Anshuman Singh

[GameVerse @TheGameVerse]

GTA 6 has to top this mission.

[Anshuman Singh @anshuman_eth]

@DovySimuMMA He is enjoying it 😂

[Anshuman Singh @anshuman_eth]

@kilobolt_ @terencecrawford If Bud can do porn Onlyfans he can even make more money than 50m? Wtf is your logic? Everything is not about money. I have commented about different aspects.

[Sanju Choudhary @saaaanjjjuuu]

Bad News 🚨
PCB (Printed Circuit board) manufacturers might hike prices by 50% due to supply chain issues:

• Raw material costs have increased due to shortages of essential PCB material such as copper foil, epoxy resins, glass fiber, and laminates.
• Current PCB prices are not sustainable due to WAR
• They are seeking help from government
• PCBs are the foundation of almost every electronic device, including smartphones, tablets, computers, networking equipment, automobiles, and AI servers.

If this happens then the device manufacturing costs (including smartphones) will increase much more than expected.

[Anshuman Singh @anshuman_eth]

12 nm TSMC node is impressive 🫡

[Abhishek Yadav @yabhishekhd]

🇮🇳 INDIA JUST GOT ITS OWN AI CHIP.

Kerala-based startup Netrasemi, backed by Zoho, has launched its first Edge AI System-on-Chip (SoC) called A2000, and customer trials have already started.

Highlights:
⚡ Built on TSMC's 12nm process
⚡ Designed for smart cameras, drones, robots, CCTV systems and industrial automation
⚡ Can run AI tasks directly on the device instead of depending on the cloud
⚡ Supports real-time video analysis and computer vision
⚡ Built for high performance while keeping power consumption low

The chip has already completed testing and is now production-ready.

🔹 Edge AI means AI runs directly on the device itself instead of sending everything to cloud servers. This makes responses faster and improves privacy.

🔹 SoC (System-on-Chip) means multiple important components are packed into a single chip.

🔹 CPU (Central Processing Unit) is the general-purpose processor that handles everyday computing tasks.

🔹 NPU (Neural Processing Unit) is a dedicated AI engine that speeds up tasks like object detection, image recognition and AI inference.

🔹 VPU (Vision Processing Unit) is designed to process photos and videos efficiently.

🔹 ISP (Image Signal Processor) improves image and video quality before the AI analyzes them.

In simple terms, a device powered by A2000 can see, understand and react to what's happening around it much faster, whether it's identifying a vehicle on a road, detecting a person on a security camera, or helping a drone navigate.

A few years ago, the idea of a startup from Kerala building an advanced AI chip would've sounded unbelievable to many people.

India still has a LONG way to go in semiconductors.

[Anshuman Singh @anshuman_eth]

Boxing is a joke. What a horrible stoppage.

[Ariel Helwani @arielhelwani]

That’s an absolute disgraceful stoppage. Rico got jobbed. What the hell.

[Anshuman Singh @anshuman_eth]

@MiamiDogbtc @Josh_HokitUFC For your info, Ilia has 7 cold knockouts. Josh has 0 cold knockouts.

[Dexerto @Dexerto]

“Forza Horizon 6 delivers on a near flawless road trip through Japan and sets the benchmark for fun-first racing”

[Anshuman Singh @anshuman_eth]

@sonyxperia Wtf what kind of AI is this😂

[Anshuman Singh @anshuman_eth]

@madhavbanka_ 😂😂

[Anshuman Singh @anshuman_eth]

@McGregorRousey2 He was always a fraud. Good wrestling, bad cardio, super bad boxing.

[Anshuman Singh @anshuman_eth]

@DovySimuMMA Where can i find this full match🥵

[Wellu @Wellutwt]

In gore ki gamd 20-21 mai fatti jaa rhi yh maa k lavde 200 saal raaz kaise kr liye

[Paul Asadoorian @[email protected] @securityweekly]

For sure

[Shiv Aroor @ShivAroor]

LMAO! Pak PM @CMShehbaz accidentally posts tweet along with what looks like a White House instruction as a header. 🤦🏽‍♂️😂🤡

[Calmer Fun @akt206]

broo @MicrosoftEdge @brave

[Abhishek Yadav @yabhishekhd]

FINALLY memory chip prices might actually go down.

Google just released TurboQuant. It lets AI use 6x less memory and run 8x faster. Just a software update. No new chips needed.

Samsung, SK Hynix and Micron all crashed today because of this news. If AI needs less memory, companies buy fewer chips. Simple.

AI has been making everything expensive for the past 4-5 months. More RAM in laptops. Higher server costs. Even your phone got pricier. This could be the thing that slows all that down.

But one thing to know. TurboQuant only helps when AI is answering your questions. Teaching AI from scratch still needs the same amount of memory as before.

Also Google hasn't shared the actual code yet. Still just a research paper.

So prices may not fall tomorrow. But this is the first real sign that the AI memory madness has a limit 👇

[Anshuman Singh @anshuman_eth]

@MKBHD All look the same after 5