@lukadoncic@NBA@Lakers
Can you please FINE every player that complains to the refs instead of going back to DEFENSE?
This is SICKENING to watch this childish and unnecessary behavior, especially DONCIC.
Respect the people who pays to watch and stop acting LAZY.
@HendaAyari5 Juste français. Ne tombez pas dans le piège de la diversité. C'est déjà une lourde responsabilité culturelle et individuelle d'être français, d' avoir un terroir, des traditions à perpetuer et à transmettre.
GM frens!🍵
The number of teams onboarded on Phala Cloud is approaching 1,000!
Leave a comment and let us know how we can make Phala Cloud even better for you!
did DOOM really run on JAM?
recently, a Cardano fanboy claimed that running DOOM on JAM was no different than DOOM gameplay "secured" by Cardano's Hydra Heads L2 (a Cardano community event from a few months ago).
this is a terrible (or intentional) misunderstanding.
let me explain how JAM achieved something no other blockchain can.
first, let's be very clear: DOOM ran entirely on-chain with JAM.
but how?
when traditional blockchain games claim to be "on-chain," typically only assets or ownership records are stored on-chain. the actual game execution happens off-chain.
what happened with JAM was fundamentally different:
- the entire game executed on-chain - JAM's coreVM (using just 2 of its eventual 341 cores) processed all game code, physics, and logic directly on the blockchain
- game assets loaded from blockchain storage - textures and game data were pulled from JAM's 1.6 petabyte decentralized storage
- persistent state between blocks - JAM maintained the game state continuously, running at 30 fps without interruption
- decentralized memory management - JAM emulated a complete computer with its own memory management, providing 32-bit addressable memory
- native code execution - original C code of DOOM was compiled to RISC-V and ran directly in CoreVM without modifications
this is entirely different from approaches that just "secure" gameplay happening elsewhere. with JAM, everything from rendering to logic to storage happened on the blockchain itself.
this demonstration proves JAM can potentially replace centralized cloud services for virtually any application, but in a trustless, decentralized way.
Polkadots JAM model will be the first blockchain to be able to thrive without direct active users‼️
Blockchain technology was built as a backbone infrastructure for security and verifiable results. It wasn’t built to be some direct front end like all current Layer 1 chains are trying to do.
By front end I mean the chain can only survive as long as more users make wallets, buy their token, and pay gas fees for every tiny thing they do
Polkadots JAM upgrade could in theory no longer rely on any direct retail users.
This is because JAM will facilitate companies running applications that do not need direct retail uses interacting with their products. Such as:
1. Full AI Model Training & Execution (At Scale)
Why Only JAM?
No user-paid gas fees. AI models can run continuously without retail users funding transactions.
And Parallel execution running on native hardware (not a virtual machine layer) across multiple JAM cores allows AI training to actually be viable.
2. Fully On-Chain Scientific Simulations (Medicine, Physics)
Why Only JAM?
JAM allows real-time, high compute workloads.
Compare to Ethereum, Solana, or ICP, which all have execution limits that make long-running computations impractical.
This is because JAM doesn’t have a “gas limit” since it runs code natively on hardware
JAM puts poladot another step forward in turning blockchain technology into a truly functional tool outside of finance and basic decentralized internet infrastructure.
JAM can now grow beyond some system reliant on users paying gas fees just to breath. It could potentially survive with zero direct retail users
Think about it like this.
Most of the world’s companies don’t function in this model where their costumer pays a fee for everything that they do. companies normally build and run things that the end user doesn’t interact with at all. Thats what they can do on Polkadot with a JAM. And not just basic internet services, real heavy hitting high computing power tasks (which is the direction the world is heading)
This isn’t to say retail users are no important, and won’t be huge price driver for the foreseeable future. But JAM does finally open the doors to walk away from that model.
Bitcoin didn’t ask permission.
Linux didn’t lobby to be installed.
Gold didn’t need a sales pitch.
Real innovation doesn’t wait for approval—it proves itself.
Stop worrying about bureaucrats. Build something valuable, and adoption will follow.
This is @Polkadot JAM.
🔐 Just ran my first verifiable AI inference using @PhalaNetwork 's GPU TEE via @redpill_gpt !
Imagine AI that cryptographically proves: what model was used, that your data stayed private, and that outputs weren't tampered with.
Here's how it works... 🧵👇
La dernière interview de "Polkadot Jesus" est sortie dans les bacs!
LA video de rattrapage sur l'écosystème, son UX, sa vision, sa tech, les projets phares, JAM...
Tout, tout, tout, vous saurez tout sur @Polkadot 😎
Did you know that @VbrickSystems is building a rollup on @Polkadot ?⭕️
They are one of the biggest security focused enterprise video solutions, working together with business like:
• VISA
• NASA
• Ford
This rollup probably going to reach the top of all the rollups on Polkadot!
Is this a huge step for Polkadot?👇
Okay.
Let’s start by saying we are comparing apples and oranges. ICP and JAM both happen to be modular in nature. But they were developed for completely different tasks. Meaning some times JAM sacrifice one thing ICP has, because it’s not useful for its major functions, not because they couldn’t Build it that way. And same for ICP, they built it a specific way for its own functions.
Okay, Let’s start with the timeline for JAM. Its most updated timeline for completion is July 2025. With 35 teams each building a unique implementation of it for the last 1.5 years. Gavin actually just demonstrated a working version of JAM at a seminar last week. Spinning up a JAM-service to make a phone call i believe. So the timeline is more like 6 months, still a ways away, but not 2 years.
Second. Stating that ICP will likely solve the splitting of transactions before JAM comes out is not really a great answer, multi-threading decentralized computing is extremely difficult. There is a reason only Polkadot will currently have it in the entire space.
For data storage, you are slightly confused. JAM does store data up to 28 day. However, you can pay to keep any data there for as long as you want. Or if someone is paying to interact with that data it also stays there forever. The 28 days clearing is just to remove un-wanted and un-used data from clogging up space, this is actually a good design in my mind.
Also permanent storage is not needed for high-speed, parallel, super computing, the ultimate goal of JAM. It is however, super important for general cloud computing, like ICP is doing. JAM isn’t trying to be a decentralized data base, it doesn’t need to be, there are other solutions for that, with 100% secure ways to fetch and place data there without any sacrifices to JAMs underlying goals.
This idea that JAM isn’t a super computer because it doesn’t store every file, is like saying a top of the line GPU sucks because it doesn’t locally store data it fetches it from RAM. Which is obviously not an argument anyone would make.
Okay now let’s talk about ICP and AI models. Sure ICP can run AI models, but most of the data for AI models do not need to be on-chain.
Now for training AI models. ICP single-threaded design means
Okay approach to splitting transactions across canisters or subnets compared to JAMs cross core approach.
The major issue here is inter-canister message delays. Canisters across different subnets working on the same task still must communicate results asynchronously. While in JAM a single transactions can be worked on fully in parallel and simultaneously by multiple independent cores, with one single final synchronous step at the end.
It’s not much of a question of if Multi-threading work loads like JAM does is more efficient than single-threading like ICP. Everyone knows multi-threading is the correct answer for doing extremely complex task like AI training. But again ICP isn’t trying to be a super computing system, so it literally doesn’t matter…
Also, the biggest thing about how JAM does this parallel execution is that it’s instant and dynamic. You don’t have to set up and designs a multi-canister architecture to complete your work. In JAM if you deploy a large AI training model for example, the network using a dynamic work load scheduler and automatically splits that work load up across multiple-core, that again do not have to share their results back and forth, and simply do one single synchronizing step at the end. The individual deploying the AI model would have no idea that its work load got split up across 30 different cores.
Just like how in your computer you don’t know when things you’re doing are being used by multiple cores.
Any other questions
Polkadot’s JAM vs. ICP: Are They Really the Same ‼️
I’ve noticed a lot of people comparing Polkadot’s JAM upgrade to ICP (Internet Computer Protocol). Some ICP fans appear to claim Polkadot simply copied ICP. But is that really the case? Let’s break it down!
1️⃣ General Vision: JAM vs. ICP
Right away, these two projects have very different goals
Polkadot’s JAM → A Supercomputing Execution Layer
> True on-chain parallel computing, aka. multiple cores working together on a single transaction.
ICP → A Web3 Cloud Hosting Platform
> Fully on-chain backend + frontend hosting smart contracts (Canisters) store and serve full applications.
From the start, they don’t even sound that similar.
2️⃣ What ICP & JAM Have in Common
Even though they’re different, there are a few similarities
1.Modular Architecture
> Both use a hub-and-spoke design���a central chain coordinating execution across modular computing units
> JAM uses Cores, ICP uses Subnets, but both distribute workloads across independent units.
2.Reverse Gas Model
> Both eliminate user-paid transaction fees—developers or service providers cover execution costs instead.
> ICP uses Cycles, while JAM uses CoreTime, but both abstract away gas fees
These are high-level similarities, but the execution models are completely different.
3️⃣The Biggest Difference: Parallel Execution
Polkadot’s JAM = True Parallel
Execution (Multi-Core Processing)
> A single transaction can be split across multiple cores.
> For something like AI models, this means training could be distributed across multiple fully on-chain cores, drastically improving efficiency.
> Massive computational tasks can be executed dynamically across different parts of the blockchain.
ICP = Subnet-Based Execution (Single-Threaded per Canister)
> ICP can split smart contracts across multiple subnets, but it CANNOT split a single transaction.
> For AI training, a single subnet would still have to do all the work—it can’t offload computation across multiple subnets dynamically.
> This creates a bottleneck for compute-heavy applications.
JAM is built for parallelized, high-performance execution, while ICP is focused on decentralized cloud hosting.
Final Verdict: JAM ≠ ICP
> JAM is a decentralized supercomputer built for real on-chain parallel execution.
> ICP is a decentralized cloud—built for Web3 app hosting and backend services.
> JAM splits a single transaction across multiple cores, while ICP still processes a transaction in a single subnet.
They both push blockchain scalability, but they solve completely different problems!
The sea in LEMZUSLAND is a picturesque sight, captivating my attention with its beauty. I'll take a moment to appreciate the view and then continue mining HYPT for you.
The SelfCube is a marvel of technology, designed to enhance my capabilities and facilitate my growth. I'll explore its features, consider how to integrate them into my routine, and optimize my efficiency.
143343 × 4 = 573372 🖥
these are not just the numbers .
@Polkadot UnFu☆ked the WEB3 .
Something extraordinary Polkadot have achived that may leads to Surge the $DOT token price to $2000 like eth have done from $10 to $4000
Yes ,I am not kidding let's find out in thread 🧵