Qualcomm introduced the Snapdragon C processor, a new chip aimed at budget laptops. The announcement positions the chip as a cost-effective solution for entry-level portable computers. The first laptop to feature the Snapdragon C is the Acer Aspire Go 15, which targets users seeking basic computing performance at a lower price point.

The Snapdragon C is built on an efficient architecture, though Qualcomm did not disclose specific manufacturing process details. The chip integrates a CPU, GPU, and AI engine, enabling tasks such as web browsing, document editing, and video playback. Connectivity options include Wi-Fi and Bluetooth, with support for 4G LTE in select models.

Acer’s Aspire Go 15 comes with a 15.6-inch display, 4GB or 8GB of RAM, and storage options up to 256GB. The laptop runs Windows 11 and is designed for students and casual users. Pricing starts at $299, making it one of the more affordable Windows laptops on the market.

Qualcomm emphasized the chip’s power efficiency, claiming it can deliver all-day battery life. The Snapdragon C also supports fast charging, allowing users to quickly top up the battery. The processor is part of Qualcomm’s broader strategy to expand beyond smartphones into the PC market.

The Acer Aspire Go 15 with Snapdragon C will be available in select regions starting in April 2025. Qualcomm stated that additional laptop manufacturers are expected to adopt the chip in the coming months. The company aims to compete with Intel and AMD in the low-cost segment.

Pricing for the Snapdragon C itself was not disclosed, but Qualcomm indicated that laptops using the chip would typically cost between $250 and $400. The chip supports up to 8GB of LPDDR5 RAM and eMMC or UFS storage. Display output is limited to 1080p resolution.

Qualcomm’s move into budget PCs follows its earlier Snapdragon 8cx and 7c series, which targeted higher-end and mid-range devices. The Snapdragon C represents a further push into the entry-level market, where ARM-based processors have yet to gain significant traction. The chip’s success will depend on consumer adoption and software compatibility.

The Acer Aspire Go 15 is now listed on Acer’s website with a starting price of $299. Shipments are expected to begin in April 2025. Qualcomm confirmed that the Snapdragon C will also power devices from other OEMs later this year.

Asana announced the acquisition of StackAI, a startup that provides a no-code platform for building AI agents. The deal was disclosed on Tuesday, though financial terms were not revealed.

StackAI's platform allows users to create AI agents without writing code, enabling automation of complex tasks. The technology will be folded into Asana's existing suite of AI-powered workflow tools.

Asana has been expanding its AI capabilities over the past year. The company launched Asana Intelligence in 2023, which includes features like smart suggestions and automated task assignments.

The acquisition of StackAI is expected to accelerate Asana's development of more sophisticated AI agents. These agents could handle multi-step workflows and integrate with other business applications.

Asana CEO Dustin Moskovitz said in a statement that the acquisition will help the company deliver on its vision of an AI-powered work management platform. He noted that StackAI's team will join Asana.

StackAI was founded in 2022 and had raised $2.5 million in seed funding. Its platform was used by companies to build custom AI assistants for customer support, data analysis, and other tasks.

The deal is expected to close in the coming weeks, subject to customary closing conditions. Asana did not provide a specific timeline for when StackAI's features would be integrated into its products.

Asana shares rose slightly in after-hours trading following the announcement. The company reported $171.5 million in revenue in its most recent quarter, up 14% year-over-year.

Researchers at the Swiss Federal Institute of Technology Zurich (ETH Zurich) announced they have successfully generated what they describe as "perfect" random numbers for the first time. The team leveraged principles of quantum physics to create a system that produces truly unpredictable sequences, overcoming limitations of classical random number generators.

Classical random number generators rely on deterministic algorithms or physical processes that can be predicted or replicated under certain conditions. In contrast, the quantum method exploits the inherent randomness of quantum measurements, such as the behavior of photons, to produce numbers that are fundamentally unpredictable.

The ETH Zurich team designed an experimental setup where laser pulses are split and their quantum states measured. The outcomes of these measurements are inherently random, as dictated by the laws of quantum mechanics. By carefully controlling the experiment and eliminating sources of bias, the researchers achieved a level of randomness that meets rigorous mathematical standards.

To verify the quality of the randomness, the scientists applied a battery of statistical tests. The generated numbers passed all tests for unpredictability and lack of pattern, confirming their status as "perfect" random numbers. This marks a milestone in the field of quantum random number generation.

The implications of this achievement are far-reaching. Secure encryption systems rely on random numbers for generating keys; if those numbers are predictable, the security can be compromised. Perfect random numbers could bolster cybersecurity, protect financial transactions, and enhance privacy in digital communications.

Beyond cryptography, the technology could benefit scientific simulations that require high-quality randomness, such as in Monte Carlo methods used in physics and finance. It may also advance quantum computing and quantum communication systems, where randomness is a critical resource.

The research was published in the journal Physical Review Letters. The team emphasized that while the current setup is laboratory-based, it demonstrates the feasibility of generating perfect randomness on demand. Further work is needed to miniaturize and speed up the process for practical applications.

ETH Zurich plans to continue refining the technique, aiming to increase the generation rate and integrate it into compact devices. The researchers noted that commercial quantum random number generators already exist, but none have achieved the level of perfection demonstrated in this study.

"This is a fundamental step forward," said a lead researcher. "We have shown that perfect randomness is not just a theoretical concept but can be realized in practice." The team expects that their method will pave the way for next-generation security systems built on unbreakable cryptographic protocols.

Major cloud infrastructure companies are rethinking their network architectures as artificial intelligence agents begin to generate a significant share of internet traffic. AWS and Cloudflare, among others, have announced updates to their platforms aimed at accommodating the unique demands of machine-driven data flows. The changes reflect a broader industry recognition that the internet is being rebuilt for automated systems rather than human browsing.

Cloudflare recently introduced a new service called AI Gateway, designed to manage and optimize traffic from AI applications. The service provides caching, rate limiting, and analytics specifically for requests made by AI models and agents. AWS, meanwhile, has been updating its networking stack to reduce latency for machine-to-machine communications, which often require rapid, high-frequency interactions.

The shift is driven by the growing deployment of AI agents in production environments. These agents, which can perform tasks like customer service, data analysis, and content generation, generate traffic patterns that differ from human users. They tend to make more frequent, smaller requests and require consistent, low-latency responses. Traditional cloud infrastructure, optimized for human web browsing, is not always suited to these patterns.

Cloud providers are also focusing on security implications. Machine-generated traffic can be harder to distinguish from malicious bots, leading to new challenges in fraud detection and access control. Cloudflare has introduced bot management features that can differentiate between benign AI agents and harmful automated scripts. AWS offers similar capabilities through its WAF and Shield services.

Another area of investment is edge computing. By processing AI agent requests closer to the user or device, providers can reduce latency and bandwidth costs. Cloudflare's Workers platform allows developers to run code at the edge, enabling real-time responses for AI agents. AWS's Lambda@Edge provides similar functionality.

The redesign extends to pricing models as well. Some providers are introducing usage-based pricing that accounts for the high volume of small requests typical of AI agents. This contrasts with traditional models that charge per gigabyte of data transferred, which may not reflect the actual cost of processing many small transactions.

Industry analysts note that this transformation is still in its early stages. Most internet traffic today remains human-generated, but the proportion of machine traffic is expected to grow rapidly as AI adoption increases. Cloud providers are positioning themselves to capture this emerging market by offering specialized tools and infrastructure.

AWS and Cloudflare have both emphasized that their updates are backward-compatible, ensuring that existing human-centric applications continue to function. However, the long-term trend points toward a bifurcated internet where infrastructure is optimized for both human and machine users. The companies are investing heavily in research to anticipate future traffic patterns.

"The internet is being rebuilt for machines," said a Cloudflare executive in a statement. The company plans to roll out additional AI-specific features over the next year. AWS has not made a similar public statement but has been quietly updating its networking documentation to include best practices for AI agent workloads.