even said demand is "structural, multi year and AI-driven" shown by AI-driven DCI being their fastest growing part of the order book as new long-haul routes get built for latency and bandwidth. COHR - upcoming CPO ramp (Nvidia spectrum-x) will speed things up, these prices will look cheap when we look back imo. CRDO - Personally bought a ton last week post-earnings drop. Like Ciena, v. high quality compounding hold through the whole AI supercycle. Crazy high margins. Obviously compete w/ Marvell/Broadcom on SerDes, but also need to factor in the 1.6T switch replacement cycle into late 2026. DELL - Trump effect. I've learnt my lesson and will listen to him next time. FN - v. low drama way to ride transceiver demand + iPronics sipho line for cpo. New datacom wins also extending into next FY, although some Nvidia conc. risks. Put them in Buy just to be generous as was unsure tbh. FORM - Important for HBM, adv packaging and CPO for higher yields. Foundry test intensity only set to increase w/ production. GLW - Lead glass core substrates which are an advanced packaging bottleneck. LTP w/ Nvidia to expand US optical manufacturing for AI infra too. JBL - Stock has done nothing for a month, but earnings coming up could be a nice catalyst for a push higher from their DC infra segment growing + outpacing drag from legacy mobility/ev exposure / margin mix. LITE - CPO ramp + Nvidia qualification like Coherent. MDB - AI is not replacing them. Imo they win vs. bolt on vector stores since their architecture is so simple. MRVL - going to $1T according to Jensen. Underlying business is solid though esp. w/ Celestial acquisition for photonics. SPY inclusion last week too is a big positive. MSFT - Current valuations are a joke tbh, markets probs punishing some margin compression. Rev +18%, Azure +40%, AI run rate +123%. So, v. clear enterprise monetisation path. Will be buying next week in retirement account. NBIS - Best neocloud by far. They're a $100B biz vs. ~$57B currently. Jensen: "Nebius will take care of you." NOW - AI is not replacing them. No enterprise CEO/CTO is dumb enough to offboard them at this point. NVDA - Same as Microsoft. Been buying this whole time, but am now even more confused at current cheap valuations. RDDT - AI is not replacing them. Cash printer. ARPUs improving also in legacy segments like international. RKLB - #2# in commercial launch after SpaceX + their IPO should re-rate the entire space comp set where RKLB is the main liquid proxy. Unbelievable earnings also, just executing so well rn. SIVE - everyone on X knows at this point? Hold: ARM - current valuation prices in flawless execution imo. But their IP is growing in DC CPUs e.g. Nvidia grace, AWS Graviton etc. ASML - Elon said yesterday: "ASML should be treasured and supported. It is arguably the greatest company in Europe." - I agree. Also Terafab fireside chat next week High-NA EUV is the next leg, locking in the roadmap through the decade. Could also be a "Buy" for more risk averse people. AVGO - CEO didn't raise >$100B FY27 target + flagged that Google will multi-source. Current AI mix is also diluting margins slightly. Just needed a pullback before the thesis starts working again. AXTI - InP substrate bottleneck, crucial for AI buildout rn. Could also buy rn, just a slow dca since they've run up a ton already + raise completed ($632M) to 2x InP capacity. BE - SOFC winner imo (Ceres 2nd). Don't think it's a buy just yet due to some valuation vs. profitability gaps. META - hold based on capital allocation mainly. Market seems wary of the ROI on their AI capex hence the continuous dips. Also potential raise to fund capex like Google too - once that digests, I'll personally look to buy. MTSI - Big fan of their investment into $IQE since it de-risks operations a lot, but just think COHR/LITE are better options for 800G/1.6T transition. PLTR - Relatively poor Risk:Reward at current multiples. SOFI - rate sensitivity. Loan book + credit performance carry macro risk which caps conviction rn. Some positives though w/ young + growing member base. Would need to look at credit trends + Fed path in June FOMC to re-assess. Avoid: CBRS - avoid at current prices. Would want it to come down closer to ~$40B mc before I look to dca. Would love to hold since they own genuinely unique tech. CRWV / IREN - Financing for both is a mess...debt/dilution. Nebius are just a better multi yr neocloud. HIMS - Forced out of higher margin GLP1s into lower margin braded GLPs from Novo/Lilly. Feel like their moat was to do w/ regulatory arbitrage on compounding. With that gone, it's a customer acquisition + churn biz buying branded drugs at lower margin. IBIT / BTC - Macro setup is hostile. Higher rates for longer (10Y ~4.54%, 30Y >5%) raise opportunity cost. Pure liquidity/risk appetite instrument + both are tight rn. ETH - same as bitcoin. MELI - personally a little confused - either a hold/avoid. Seeing some margin compression via their credit book growing faster than revenues. Talks of margin recovery next year, at which point the stock could re-rate. SNAP - Absolute worst social media app + CEO is a weirdo. Platform keeps losing share to Meta/Tiktok. TSLA - Huge competition from other EV makers shown by production > deliveries volumes. Humanoids will be their next key growth driver, just a little while away. SPCX (SpaceX) IPO: I never personally participate in IPOs + SpaceX specifically is way too overvalued for me. Will be going long eventually though. Rough ballpark would be ~$1.5T if it gets there post IPO. --- Just for very high level notes at current stock prices (NFA). I'm personally staying long despite the current macro backdrop, mainly in AI supercycle names e.g. memory, semis etc. But then you also have great companies at depressed prices, mainly in SaaS which I'm DCA'ing currently. I don't hold positions in all of these names. This is just a subset that overlaps my "Close Tracking" list + X's favourite names.

This chart shows just how much bigger the AI infrastructure boom is about to become. UBS estimates the annual rack capacity deployed for AI will increase from 9.5 GW in 2024 to 104.5 GW in 2030. That is an elevenfold expansion in only six years, with most of the acceleration happening after 2026. The important part is that every new gigawatt of AI rack capacity creates demand across the entire infrastructure stack. More racks mean more GPUs and custom chips, more HBM, more networking and optics, and significantly more spending on power and cooling. Now here are some of the stocks that will benefit from this. $NVDA remains the largest winner because its annual rack capacity is expected to increase from 5.9 GW in 2024 to 50.5 GW by 2030. This means Nvidia could still represent almost half of all new AI rack capacity despite growing competition. The biggest change is the rise of custom AI chips because AI ASIC capacity is forecast to increase to 45.1 GW by 2030, nearly matching Nvidia. These chips are designed for large customers such as Google, Amazon, Meta and Microsoft that want better cost and power efficiency for specific workloads. This does not necessarily mean custom chips replace Nvidia. Nvidia should remain the preferred platform for flexible training workloads because of its hardware and CUDA software ecosystem. Custom ASICs are likely to gain share in large, predictable workloads where hyperscalers can optimize the chip for their specific models. $AMD also benefits but UBS expects it to remain a smaller third player. Its projected annual rack capacity reaches 8.9 GW in 2030, giving it significant growth without suggesting it will overtake Nvidia or custom silicon. The larger opportunity extends beyond the accelerator companies. Whether a rack uses an Nvidia GPU, an AMD GPU or a custom ASIC, it still needs HBM, advanced packaging, networking, optics, storage, power delivery and cooling. $MU, $SKHY and Samsung should benefit from rising HBM and server-memory demand. TSMC, ASE and Amkor should benefit from manufacturing and packaging increasingly complex AI chips. Broadcom, Marvell, Credo and Astera Labs should benefit from connecting those chips inside and between racks. Super Micro, Dell and HPE should benefit from building the servers and racks around the accelerators. Vertiv, Eaton, Schneider Electric and Modine should benefit from the power and cooling challenge. AI racks are becoming more power-dense, meaning data centers need more advanced electrical equipment and liquid-cooling systems. Bullish on the entire AI infrastructure stack and I’m already positioned across memory, networking, power, cooling, and neoclouds, and if you want to see exactly what I hold to capture this buildout, check out my full Milk Road Pro portfolio below.

Interesting. According to FundaAI, Qualcomm is expected to begin shipping an LPU-like AI ASIC to a Chinese CSP by the end of 2026. Estimated shipment volume is around 1 million units, with an ASP of roughly $4,000. For general-purpose server CPUs, shipments are expected to begin in the second half of 2027, or toward the end of 2027. 3 million units… that’s a lot. Qualcomm is also reportedly working with two U.S. CSPs. $QCOM