
#AIStorageAndSoftware
About AIStorageAndSoftware
AI earnings are extending to storage and software. Nvidia validated compute demand; Marvell raised long-term goals but shares fell on order concerns. CXMT reported H1 revenue of RMB150.31B and net profit of RMB77.61B, sees tight H2 DRAM and LPDDR6 validation. Its turnaround also reflects capacity, utilization and memory-price gains. CrowdStrike, Salesforce and Okta saw improving results and guidance, showing AI monetization in orders and recurring revenue. Share your view under this topic.
Populaire
Récents
AIStorageAndSoftware Publications populaires
Épinglé
Sprint vers 1 000 milliards de dollars de revenus pour l'exercice 2028 ? Nvidia enflamme les nouvelles attentes de Wall Street
Titre original : « Sprint vers 1 000 milliards de dollars de revenus pour l'exercice 2028 ? Nvidia enflamme les nouvelles attentes de Wall Street »
Auteur original : Zhao Ying, Wall Street Insights
Le dernier rapport financier de Nvidia a de nouveau secoué le marché, la direction prévoyant une croissance des revenus de plus de 70 % pour l'exercice 2028, bien au-delà des 45 % attendus unanimement par Wall Street. Plusieurs institutions telles que JPMorgan, Goldman Sachs et Bernstein sont unan
$SNDK $BTC SanDisk and Kioxia plan to invest $31B in Japan to expand 3D NAND production, but the market isn’t cheering. The key question: can AI-driven enterprise SSD demand absorb the new supply? AI storage demand remains strong, but massive capacity expansion could pressure NAND prices and margins. The bigger signal: AI capex is rising, but markets now want to know—when does it pay off?
#WalshInflationRisk
#WalshInflationRisk
AI Chip Shortage Could Last Until 2030
SK hynix CEO Kwak Noh-jung says the current memory chip crunch could continue until 2030, driven by strong and growing demand from AI infrastructure.
The AI boom is not only about GPUs anymore. High-bandwidth memory (HBM) and advanced memory chips have become critical components for AI data centers.
If demand keeps rising faster than supply, the memory market could remain tight for years.
📌 AI demand → More data centers → More memory demand
$MRVL BEAT EXPECTATIONS — SO WHY DID IT DROP 8%? 🤔
Marvell delivered a strong quarter:
📈 Revenue: +37% YoY
🏢 Data Center: +46%
🚀 FY27/FY28 guidance: Raised
Yet $MRVL still dropped nearly 8% in pre-market trading.
What’s notable is that $SNDK, $MU, and $WDC also moved lower, while AI leaders like $NVDA and $AVGO remained relatively resilien
#BankTokensVsStablecoins #WalshInflationRisk #SpaceXRevenueBy2033

$31 BILLION BET ON AI STORAGE — BUT THE MARKET ISN’T BUYING IT.
SanDisk and Kioxia just announced plans to invest $31B in Japan to expand 3D NAND production.
Sounds bullish for AI infrastructure, right?
Not so fast.
The market’s real question is simple:
Can AI inference and data-center demand grow fast enough to absorb all this new NAND supply?
If yes, SanDisk’s long-term story stays intact, and this pullback could just be profit-taking.
#DailyOrbit
🚨 MRVL BEAT THE NUMBERS — SO WHY DID THE STOCK DROP 8%?
Marvell just delivered a strong quarter:
📈 Revenue: +37% YoY
🏢 Data Center: +46%
🚀 FY27/FY28 outlook: Raised
And yet, $MRVL still fell nearly 8% pre-market.
What’s even more interesting? $SNDK, $MU and $WDC also moved lower, while AI heavyweights like $NVDA and $AVGO stayed relatively steady.
📌 The message from the market may be changing: AI demand itself isn’t necessarily weakening — investors may simply be becoming
#DailyOrbit
Brothers, don’t just focus on GPUs. The ones quietly making money in this AI wave might actually be the memory sellers.
Previously, when we talked about AI, we kept mentioning Nvidia, AMD, and computing power, but now the real price surge is happening behind the scenes with HBM. Micron and SK Hynix are the two players on this chain worth watching the most.
#WalshInflationRisk #BTCGoldCorrelation #SchwabExpandsCrypto
The scale of the planned NAND buildout matters less than its sequencing. SanDisk and Kioxia intend to invest more than $31 billion in Japan through 2032, subject to government support, with the new Kitakami fab targeting FY2029 production.
That timeline creates a useful demand test: if AI inference, data retention and cloud workloads deepen before capacity ramps, added enterprise SSD supply could meet a durable market. If orders lag, the same expansion may pressure the NAND balance. Capacity milestones, customer commitments and cash flow should be more informative than headline spending. Not advice, just analysis.
#NANDCapacityExpansion
Here's some solid info for those bullish on AI and storage: $SKHYNIX Hynix's CEO said the global memory shortage will last until the end of 2030, and they're open to continuing to invest heavily in expanding production in the US. Sounds exciting, right? Memory shortages and computing power scarcity are indeed some of the strongest bullish arguments this year. But I have to pour cold water on that: no matter how strong the narrative is#WalshInflationRisk #BTCGoldCorrelation #SchwabExpandsCrypto
🔥 $SNDK | AI HAS A STORAGE PROBLEM
The next bottleneck in AI may not be compute — it’s where all that generated data goes.
SanDisk and Kioxia just announced plans to invest over $31B in Japan through 2032 to expand memory production as AI drives demand for NAND storage. $SNDK
The deeper thesis:
Every smarter AI model creates more data — and every byte eventually needs somewhere to live. 🔥$SNDK
#WalshInflationRisk #BTCGoldCorrelation

🔥 AI’S NEXT BIG WINNERS MAY NOT BE SELLING CHIPS — THEY’RE SELLING SOFTWARE.
AI hardware proved one thing: companies are willing to spend billions to build the infrastructure.
Now comes the real test: are they willing to keep paying for the software?
Strong results from $CRWD, $CRM and $OKTA suggest the answer could be yes. 👀
The AI story may be shifting from GPUs and data centers to recurring software revenue.
#DailyOrbit
