Acala Network is set to shut down, with users required to complete asset migration by the end of 2026.
According to official announcements, the Acala community has voted to shut down its network, with block production expected to cease after February 10, 2027. The project team is reminding users to complete asset transfers via the Astar Portal cross-chain bridge by December 31, 2026: users holding aSEED, LDOT, or ACA on the Astar network must move their assets to Acala; users holding ASTR on the Acala network must transfer their assets to Astar. The Astar Portal is expected to remove the Acala cross-chain channel in early January 2027, after which a governance proposal will be passed to shut down the cross-chain connection between Astar and Acala. Any ASTR remaining on the Acala network at that time will be non-withdrawable.
5 minutes ago
AMD completes approximately $2.9 billion acquisition of Taalas: Embedding AI models directly into chips, enabling 17,000 tokens generated per second.
Beating AI News Flash: AMD announces the completion of its acquisition of Canadian AI chip company Taalas. The two sides reached an agreement in August this year, with a total transaction value of up to approximately $2.9 billion, of which up to $1.4 billion will be paid contingent on meeting agreed conditions.
Taalas has taken a completely different approach from traditional GPUs: when manufacturing its chips, it directly hardcodes AI model parameters into the circuits. Conventional GPUs need to repeatedly read model parameters from video memory, while Taalas integrates storage and computing, reducing data movement and eliminating the need for expensive high-bandwidth memory (HBM) and related complex packaging.
In February this year, Taalas launched its first chip, the HC1, designed specifically to run Meta’s Llama 3.1 8B model. According to official tests, with 1000 input and output tokens each, the single-user generation speed reached approximately 17,000 tokens per second, nearly 9 times that of Cerebras at the time. However, this design comes with a tradeoff in flexibility. Conventional GPUs can load different models at any time; while the HC1 supports LoRA fine-tuning, switching base models still requires modifying manufacturing masks and re-producing chips, a process Taalas says takes about two months.
The HC1 also uses aggressive 3-bit and 6-bit quantization, which while significantly compressing model parameters, also results in some loss of model quality. The current 17,000 tokens per second figure only applies to Llama 3.1 8B; the actual performance on larger models remains to be verified.
AMD plans to integrate this technology into its future AI accelerator products, pairing it with its existing Instinct GPUs, so that the dedicated chip handles high-frequency inference tasks for fixed models.
5 minutes ago
Binance Wallet's subscription quota for Anthropic's IPO has hit the 500 million USDC cap.
According to an official announcement, Binance Wallet has announced that its xStocks subscription campaign for Anthropic's IPO has hit the total cap of 500 million USDC, and is no longer accepting new subscription applications. Eligible participants will receive their corresponding tokenized stocks after relevant processes are finalized, with applicable refund amounts also being returned to their accounts. Binance Wallet added that a separate announcement will be issued once relevant details are confirmed.
5 minutes ago
Second Secures $10 Million in Funding: Automatically Records Computer Activity to Enable Memory Sharing Across Different AIs
Beating AI News – AI startup First Person Labs has launched a personal memory tool called Second, and announced it has closed a $10 million seed round led by a16z. Once installed on Mac, Second runs in the background to continuously access the user’s computer activity, automatically organizing ongoing projects, to-do items, contacts, writing habits, and recent actions. With long-term use, it gradually learns the user’s preferences, work patterns, and priorities.
While products like ChatGPT and Claude already have memory capabilities, the information they store is generally not directly shareable. Second decouples this memory layer, making it available to various AI agents via APIs and MCP. For example, when you discuss a project in Claude, switch to Codex for programming, or use Grok Bot to continue a task, you can call the relevant context stored in Second, eliminating the need to repeatedly explain prior work.
By default, these memories are stored in local folders; users can directly view, edit, back up, or migrate them, and control which AI tools can access the data. Second supports integration with existing Claude, ChatGPT subscriptions, and other models, and also offers optional cloud services. It is currently in early access and not yet fully launched. The company also plans to open-source key components to enable self-deployment.
5 minutes ago
Elon Musk: SpaceX is not expected to sell any computing power to OpenAI.
Elon Musk posted on social media that SpaceX has set its future sales target for OpenAI at zero. If OpenAI does not see rapid growth in its own computing power needs, there may be issues with its products. Earlier, OpenAI CEO Sam Altman said chipmakers overestimated OpenAI’s demand, assuming the firm would buy all their chips, but companies like OpenAI and Anthropic have limited needs—including for computing power—and OpenAI cannot possibly require all chips from every chipmaker.
5 minutes ago
Morpho officially deployed on X Layer, Gauntlet’s USDC Vault launches simultaneously.
According to official announcements, on-chain lending protocol Morpho has been officially deployed on X Layer, with the USDC Vault managed by Gauntlet launching simultaneously, further enriching X Layer’s stablecoin yield and lending scenarios. Users can deposit USDC to earn yields, or use tokenized assets including xStocks (tokenized stocks) and xBTC as collateral to borrow funds, unlocking liquidity without selling their assets. Additionally, OKX Wallet has rolled out a limited-time incentive program totaling $350,000, covering deposits, collateralization, and borrowing scenarios, further lowering the threshold for users to participate in on-chain lending and yield activities on X Layer.
5 minutes ago