@OnodaCapital Orbital doesn't work until you figure out 400 ft radiator mass
English
Elapsus
179 posts
@aleabitoreddit Why is Leopold so bullish on $IREN he seems smart enough to realise the above yet still holds so much of it
English
$IREN back down -34% from $70 to $46.
I wonder if one of the dumbest communities on X finally learned to read?
$NBIS is objectively the better Neocloud, with actual financing.
-> Nvidia didn’t fund $IREN at all. They got a free purchase agreement to let IREN use their logos and dilute for GPUS.
$NVDA actually gave $NBIS capital.
-> $IREN is facing endless dilution like $BKKT, $ASST, $SLNH as retail wealth transfers capital over from $6,000,000,000 ATMs, on a dwindling “5 GW capacity” moat.
$NBIS actually uses equity appreciating financing structures.
And this is reflected in the YTD differences between them both.
I’ve said the same thing last year too. One is up ~100%.
The other is flat, and even negative depending on entry points.
IREN is literally a marketing company at this point by how they manage to convince retail to wealth transfer over capital.
Serenity@aleabitoreddit
I still am bearish on $IREN. Algorithms/retail probably read $NVDA + $IREN partnership and bought it up. However, if you look at the realtity, it's just looks like brand agreement giving $NVDA risk-free convertible notes. So $IREN can continue selling their $6,000,000,000 ATM into retail investors. It's the equivalent of a startup using AWS and saying they have an Amazon partnership so give them $6B. This wasn't Nvidia directly funding $IREN yet, just a risk free option to. There's a "5 GW deployment" but I'd rather not be the one buying into the dilution to fund it.
English
Seeing concern now that this Karpathy move indicates Anthropic already won and is therefore bearish for the other labs.
My take is this indicates that we are close to RSI and therefore an accel in model IQ increases.
In that scenario the value of compute is going to explode as supply chain scale ups are linear while demand-creation is non-linear.
Anyone with compute is sitting pretty regardless of lab talent.
Full stop. Every GPU will explode in value. If you can run a million Von Neumann in a datacenter we will quickly have AI inventing use cases for token consumption faster than we can supply them.
New fields of science. Reverse aging. The goonasphere.
It all gets pulled forward and it will all require compute.
English
@wallstengine my free GPRO stock from Stake finally about to pull an Nvidia
English
GoPro won its patent case against Contour after a California federal judge invalidated the remaining patent claim and wiped out the $8.2M damages award. $GPRO now faces no liability in the case, though the ruling can still be appealed.
English
@MichaelMOTTCM PPT would fill orders in just to make sure Nvidia beats earnings est.
English
Nvidia's Earnings Could Pop The Options Bubble
seekingalpha.com/article/490578…
English
@Bluebearmonkey @lithos_graphein Do you realise the software can fry the hardware?
+10 to your social credit score for the NPC answer though
English
@ausydau @lithos_graphein Software killed machines should not be difficult to reverse.
If not, having the physical machine to study is valuable in itself.
English
@lithos_graphein How exactly is the US going to do that? The US Navy won’t be anywhere near Taiwan.
A missile attack on Taiwan from Japan or the Philippines would be seen as an attack on China.
So no… nobody is touching TSMC.
English
"Shifting from eFlash to RRAM (ReRAM) and MRAM for advanced nodes (≤28nm) in Auto, AI Glasses, and Edge AI."
So this is why the shitco Everspin is rising
Dan Nystedt@dnystedt
TSMC 2026 Technology Symposium (Taiwan 5/14) Summary: Global Expansion & Operations · Global Build-out: Constructing or modifying 18 fabs worldwide (includes 5 advanced packaging plants). · Taiwan Core: 12 of these 18 facilities are located in Taiwan. · Acceleration: Expansion speed has doubled to 9 new plants per year (2025–2026) vs. 4 per year (2017–2024). International Site Progress · Arizona (USA): -Fab 1 output to grow 1.8x in 2026. -Fab 2 equipment move-in 2H-2026 (N3 production 2H-2027). -Fab 3 broke ground 1H-2025. -Fab 4 and first advanced packaging plant (AP1) are in initial construction. · Kumamoto (Japan): -Fab 1 (28/22nm) yield already equals Taiwan levels; 2026 output target is 2.3x YoY. -Fab 2 began construction in 2025 for N3 production. · Dresden (Germany): -ESMC facility focusing on 12nm–28nm for Automotive/Industrial. Advanced Process Roadmap (N2 / A14 / A12) · N2 Momentum: Fab 20 (Hsinchu) and Fab 22 (Kaohsiung) are in mass production. Taichung Fab 25 to follow in 2028. · N2 Performance: First-year output expected to be 45% higher than N3's first year. Capacity expansion CAGR of 70% seen 2026–2028. · Yield Maturity: N2 yield learning curve is 2 quarters ahead of N3 at the same stage, driven by AI-optimized manufacturing. · A14 (1.4nm): Already achieved >80% yield on 256Mb SRAM; offers 10–15% speed boost or 25–30% power reduction vs. N2. · New Nodes: -A12 (Performance-focused with Backside Power/Super Power Rail) and A13 both targeting 2029 production. -N2U(enhanced 2nm) slated for 2028. · Next-Gen: CFET architecture reduces SRAM cell area by 30%vs. Nanosheet. Advanced Packaging & Silicon Photonics · CoWoS Scale: 5.5x reticle size now at 98% yield. Scaling to 14x reticle for 20 HBMs by 2028 and 24 HBMs by 2029. · System on Wafer (SoW): Future "Super Exchange" chips to integrate 64 HBMs/16 CoWoS units, exceeding 40x reticle size with 100TB+ bandwidth. · Growth: 3DIC/CoWoS capacity CAGR of >80% 2022–2027. · COUPE: All-optical interconnects to replace copper as data centers scale to millions of GPUs and power needs surge 200x. Specialty & Memory Shift · The Memory Divorce: Shifting from eFlash to RRAM (ReRAM) and MRAM for advanced nodes (≤28nm) in Auto, AI Glasses, and Edge AI. · Edge AI Specs: -N4PRF (RF process) for 39% power reduction -N16HV(16nm High Voltage) specifically for AI Smart Glasses/AR displays. $TSM $NVDA $AAPL $AMD $AVGO $MRVL $INTC $AMZN $GOOGL $MSFT $META #semiconductors Links: news.cnyes.com/news/id/6458515 cna.com.tw/news/afe/20260… ctee.com.tw/news/202605147… money.udn.com/money/story/56…
English
AMAT’s strength is that even if each individual tool may lag competitors in performance, AMAT has an extremely broad equipment portfolio, allowing it to propose bundled solutions across multiple tools. That gives them a structural advantage over companies that focus on refining individual tools one by one. After all, a fab does not run on a single piece of equipment.
$AMAT
English
Korean value stocks in 2026 are one of the largest pools of mispriced equity left on earth, sitting in hundreds of profitable, debt-free, asset-rich small and mid cap companies trading at fractions of liquidation value, in a developed market with a functioning legal system, run by founding families who have hoarded cash for two generations and are now, for the first time in 30 years, under coordinated political and regulatory pressure to do something about it. The Corporate Value Up Program launched in February 2024. The Value Up Index is now promoted by the Korea Exchange ahead of the KOSPI itself. The dividend tax was cut from 45% to a range of 14 to 30% in December 2025. Activist campaigns, both domestic and foreign, are arriving at a pace the market has never seen before. You build a basket of 30 to 50 names, sized small, hold for a decade, and let the math do the work, because the catalyst is no longer a thesis you have to defend, the catalyst is the price itself combined with a regulatory regime that is, in real time, dismantling the structural reasons the discount existed in the first place.
English