Astro

[Astro @AstrosPhoto]

As we enter 2026, here are three critical market themes for next year:

1. A "show me" market. Look at the CAPE ratio. Investors want earnings proof next year.

2. Fed policy. We going lower, higher, or not moving?

3. Conflicts - China/Taiwan, U.S./Venezuela, Iran/Israel

[Bilawal Sidhu @bilawalsidhu]

holy crap! apple just beat google to the punch -- 3d gaussian splatting is coming to apple maps.

these 3d scenes are made from oblique aerial imagery. but unlike blobby photogrammetry -- no more broccoli trees, no more melted powerlines -- ground level detail that actually holds up.

here's hoping google maps/earth follows suite soon -- they have a significantly larger corpus of sensor data to work with. time to splat the world!

[Astro @AstrosPhoto]

@SilentAlert1 @Hyliion What makes you think they can't?

[Astro @AstrosPhoto]

@USnuclearfamily $HYLN could offer more MW/acre than $BE with the benefits of fuel agnosticsim, and low noise and maintenance.

[Astro @AstrosPhoto]

@gautaminvests $EOSE sounds like the better storage bet for AI to me.

[How To AI @HowToAI_]

Google has published a paper that might end the transformer era.

For the last 7 years, every major AI, ChatGPT, Claude, Gemini, has been built on the exact same architecture: The Transformer.

But Transformers have a fatal flaw.

To remember context, they have to process every single word against every other word. It’s called quadratic complexity. As your prompt gets longer, the compute cost explodes.

The alternative is the old-school RNN (Recurrent Neural Network). RNNs are incredibly cheap and fast, but they have a fixed memory size. If you give them a long document, they get amnesia.

Until today.

Google researchers published Memory Caching: RNNs with Growing Memory.

And it fixes the biggest bottleneck in AI.

Instead of an RNN having a fixed, rigid memory that constantly overwrites itself, Google gave it a "save" button.

The technique allows the RNN to cache checkpoints of its hidden states as it reads.

The memory capacity of the RNN can now dynamically grow as the sequence gets longer.

They built four different variants, including sparse selective mechanisms where the AI actively chooses exactly which checkpoints matter most.

The results rewrite the rules of efficiency.

On long-context understanding and recall-intensive tasks, these new Memory-Cached RNNs closed the gap with Transformers.

They achieved competitive accuracy without the explosive, quadratic compute cost. It perfectly bridges the gap between the cheap efficiency of an RNN and the massive capability of a Transformer.

We have spent billions scaling Transformers because we thought they were the only way an AI could remember a long conversation.

But Google just proved we don't need to process the whole history every single time.

We just needed a smarter cache.

[Astro @AstrosPhoto]

@bdinvestingg Why should $MU $AAOI or $SNDK trade at a PEG of 1?

[Astro @AstrosPhoto]

@kishwarAI @ChairmansLedger For LNA, their focus was mostly on cost as innovation wasn't necessary for years except now. $AMPG for over 20 years had a focus on LNAs.

AMPG - 0.8 dB
$NOK - anywhere b/w 1.5-2.5 dB

Both in midband

Note: Nokia doesn't release numbers, but we know they follow standard LNA

[Astro @AstrosPhoto]

@kishwarAI @ChairmansLedger $NVDA broke up the telecom market by bringing their GPUs to cell towers. Now it's about increasing tokens per watt. Nokia LNAs are inferior to $AMPG

[Astro @AstrosPhoto]

That's not how it works

[Athu Invests @athuinvests]

$AAOI PEG is 0.7, below 1.0. Growth is still mispriced. That’s why $150-$180 is a VERY good entry, in my view.

[Astro @AstrosPhoto]

Not really, this seems like 2022 all over again. Stagflation is the narrative. Two rate hikes early 2027 is now the consensus. Bond yields are ripping. This is not an environment where high beta stocks perform. Friday was just a teaser.

[TW @TW_trades_]

Names to buy on today’s dip:

$AAOI
$FLNC
$SIVE
$NBIS
$SNDK
$XFAB
$DRAM
$NVTS

The sell off today seems like an over reaction.

Good time to add to ur positions 🙌

[Astro @AstrosPhoto]

@tw_crypto_ I’m worried about its transceiver business and the switch to CPO architecture. Nvidia has already standardized the architecture and everyone, including Oracle, is adopting it.

Lasers only a commodity like memory. I can’t apply a fwd multiple higher than 35.

[Astro @AstrosPhoto]

@Elfieeen @TheLongInvest Overall, the business is positioned to perform well if execution is there. They have a stock buyback program of $20m. With profitability, committment should be made by management to increase this over time, reversing previous dilution to fund growth.

[Astro @AstrosPhoto]

@Elfieeen @TheLongInvest 4. I agree with this point. One vendor is extremely risky and the business needs to diversify from SMCI. That should happen if the business scales.
5. There is geopolitical and FX risk (Egypt) that you didn't mention. I find these to be the highest risk for investors.

[L @Lg3ux]

The $LPKF AGM results are out!
Shareholder confidence is high as CEO Klaus Fiedler and the board received a strong 94.92%vote of approval for the restructuring course. With internal debates settled,the path is wide open for the highly anticipated Q2 LIDE semiconductor mass comm.

[Astro @AstrosPhoto]

@chinoalemano He’s 100% correct

[Astro @AstrosPhoto]

Data centers are rapidly evolving. AI is moving onto our devices (the “edge”), but signal loss from data centers isn’t power efficient, and energy is currency in the AI factory business.

Cellular towers will become neoclouds leasing compute to hyperscalers. $TMUS $AMPG $NVDA

[Astro @AstrosPhoto]

@chandra2685 @UncleAlpha007 @RJCcapital Error from SPAC merger

[Astro @AstrosPhoto]

$AAOI order book is wild lol