Andrew Schmitt

@Truegreta This is so obviously AI the people are standing in exactly the same place.

I stood at this pool, at both monuments and saw both reflections… He’s a God damn idiot, as are the fools that support him. The “Reflection Pool” wasn’t designed by American architect Henry Bacon a hundred years ago to look like a swimming pool. It’s designed to have a darkened characteristics that has reflective qualities to reflect the monuments. That way, the Washington Monument is reflective to you when at the Lincoln Memorial, and when at the Washington Monument, the Lincoln Memorial is reflective to you. It’s designed to enhance the grandeur of monuments, create an illusion of reflection, and inclusion of expansive space of unity. He’s a tacky vulgar person that vulgarizes everything he touches. America isn’t becoming great, it’s becoming vulgar. Credit - Mathew Reed

@bentboolean @insane_analyst He's legit good for it.

@insane_analyst I’ll totally take you to lunch / dinner :)

I dont charge money but freer food is an exception I accept.

@insane_analyst Still waiting for remote surgery and self driving cars that eliminate the need for stoplights.

That bullshit was promised in 5G. Don't beleive any of this crap. 5G and 6G both economic hoax.

@insane_analyst I’ve made this argument for decades for different semis without success. In the end public executives and investors value margin more than margin dollars. Fair warning.

Bro do you even book value? HBM post in progress. To the retards who think I am bearish memory, do you think semis world want X GB of capacity at 80% gross margins or 3X GB of capacity at 60% GM? Everyone wins in second scenario. HBM is a mistake and you all too retarded.

@insane_analyst Exactly. But copper links will operate at this rate too.

@aschmitt Coherent lite tranceivers?

If MTK TPU succeeds, SMTC is a 2-4x from here. If MTK TPU fails, SMTC is a 1.5-2x from here. SMTC CTLE can give 4-7 dB SNR gain. ~300G SerDes at PAM6 has 4-5 dB SNR penalty from theory. MTK would need 2-4 dB improvement in design + CPC + SMTC CTLE + good EMIB pkg to make it.

@peptidecap @FundaAI @CignalAI That's great really helpful.

@aschmitt @FundaAI @CignalAI There is a particular coherent lite vendor that most public markets folks do not know about that is going to rip because of this…

From some discussions we came across today on TPU v9 networking, the direction seems broadly consistent with what we had touched on in our earlier report. The next step-up in Google's TPU networking bandwidth appears to come from two directions. First, the new scale-out network architecture, which seems to be increasingly corroborated by the Boardfly direction and Virgo Network. Second, the continued upgrades in scale-up / ICI bandwidth. We had also flagged the possibility that future TPU systems could move toward a higher-dimensional torus topology, potentially a 6D torus. If TPU v9 upgrades the topology, optical module speed, and port ratio at the same time, a roughly 4x increase in ICI bandwidth versus TPU v8 may not be entirely out of reach. This is likely not just a matter of "buying more optical modules." It could reflect a system-level architecture shift in Google's TPU roadmap toward higher bandwidth and higher optical interconnect content. One point that did not get much airtime in today's discussion, but which we had spent more time on in our earlier report, is the potential introduction of coherent-lite technology. If coherent optics does make its way into this part of the architecture, the dollar content per optical module could rise meaningfully — in some designs, perhaps even close to doubling. It would also bring in a more complex modulation stack, which may call for more advanced silicon photonics, thin-film lithium niobate, or other higher-end PIC platforms. That matters because the value migration would not be limited to the module layer. It could also benefit upstream components, including lasers, modulators, PICs, DSPs, connectors, and advanced packaging. If this plays out, the TPU networking upgrade would meaningfully shift more AI cluster value into optical and connectivity content. Names we'd keep an eye on as potential beneficiaries include $LITE , $COHR , InnoLight, $NOK , $MRVL , $SMTC , $AXTI , and $TSEM .

We've released our first CPO forecast based on ELSFP shipments. We aren't late. In the infamous words of Google's Hong Liu and others - CPO is a technology that is always arriving in N+2 years, where N is the current year.

One of the things Musk vs Altman shows is how much more promising AI is than anyone expected. Sam could have started it as a for-profit company. His life would be much simpler now if he had. But he didn't realize in 2015 that AI would warrant more than you can raise in donations.

@insane_analyst You did see the movie and know what happens to the Death Star right?

@FundaAI What’s your source for 2.4T at Innolight today. I don’t see anything disclosed today.

Research| $LITE : From Cyclical Optical Name to AI Datacenter Bottleneck LITE rallied today, and on the surface it looks like passive flows tied to the Nasdaq-100 inclusion. But the real story runs deeper. Last week’s print already validated the demand intensity in 1.6T, InP/EML, and OCS — revenue, gross margin, and operating leverage all started to come through. So today’s index flow is more of an catalyst; the substance is that the market is beginning to reprice LITE from a legacy optical components cycle stock into a critical bottleneck asset in AI datacenter optical interconnect. In China, Innolight’s investor communication today flagged 2.4T and coherent-lite, and that signal is worth paying attention to. The market is no longer just trading the 1.6T ramp — it is already looking ahead to the next-gen architecture. 2.4T / coherent-lite is not a simple bandwidth upgrade: it implies higher lane rates, more complex modulation schemes, longer-reach links, and substantially higher photonic-electronic integration requirements. The higher the technology bar, the better positioned the names with real depth in EML, InP, laser, and OCS/CPO — namely LITE and COHR. We flagged this technology direction two months ago. Detailed Report fundaai.substack.com/p/researchlite…

@insane_analyst I feel you. Except signal processing.

I failed the final exam of my first undergrad upper division analog design class but dammit its time for redemption. That class led to 50 hour weeks JUST FOR THAT CLASS, mostly locked in a lab with no windows.

@aj_kourabi There is no way those guys are doing $100m. Show the receipts.

slop like this is what i point to when people ask me how can SemiAnalysis make $100M

Maybe maybe maybe

@bentboolean the strategy is to be maximally insane yet still squeak out a victory against arguably the most unpopular leadership to date

In addition to why your rent is so expensive and your health care is so expensive and the cost of college is so expensive…. Modern progressive politics is a death cult.

I’ve been saying this for years. CPO is a process, not a product. And he’s right that TSMC is going to use this process to build a moat.

@insane_analyst Gas for car > New Phone

Im wondering if long memory short Qualcomm is worth sizing huge. Like 20%+ position. I think high end smartphone is also gona hit 30% demand destruction minimum by H2. Gona buy a $20 replacement battery kit for my S23 ultra. Great phone, degraded battery. Need to sleep.

@techjeet5 @insane_analyst Yeah well nothing is more confusing than “Coherent” in conversations.

Whoever named Semtech and Samtec and Santec can't keep getting away with this

We've shipped our OFC show report cignal.ai/2026/03/ofc-20… - CPO is Imminent - New MSAs - Coherent Lite Lives – Google, TPUv9, OCS - Optical Circuit Switching Accelerates - ELSFPs - Broadcom Ushers in 400G - Pluggable Coherent - Multi-Rail Line Systems - Full-Band Transponders