Mojo

A full wash & wax on $NVDA #AI Ponzi scheme by @FT While they didn't get into $SMCI $DELL inventory stuffing & round tripping to their customers like $MSFT The entire house of cards on which #Coreweave & others funded by #Blackstone $MS $JPM etc is exposed This ends badly 💣

@FosterTraveler @danielnewmanUV #Taiwan needs LNG. The island's electricity runs on it. They have a few weeks left. Then $TSM has a real problem

@Mojo_flyin @danielnewmanUV Qatar produces 1/3 of world supply! Sourced as byproduct from NatGas production… prices moving up and longer could impact many fabs. No substitute….according to my simple search. Give few more weeks or long no bueno

@Mojo_flyin @danielnewmanUV Must read…watched early morning Bloomberg and they said same. POTUS ability and is over playing the Taco 🌮…this is not on anyone’s card

@Silicon_Fly Medium or well done please?

@Mojo_flyin I’d like a dozen pls

Considering the ASML backlog, has Elon already placed EUV orders with ASML for his new Terafab? Or is he gonna built them all by himself in his garage?

@theaustinlyons On that I agree. They need to talk a lot more on the role of CPUs and X86 inside enterprise agentic workflows @asheaw

@Mojo_flyin I’m talking about agentic AI CPU racks. Intel is not saying anything on the subject. But they should!

Didn't see this coming! The two merchant silicon CPU vendors dominating the agentic AI rack narrative? Nvidia and Arm. Not AMD, not Intel. Crazy.

Not sure how $ARM is winning on the narrative. If they were why would Jensen sign a technical collab with $INTC. Why would $INTC be signing LTCs with hyper scalers - something the hyper scalers have never done before and why is $INTC sold out on Xeon this year and likely next. The only at scale $ARM CPU in the market is $NVDA (let's ignore $QCOM) and they don't have a 100% attach rate with their own GPUs. They need $INTC Xeons too.

@Mojo_flyin Didn’t say Arm would win, just saying they are winning on the narrative. That said, are agentic CPU racks near GPUs where the legacy x86 workloads will run ? I’m not so sure.

@Gfilche Cos it's not real.

haven’t seen one mainstream media article on how the Tesla/SpaceX TeraFab is the West’s most ambitious, strategic and important infrastructure project in decades. which will cause a massive shift in all of geopolitics and business rippling for decades (if not centuries) to come

@aakashgupta @Perplexity has a lot of work ahead of it. Claude beats them hands down. Their wrapper is mediocre. I just cancelled my subscription and I have equity. Economic model is great. But user experience value creation needs work. @AravSrinivas

Perplexity is a $20 billion company that built zero AI models. Their product sits on top of 19 models made by other companies. Claude for reasoning. Gemini for research. GPT-5.4 for long context. Grok for lightweight tasks. Nano Banana for images. Veo 3.1 for video. You write one prompt. Computer picks the best model combo for the job, spawns sub-agents in parallel, and runs the whole thing in a cloud sandbox while your laptop is closed. 400+ app connectors. Gmail, GitHub, Snowflake, Salesforce, Ahrefs, Shopify. Read and write access. One prompt can scrape your competitors, pull live financials from FactSet, query your data warehouse in plain English, and push a finished report to Google Slides. No API keys. No terminal. The enterprise usage data tells you where this is heading. In January 2025, 90% of enterprise tasks on Perplexity ran on two models. By December, no single model held more than 25% of usage. A new frontier model launched every 17.5 days in 2025. Each one brought different strengths. The era of picking one model is ending. Perplexity built none of the intelligence. They built the routing layer that makes the intelligence usable. Stripe didn't build the banks. Google didn't build the websites. The value is in making complexity disappear. Four of the Mag Seven already use Perplexity's search API in production. Every model provider is now building orchestration in-house. The question is whether the routing layer stays independent or gets absorbed. I wrote the complete guide to using Computer without wasting credits. 6 use cases, the prompt spec that controls cost, honest limitations. aibyaakash.com/p/perplexity-c…

If anyone's interested, here's a summary of my work at @wccftech , particularly what I have been doing in the past few months: - Covering the entire AI supply chain, starting from upstream, which includes hyperscaler deployments and AI architectures, all the way to downstream, which includes semiconductors and supply chain elements. - Another front which I recently became interested in is predicting future compute trends, and the shift of workloads and the demands it will bring for compute manufacturers. This includes talking about disaggregated inference and latency-sensitive workloads within the AI stack. - Consumer/PC hardware reports as well, whether it includes the latest news, dissecting launches from unique angles, and telling users about what's best for them in terms of purchase. - Finally, connecting the AI supply chain with industry matters, and this includes quotes from executives, personalities like Jensen and Lip-Bu. I also take great pride in coming up with angles that have never been discussed in the mainstream media before. I have spun out a total of 2,800 pieces, including a mix of the rumor mill, analysis, op-eds, interviews, event exclusives, and more. My work has not only created an audience that spans 'hundreds of thousands' in readers, but has also received accreditation from C-suite executives, experts in the industry, and several people that I admire.

@jukan05 @Pirat_Nation Because edge and hybrid compute is critical for broad scale #AI. $INTC makes a boat load of high margin CPU money in the mid market. That GPU business was all. $NVDA. No longer.

@Pirat_Nation I don’t understand why Intel is doing this kind of business. Is it some kind of charity program to create jobs for GPU architects?

Intel launched the Arc Pro B70 GPU at $949. It has 32 Xe cores, 32GB GDDR6 memory, up to 367 peak TOPS for AI, and 608 GB/s bandwidth.

$NVDA #AI moat is developing cracks. Inference is entering the CPU era. $INTC $AMD

This isn't the first time $SMCI servers were found at a university research shop in #China with military links. $DELL was too. But this time $SMCI is well and truly f***ed The key question remains - will Liang take $NVDA down too? tomshardware.com/tech-industry/…

@FosterTraveler @danielnewmanUV Yes

@Mojo_flyin @danielnewmanUV Helium is shortage I’m reading …key component??

Undoubtedly #China is watching the #US war with #Iran & learning how to use low cost saturation asymmetric warfare to exhaust #Taiwan defenses Probably also observing the #US is depleting significant inventory & will be a diminished force if they even decided to ever assist #Taiwan All #China needs to do is block all semiconductor & electronic shipments in and out via #Taiwan Straits & the island's & the global economy grinds to a halt The ultimate leverage isn't nuclear. It's geography & economics. $TSM $INTC

Upgrading your RAM is now unnecessary. Introducing our new ComfyUI Dynamic VRAM optimization. Running local models is now possible on even the most memory constrained hardware. Read more here: blog.comfy.org/p/dynamic-vram…

#Asia is screwed. Singapore spot prices on #Dubai crude are at a significant premium to #WTI & #Brent The lesson they are learning is delever your economies from oil. As it is this war has accelerated #EVs for #China. It will do the same for #Asia $BYDDY $XPEV $NIO #Zeekr

Meant to share this insane chart.

@FosterTraveler @sentdefender @FosterTraveler if 👇🏼 is true they are now openly goading him. Even if he tries to take the off ramp until they get what they need from any settlement they won't let this rest. They have no choice because they cannot trust either US or their neighbours. x.com/i/status/20374…

@sentdefender @Mojo_flyin the account does good job on reporting… look at recent post…looks like more Iranian drones/missiles are hitting the ground …possibly indicating diminishing stock of anti drone missiles to stop them = possible soon cease fire? US claim victory before hits tally up?

Satellite imagery appears to confirm a U.S. Air Force KC-135 Stratotanker has been destroyed and several others possibly damaged, as a result of an Iranian attack earlier today, utilizing ballistic missiles and drones, against Prince Sultan Air Base in Saudi Arabia. Image credit: @tom_bike

NIO's "Shenji NX9031" chip, using 5nm technology, launches in the ET9 by March 2025. It offers high efficiency, low power consumption, and advanced safety features. Finances: $2.2 billion funding, $10 billion valuation.