By Lauren Berg ( December 11, 2024, 11:58 PM EST) — A split Fifth Circuit ruled Wednesday that Nasdaq cannot implement U.S. Securities and Exchange Commission-approved rules requiring that companies listed on the exchange disclose board diversity data, finding that the stock exchange’s rules run afoul of federal securities law…. Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Training a large language model (LLM) is among the most costly and time consuming exercises for enterprises. A new open-source model being released today by ServiceNow could make a big difference, with the promise of training 20% faster, saving enterprises time and money. The Fast-LLM technology has already been in development inside the company, helping ServiceNow to accelerate its own LLM training efforts. Fast-LLM helped train ServiceNow’s StarCoder 2 LLM, which the company released earlier this year. StarCoder itself is an open source effort, as well, which benefits from the contributions of Hugging Face, Nvidia and others. ServiceNow also uses Fast-LLM for large, trillion-token continuous pre-training from existing models, as well as for fine-tuning jobs. Because it is an open source technology, anyone can use Fast-LLM to help accelerate AI training, including fine tuning operations. The intent is that it can be a drop-in replacement to an existing AI training pipeline with minimal configuration changes. The new open source project aims to differentiate against commonly used AI training frameworks, including the open-source PyTorch, with a series of innovations for data parallelism and memory management. “When you’re dealing with compute clusters that cost hundreds of millions and training runs that cost millions of dollars, 20% can be a huge saving in terms of both dollars and time and the overall CO2 footprint,” Nicolas Chapados, VP of research at ServiceNow, told VentureBeat. The innovations that enable Fast-LLM to accelerate AI training The AI industry well understands the challenge of training AI more efficiently. VentureBeat Transform 2024 featured a panel that discussed that very issue, detailing options for scaling infrastructure. The Fast-LLM approach isn’t about scaling infrastructure; it’s about optimizing the efficiency of existing training resources. “We carefully looked at all the operations needed to train large language models, especially transformer based large language models,” Chapados explained. “We carefully optimize both the way in which the compute is distributed to the individual cores within the GPU, as well as how the memory is being used by the models themselves.” Fast-LLM’s competitive advantage stems from two primary innovations that help to differentiate it. The first is Fast-LLM’s approach to computation ordering, which defines the order in which computations occur in an AI training run. Chapados explained that Fast-LLM uses a new technique that ServiceNow calls “Breadth-First Pipeline Parallelism.” “This is the fundamental scientific innovation around the way that compute is scheduled, both inside a single GPU and across multiple GPUs,” said Chapados. The second major innovation addresses memory management. In large training operations, memory fragments over time. This means memory becomes broken into pieces over time as training progresses. The fragmentation creates memory inefficiency, preventing training clusters from using all available memory properly. “We’ve been very careful in the way that we design Fast LLM to almost completely eliminate the problem of memory fragmentation when training those large language models,” said Chapados. How enterprises can use Fast-LLM today to accelerate training  The Fast-LLM framework is designed to be accessible while maintaining enterprise-grade capabilities. It functions as a drop-in replacement for PyTorch environments and integrates with existing distributed training setups.  “For any model developer or any researcher, it’s just a simple configuration file that lets you specify all the architectural details that matter,” said Chapados . Running training operations faster has multiple benefits and can allow enterprises to experiment more. “It makes the risk of large training runs smaller,” said Chapados. “It equips users, researchers and model builders with a bit more ambition to train larger runs, because they will not be afraid that it will cost so much anymore.” Looking forward, the expectation is that as an open source project, Fast-LLM will be able to expand faster, benefiting from external contributions. ServiceNow has already been successful with that approach with StarCoder. “Our goal is really to be very, very transparent and responsive to the community contributions in terms of the use of this framework,” said Chapados.” We’re still getting early feedback about what people like, what they are able to do with it and our goal is really to scale this.” source

Tech executives at global organizations manage their public cloud environments by keeping into account many different factors worldwide, from different regulations to country-specific technology and business requirements. There are a few trends that are stable, though, as we found out in our research. Every year, we produce a series of reports that dive into cloud adoption data in different regions. We have published the first three reports in the 2024 series: The State Of Cloud In The US, 2024; The State Of Cloud In Canada, 2024; and The State Of Cloud In Europe, 2024. We will also publish six APAC-focused reports over the coming months that cover Australia and New Zealand, India, China, Japan, Malaysia, and Indonesia. Each report also includes a map of the data centers in these regions from the global and regional cloud vendors. Global Trends Have Their Own Local Peculiarities The importance of cloud, together with the preponderance of global vendors in each local market, has recently cast new focus on digital sovereignty at global organizations in an attempt to reduce dependability on foreign legislations and grant autonomy and business continuity independent of the cloud vendors’ applicable jurisdiction. In 2024, the cloud continues to serve as the heart of major technology innovations, especially amid the age of AI. Cloud vendors are opening new data center regions to support their customers’ demand in the AI and generative AI space. As an example, Microsoft is set to invest €4.3 billion just in Italy to enhance its cloud and AI data center infrastructure. In addition to AI, companies are prioritizing FinOps to get more out of their cloud spend. Enterprises adopt FinOps for collaboration, accountability, and transparency on cloud spend — and this is gaining traction among tech executives, not just cloud leaders. Here is a sampling of our findings from our recent research on cloud adoption: Digital sovereignty norms influence cloud choices in the US, Europe, and Canada in different ways. While enterprises in the US, Europe, and Canada say that they value digital sovereignty, the three markets are moving in different directions. More than half of US enterprises say digital sovereignty “completely” constrains their choice of cloud vendor, which is more than those in Canada and Europe, but the US federal government continues to take a laissez-faire approach to the cloud overall. The European Union and its member states have laws in place that refer to digital sovereignty and serve as models for non-European legislation in cloud and elsewhere. In 2024, the EU and its member states continued to promulgate cloud regulations that touch on areas around and beyond data residency, such as France’s SREN Law, which aims to make the cloud market more competitive. Meanwhile, Canada loosened its data residency laws. For example, British Columbia’s Freedom Of Information And Protection Of Privacy Regulation and Quebec’s Privacy Legislation Modernization Act ignored the question of digital sovereignty. AI is driving cloud adoption and talent hiring in North America and Europe. Enterprises in both North America and Europe are hiring and retraining their staff so they can better utilize AI. For example, some US enterprise cloud decision-makers in 2024 are either hiring new staff (24%) or retraining existing staff (25%) for AI and machine learning (ML). The National Hockey League uses cloud-based AI/ML to track player performance and the likelihood of players scoring goals. Vodafone used Google Cloud to build a unified AI platform, giving its data scientists prebuilt pipelines, standardized tools, and automated deployments to cut the amount of time it takes to move from proof of concept to production-ready AI solutions. This platform allowed Vodafone to reduce its AI development time from proof of concept to production from 20 to two weeks. FinOps is a priority in North America and Europe to optimize costs and cloud usage. Many enterprises are adopting FinOps to optimize their own cloud usage: 68% of North American enterprise cloud decision-makers and 56% of European enterprise cloud decision-makers have their own FinOps practices. Enterprises such as Software Solutions Integrated (SSI) and Siemens Mobility Services use cloud cost management and optimization tools — Virtana Optimize and CloudCheckr, respectively — to optimize their cloud spend. To help you navigate the cloud markets in each of these regions and beyond, here are a few steps that you can take: If you are a Forrester client and want to learn more about this topic, please schedule a guidance session with me, Dario Maisto, senior analyst. Explore the links to related research included below. Not a Forrester client? Please contact the Forrester account team. source

（相）香港旅遊促進會總幹事崔定邦召開記者會，為業界提出訴求。 香港立法會於2023年10月23日通過《酒店房租稅條例》，計劃於2025年1月1日起恢復徵收3%的酒店房租稅。然而，距離正式實施不足一個月，業界對稅項計算方法仍有重大疑問，擔心此措施將引發旅客不滿，並對業界造成嚴重影響。 酒店房間稅徵收混亂  與OTA爭議未解 香港自1966年起徵收酒店房租稅，1998年調整至3%。2008年，因促進旅遊業發展，暫停徵收。然而，隨著政府計劃於2025年恢復此稅項，酒店業面臨多重挑戰： • OTAs的影響：網上旅行社的崛起使得訂房方式愈加複雜，稅項計算面臨新挑戰。 • 多元化套餐：現今的酒店預訂常包含附加服務，難以準確計算純房租部分。 • 資訊混亂：部分OTAs提供錯誤資訊，甚至指示酒店免收稅，與政府政策衝突。 • 缺乏清晰指引：政府尚未提供明確的稅款計算方法，導致業界面臨合規風險。 業界訴求 基於以上挑戰，香港旅遊促進會提出以下建議： 1. 延遲稅項實施：請求政府延遲酒店房租稅的實施，直到問題完全解決，並提供清晰的計算指引。 2. 提供一致政策指引：政府需明確政策內容，確保各方了解責任和義務。 3. 規範OTAs行為：要求OTAs提供準確的房價和稅款計算資訊，避免誤導性資訊。 4. 加強與業界溝通：定期與酒店業界舉行會議，聆聽反饋，確保政策符合實際需求。 5. 提供支援措施：考慮疫情後業界復甦的困難，設立專門小組協助中小型酒店解決稅務問題。 潛在影響 若在未解決問題的情況下倉促實施房租稅，將對香港旅遊業造成深遠影響，包括： • 損害聲譽：旅客因稅項問題而產生不滿，影響未來旅遊復甦。 • 運營困難：增加的合規風險可能迫使部分中小型酒店關閉。 • 經濟影響：混亂或削弱政府振興經濟的努力，影響吸引國際旅客的能力。 酒店業是香港經濟的重要支柱。我們呼籲政府： • 延遲酒店房租稅的實施； • 提供清晰的一致政策指引； • 加強與業界的合作； • 規範OTAs行為； • 提供必要的支援措施。 我們希望與政府攜手合作，確保政策順利推行，並重建外界對香港旅遊業的信心。 LinkedIn Email Facebook Twitter WhatsApp The post 香港旅遊促進會促政府暫緩徵酒店房間稅 appeared first on VeriMedia. source

You’ve got your Voice over Internet Protocol network up and running smoothly — congratulations! One of the great things about a well-set-up VoIP system is its reliability. For the most part, you can count on it to hum along and support your calls, chats, and team meetings. But even the best network setups can face stress. If your business expands to new locations, as more callers start using WiFi instead of traditional phone lines, or deploys new security measures — your previously perfect system may start to have performance problems. If you have tried all the basic VoIP troubleshooting fixes, it might be time to take a closer look at how to optimize your VoIP network. A well-optimized network can make all the difference for professionality. In this post, we’ll walk you through practical steps to ensure your VoIP network is primed and ready for your next call. 1 RingCentral RingEx Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Medium (250-999 Employees), Large (1,000-4,999 Employees), Enterprise (5,000+ Employees) Medium, Large, Enterprise Features Hosted PBX, Managed PBX, Remote User Ability, and more 2 Talkroute Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Any Company Size Any Company Size Features Call Management/Monitoring, Call Routing, Mobile Capabilities, and more 1. Prioritize VoIP traffic on the Network VoIP is super sensitive to one little thing: latency. Now, what exactly is latency? Think of it as the time it takes for your voice to travel from your computer or phone to its destination. Unlike traditional phone calls, where the connection is almost instant, VoIP relies on data packets being sent and received. If there’s even a slight delay in transmission, things can get messy really quickly. Imagine talking to someone and hearing them respond with a weird lag or, even worse, having your words cut off mid-sentence! Not exactly the smooth conversation you were hoping for. So, how do you handle this? VoIP networks have something called Quality of Service (QoS) features that allow companies to prioritize VoIP traffic over everything else on their networks. In other words, they ensure that your VoIP sessions get priority over things like web browsing so that your calls sound crystal clear. Some routers and switches have user-friendly interfaces that allow you to tweak QoS settings on your own. But if you don’t know what that means, your best bet is to check out the manufacturer’s website or user manuals. They often have step-by-step instructions tailored to their specific devices. Learn more about QoS and why organizations should use it. 2. Assess your bandwidth bottlenecks Bandwidth refers to the maximum amount of data that can be transmitted over an internet connection in a given period. In our highway example, we could say it’s the maximum number of cars (data packets) that can fit on the highway going the same speed. Now, think of rush hour. Even with multiple lanes, the highway gets congested if too many cars are trying to use it at once. Likewise, your network can get jammed when too much data has nowhere to go. If you don’t have enough bandwidth, you’ll face a hot mess! For example, a sudden surge in incoming or outgoing calls during peak hours might cause lag times, voice distortion, dropped calls, echoes, or delays. Or, overloading a low-speed DSL connection or a wireless network with too many callers might do similarly. These scenarios are especially common as teams grow or you add new connections. SEE: Learn how to fix dropped calls in just a few minutes. To figure out if bandwidth is your problem, analyze your usage patterns, considering factors like the number of simultaneous calls made, the size of files being transferred during calls, and any other data-intensive activities on your network. If your internet connection constantly struggles, it’s time to upgrade. Higher-speed options like fiber or cable can keep your communications running smoothly, even as your demands grow. 3. Upgrade your router Not all hardware is created equal. What served you well in the beginning, like a mid-range router, may start to buckle under the pressure of higher traffic volume or bandwidth-intensive activities. Equally, your router may start to stutter as your organization grows and more people need more bandwidth to work with. If so, it might be time to upgrade to a newer router built for VoIP communication. Investing in a high-quality router with advanced QoS features is one key to optimizing your VoIP experience. These routers are designed to prioritize voice traffic over other data. This helps prevent call drops, echoes, or other issues caused by your network, ensuring that you can call reliably even during times of heavy network congestion. 4. Implement a jitter buffer While the name may sound funny, adding a jitter buffer can greatly improve call quality. These buffers help regulate the flow of incoming data packets, reducing delays and choppiness to give you smoother calls. Think of them like tiny waiting rooms for data packets. As packets arrive, the buffer stores them temporarily, regardless of their arrival time. Then, it releases them in a steady stream to your device, ensuring a consistent flow of information. Kind of like how passengers gather at the gate and then all board the plane together. To implement a jitter buffer, you need to configure it within your VoIP system. The specific process may vary depending on the software or hardware you’re using. Here’s a general overview of the steps: Determine the appropriate buffer size: The buffer size should be set based on your network conditions and the level of jitter you’re experiencing. Smaller buffer sizes reduce delay but might be less effective in compensating for high jitter levels. Set the buffer delay: This parameter determines the amount of time the packets are stored in the buffer before being played out. Adjust the buffer delay based on the level of jitter you’re experiencing. Configure other settings: Depending on your VoIP system, you may

By Bonnie Eslinger ( December 11, 2024, 11:01 PM EST) — A California federal magistrate judge indicated Wednesday that Google likely won’t be able to escape a proposed class action alleging YouTube illegally collected children’s data from targeted ads, saying a seventh version of the complaint sufficiently alleged that the tech giant “engaged in highly offensive conduct.”… Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

LinkedIn Email Facebook Twitter WhatsApp The post 行業研究報告: 2024年中國消費者信心報告（BCG） appeared first on VeriMedia. source

Traveling for work or working remotely often means juggling priorities and finding ways to stay productive while protecting your most valuable gear. Lugging an expensive desktop or high-end laptop around isn’t always practical—or safe. That’s why it’s so useful to have something like this $75 HP Chromebook 11. This compact computer would normally cost $239, but now you can get a reliable, low-risk machine for under $100. Head to TechRepublic Academy to grab this offer. Under the hood This refurbished Chromebook offers enough power to handle email, web browsing, video calls, and basic productivity tasks. It’s equipped with an Intel Celeron processor and 4GB of RAM, which is modest by modern standards but perfectly adequate for lightweight tasks. The Chrome OS ecosystem enhances its usability, providing seamless integration with Google apps like Docs, Drive, and Gmail, and you can also access Android apps for some added functionality. Durability is key for any travel laptop, and this Chromebook doesn’t disappoint. Originally designed for classroom use, the HP Chromebook 11 is built to endure bumps, spills, and the wear and tear of travel. It has reinforced corners and a spill-resistant keyboard, so you can toss it in your bag without worrying about every little mishap. Its compact size and light weight make it perfect for slipping into a carry-on or backpack without adding unnecessary bulk. You won’t have to search for an outlet often. This Chromebook can last up to 13.5 hours on a single charge, which makes it ideal for long flights or days spent working in coffee shops or coworking spaces. And with fast charging, you can get back up to 90% battery life in just 90 minutes. While it won’t replace a powerful desktop or serve as a multitasking machine for heavy workloads, this HP Chromebook excels as a secondary device you can use on the go. Visit TechRepublic Academy to get this HP Chromebook for only $74.99. Prices and availability are subject to change. source

From brainstorming ideas to booking guests and outlining show topics, there’s a lot of work involved with planning a podcast episode. Podcast planning templates can help you keep track of sponsors, manage your editing and production process, and even map out promotional content once your episodes go live. These are some of my favorite options for beginners, seasoned hosts, and everyone in between. monday.com: A podcast planning template for smooth production monday.com is one of my favorite project management tools on the market. It’s a great way to plan your podcast because it has everything you need to stay organized and ensure none of your production tasks slip through the cracks. It also has a great mobile app, allowing you to manage podcast planning from the palm of your hand. You can sign up for monday.com’s free plan to unlock access to 200+ templates, up to three project boards, eight custom column types, and seats for two collaborators. It’s a great place to start or try it out but most teams will likely need to upgrade to a paid plan rather quickly. More on monday.com: monday.com Review | monday.com vs Wrike | monday.com vs Airtable. Individuals and teams alike can use monday.com to organize ideas and manage podcast productions. Image: monday.com I really like how the production flow is set up with this template. Instead of a single card for each episode that displays its status, every episode has its own list of dedicated tasks. For example, the preparation phase may include tasks related to preparing questions for guests and outlining the episode structure. You can also have a post-production phase for things like editing clips to be promoted on social media before the final episode is polished and ready to live. This kind of setup lets you get as granular as you’d like without cluttering everything up. The template also lets you estimate how much time each task should take. From there, you can enter the actuals and compare the difference. Overall, monday.com and this template are solid options for any podcaster, whether you’re working by yourself or with a full team. It’s easy to use, and the fact that you’re getting the full power of monday.com’s project management platform is a bonus that can’t be overlooked. ClickUp: Podcast planning templates for any type of show ClickUp is another all-in-one project management and team collaboration platform. While this might seem like overkill for podcast planning, there’s a free forever plan that’s perfect for organizing just about everything. Each template below is more advanced than the one before it. The right one for you depends on how much you want to lean on the tool, how many people you’re working with, and how granular you want to be with post-production tracking. More on ClickUp:  ClickUp Review | ClickUp vs monday.com | ClickUp vs Notion. Template 1: A beginner-friendly option for planning episodes This template is great for newer podcast hosts who just need something simple to plan upcoming episodes. It comes with all of the basics you need to stay organized. You can create a backlog of episode ideas, categorize episodes, add audio files, and attach links to live episodes so everything’s centralized in one place. This beginner-friendly template is great for solo podcast hosts who need some help planning and organizing upcoming episodes. Image: Clickup.com My favorite part about this podcast planning template is its simplicity. It’s organized into four sections that follow a linear flow — planned episodes, recorded episodes, shows that have been edited, and shows that have been published. Because of its simplicity, you won’t have to spend a ton of time figuring out how it works. It’s just a matter of signing up and adding all of your ideas. Within minutes you’ll be more organized and have far more insight into where everything stands in your production flow. The biggest downside is that it’s missing functionality for brainstorming, managing post-production tasks, and scaling beyond a single channel. This isn’t necessarily a bad thing, especially for newer shows. But bigger production teams will likely need to go with one of the other templates below. Template 2: Added features for managing different types of podcasts This ClickUp template follows the same core concept of the previous one in the sense that all episodes are organized by stage in the production process. But it takes it up a notch if you want more granular details about each episode. You’ll see the progress status of each episode updated in real-time as you or your team completes tasks required to get it live. This template is a good starting point if you need to involve a team in your production flow. Image: Clickup.com This template comes pre loaded with 17 custom attributes that you can use tag episodes and stay organized. These range from topic categories to listens, subscriber information, and more. You’ll also be able to categorize each item by type. For each type, you can define a set structure that’s applied when you add a new episode, so you don’t have to outline all the steps every single time you add something new. Say you’re planning an episode that includes a guest. If you tag that episode with this attribute, you can have your interview notes and questions available by default on that particular card. I also love the visual elements of this template. If you want to have a broad range of topics for your show, the color-coded topic column clearly indicates whether you have a good mix of ideas or too many episodes in one category. While you can use it to manage some post-production tasks, like your social media marketing plans, it’s a bit more limited for this use compared to the template below. Template 3: The most advanced option for managing all pre and post-production tasks If you have a well-established show with a full team working collaboratively to record, edit, and promote your show, this template has everything you’ll need to stay