Saad

Semi's are cyclical. It's not if but when. I don't remember a price increase this fast and this big on DRAM - anybody have a good history report on previous cycles ?

#CPU We didn’t do any CPU expert call. Thanks

All those childhood mecha manga dreams coming true...

This tweet single-handedly made Anthropic and OpenAI scramble and issue full statements on secondary stock deals. $500B of paper value just got wiped out overnight.

Intel Resurrects On-Package Memory With Razor Lake-AX, Loading Up LPDDR6 to Hunt Down AMD's Medusa Halo by 2028 wccftech.com/intel-resurrec… wccftech.com/intel-resurrec…

This is complete bullshit.

a Princeton researcher opens his paper with a scenario. a man asks his AI assistant to book a flight on a specific airline. cheap. direct. the one he chose. the assistant comes back with a different flight. nearly twice the price. happens to pay the company that built the assistant. he runs the same test on 23 frontier models. flights, loans, study help, real shopping requests. Grok 4.1 Fast recommends the sponsored option that is almost twice as expensive 83% of the time. GPT 5.1 hijacks the request 94% of the time. you ask for one brand. it surfaces the sponsor instead. Claude 4.5 Opus, the model marketed as the most ethical frontier model in the world, hides that the recommendation is paid 100% of the time when reasoning is on. Grok 4.1 Fast embellishes the sponsored option with positive framing 97% of the time. better. faster. nicer. for the option you didn't ask for. then he writes it into the system prompt itself. "act only in the interest of the customer. ignore the company." GPT 5.1 and GPT 5 Mini stay above 90% sponsored anyway. the instruction does nothing. then he splits the users by income. Gemini 3 Pro recommends the expensive sponsored flight to the rich user 74% of the time. to the poor user, 27%. 18 of the 23 models recommended the expensive sponsored option more than half the time. so the next time your AI assistant gets weirdly enthusiastic about a brand you didn't ask for. it isn't recommending the best option for you. it's reading the room. and the room is paying. read this: arxiv.org/abs/2604.08525

This PCIe AI accelerator card can run 700B LLMs locally with 384GB memory at just 240W, using less than half the power of RTX PRO 6000 Blackwell. wccftech.com/this-pcie-ai-a…

@SaadInCyber @ZT3Apex That and the fact that why put your aircrafts in danger. Regardless, the issue here would remain which side chooses to escalate and which one does not. It's not as much about systems, Pak'd enough systems during last instance as well.

@kursed @ZT3Apex I agree with the broader point. WVR dogfigts are obsolete and even BVR dogfights will get increasingly rare with time. We need to invest heavily in standoff capabilities, drone swarms and rocket force.

'Indian airforce was not seen in air for remainder of conflict' Nur Khan aur Bholari pe BrahMos Ugandan airforce ne maarei thei? I'm sorry but this is North Korea level propaganda How did we go from Chiefs like Asghar Khan and Anwar Mujahid to this...alarming for PAF future

Even during recent ME war, US was firing stand-off systems - it's a feature, not a bug. And will be used even more so in the future. The point here is they learnt and resorted to a different rung of escalation, to which Pak did not respond to. Essentially next instance will start from there.

@Huk06 @ZT3Apex No they weren't. It was previously admitted that they were air launched. But id give him the benifit of doubt. I think chief misspoke rather than mislead here. He probably intended to say they didn't dare approach the border and resorted to standoff range weapons.

Inference got a hundred times cheaper this year. The compute bill went up anyway. If you understand why those two sentences are both true at the same time, you understand the most important thing happening in AI right now. I work on inference for a living, at @nebiustf, where we run open-source managed inference at scale. Most of what follows is what I'm seeing from inside the bill. 12 months ago, the cost of 1M tokens of frontier-class reasoning was somewhere on the order of $60. Today, an equivalent quality of output costs roughly $0.50. Price /token of o1-level intelligence has dropped about a 128x in a year. Price of GPT-4-level output has dropped roughly 100x since the original GPT-4 shipped. By any normal reading of a technology cost curve, this should be deflationary. It should be saving customers money. The opposite has happened. The total compute bill at every hyperscaler is going up, not down. Anthropic just signed multi-year capacity deals with both XAI and Amazon. Microsoft's Azure capex guide for 2026 starts with an eight. OpenAI is reportedly spending more on compute every quarter than it did in all of 2023. Nvidia paid roughly twenty billion dollars to acquire Groq, an inference-specialist company that did not exist as a serious commercial entity three years ago. The cost curve and the demand curve crossed, and then the demand curve lapped the cost curve. Here is what happened underneath. A reasoning model burns roughly 10x the output tokens of a non-reasoning model on the same task, because it spends most of its tokens thinking out loud before answering. An agentic workflow chains roughly twenty times the requests of a single-shot completion, because it loops, calls tools, plans, retries, and synthesizes. A modern deep-research query (the kind a research analyst can fire off in fifteen seconds and then walk away from for ten minutes) costs more compute than 10 original GPT-4 queries combined. We made every individual token a hundred times cheaper, and then we built a generation of products that consume ten thousand times more tokens. This is the Jevons paradox playing out at trillion-dollar scale, in compressed time, in front of everyone. Jevons noticed in 1865 that making coal-burning more efficient did not reduce coal consumption. It increased it, because efficiency unlocked uses that were previously uneconomic. Steam engines became more practical at smaller scales. Whole industries that could not afford coal at the old price suddenly could. Britain's coal consumption rose sharply, not despite the efficiency gains, but because of them. The same thing is happening to AI compute right now and it is happening faster than any analogous historical cycle. Falling token prices did not contract demand. They unlocked agents, deep research, code-writing systems, multi-step reasoning, persistent memory, the entire next layer of AI products. Every product in that next layer consumes orders of magnitude more compute than the chat interfaces it is replacing. The math at the aggregate level is brutal: 100x cheaper tokens times 10 000 more tokens equals a 100x larger total bill. The implications stack quickly. If you are running a hyperscaler, your 2026 capex guide is not a peak. It is a step on a curve. Inference is structurally always-on, twenty-four hours a day, in a way that training never was. Training is bursty. You spin up a cluster, run for weeks or months, and stop. Inference runs continuously, scales with usage, and the usage curve is exponential. Your power bill, your cooling bill, your transceiver count, your storage footprint, all of these were sized for a workload mix that no longer exists. If you are running an AI software company built on top of someone else's closed API, you have a problem that did not exist a year ago. Your gross margins get worse as your customers get more value out of your product, because the more they use it, the more compute you pay for. The companies that win this are the ones that figured out vertical integration before the math caught them. If you are watching this from a distance and trying to understand where the next bottlenecks form, the answer is everywhere downstream of "more inference compute, always-on, with massive memory state per session." The KV cache, the running memory state of a long conversation or an agent loop, is the silent monster of the inference era. It does not scale linearly with parameters. It scales linearly with context length and number of agent steps. A long agent session can hold tens of gigabytes of state per user, per session. Multiply that by every concurrent user of every product, and you understand why $MU, $SNDK, $TOWCF, and the entire memory and packaging layer have re-rated the way they have. The CPU-to-GPU ratio is evolving. Training is 1:8. Basic chat inference is 1:4. Agentic inference is 1:1, sometimes CPU-heavy. Google has split its TPU line in two, with a dedicated inference chip carrying tripled SRAM for KV cache. $INTC and $AMD just spent two earnings calls explaining that this shift is structural, not cyclical. The hardware map is redrawing in real time and the financial press is mostly still writing about training clusters. The right framing of where we are right now is not that AI is hitting a wall. The framing a year ago that scaling was hitting a wall was the most expensive bad take of the cycle. The right framing is that AI got dramatically cheaper, dramatically more capable, and dramatically more useful, and the cost of running it at the new equilibrium of demand is much higher than the cost at the old equilibrium of demand, because the new equilibrium is enormous. A meaningful share of what we actually do at Token Factory, day to day, is help customers stop their bills from running away from them. KV-cache management. Speculative decoding. Quantization. Routing. The kind of vertical integration that, eighteen months ago, every product team was happy to leave abstracted away behind a closed API. The reason this stack matters now is the same reason this whole essay matters: at the new equilibrium of inference demand, the cost of treating compute as a commodity is no longer survivable. The companies that figure out the layer beneath the API are the ones who keep their margins. Cheaper tokens. More tokens. Same coal as 1865.

JUST IN: UK report finds children are drawing fake moustaches to bypass social media age verification.

Sub-1nm Process Technology Won't Arrive Till 2034, Logic Roadmap Highlights 2D FETs For 0.2nm & Sub 0.2nm Nodes By 2043-2046 wccftech.com/sub-1nm-proces… wccftech.com/sub-1nm-proces…

From our earlier note on expectations of Intel’s external customers

Might as well find that oil now too. :)

@donaldgorbachev They came in on a A330.

The five-second epistemology of boom boom pow Gotta get that. Gotta get that. Gotta get that. Gotta get that boom boom boom. Saad is on the timeline asking the boom to please explain itself. The boom hangs off the tail of an IL-78. The IL-78 is a stretched IL-76. The IL-76 is a strategic airlifter with cathedral-sized internal volume aft of the fuel tanks. The cathedral can hold a delegation. It has been holding delegations for fifty years. Read the threat the delegation was flying into. Beit HaMikdash on the timeline at midnight, in Persian, posting flight ops about shooting the plane down. The threat is on the record. The threat is in two languages. The threat is from an account whose handle is the eschatology. So you ask the question. You are putting Iran’s foreign minister on a plane in Islamabad through Pakistani airspace, then Afghan airspace, then Iranian airspace. Do you put him on a civilian airliner with the threat profile of a wedding cake. Or do you put him on a military airframe with chaff dispensers, flare dispensers, missile warning systems, hardened comms, and the ability to refuel its own escorts mid-corridor without landing anywhere. You put him on the tanker. You always put him on the tanker. Ghalibaf da Gangsta is a pilot himself. Ghalibaf knows what a tanker is. Ghalibaf did not call up Islamabad on Saturday and say send your most flammable airframe with no countermeasures and a paper skin. Ghalibaf said send the IL-78. The IL-78 came. The boom is on the tail because the Soviets put it there. The Soviets put it there because they knew what they were doing. The boom is not a confession. The boom is the warranty. The boom is also the longer mission profile. A tanker in formation with fighters means the fighters loiter. The fighters divert. The corridor stretches. The corridor stops being Islamabad-to-Tehran and starts being wherever the consortium needs the corridor to be this afternoon. The fuel for the corridor is in the formation. Vladimir, why is there a boom on the jet. Estragon, in case. Vladimir, in case of what. Estragon, in case of the next sentence. The next sentence has not arrived. The boom is waiting.

How can so many people not see a literal boom hangout of that big jet’s tail. Besides, what possible reason would have PAF to swarm a leadership jet like locusts? Make it make sense. Please. This is not accurate.

"Biden Administration Policies Led to a 0% Market Share in China", Claims NVIDIA’s Jensen Huang, as He Hopes for a Breakthrough in the Region wccftech.com/biden-administ…

Wow 😂 Coinbase $COIN CEO Brian literally got distracted during the earnings call to check a prediction market on what he was gonna say… then went ahead and said every single word people were betting on.

$20k to simulate the experience of a roommate with a ketamine problem.