Skip to main content
September 5, 2024

Summer news in the Tech sector that you may have missed

September has already begun and with it, we leave behind the hot summer days, the beach, and the quiet of the holidays. But, as technology does not stop, these months of July and August have been times when new products have been launched and we have discovered advances in various services in the sector. Here are some of the most important ones!

Updates in Microsoft Fabric

Fabric has already deployed AI skills in public preview, allowing users to develop a customized GenAI system tailored to the organization’s needs.

These AI skills enable the creation of conversational question-and-answer systems that can respond to data-driven questions on a case-by-case basis. By providing clear instructions and examples, accurate answers can be ensured, while maintaining robust data security and making insights more accessible to all teams.

They have also introduced Lakehouse charts, a new feature designed to help users organize and manage data more efficiently within logical data spaces. This feature allows tables to be grouped by categories, such as business domains, projects, or teams. They also improve security by allowing policies to be applied to entire groups of tables.

Intel Flex AI capabilities

Intel has made a significant leap in the AI chip market following the release of technical papers showcasing the capabilities of a broad range of processors designed to serve the entire AI ecosystem.

From data centers to edge computing to consumer devices, Intel’s latest offerings demonstrate its commitment to delivering flexible and powerful AI solutions across the board.

With data centers and cloud environments being the backbone of AI, Intel has targeted this sector by introducing its new Xeon 6 processors. These chips are designed to handle the intense computational demands of AI workloads, delivering improved performance and power efficiency.

In fact, one of our colleagues at Plain Concepts recently tested the Intel Arc A770 GPU on Windows to see how it performed with respect to its AI capabilities.

ChatGPT Leadership

Since its launch and the subsequent massive advance of generative AI, ChatGPT became the pioneer of this great revolution. After months, it still retains the title of ‘king of platforms’ and now boasts more than 200 million weekly active users, which has doubled since autumn 2023.

One of the reasons for these figures is its growing penetration in the corporate world. OpenAI reports that a whopping 92% of Fortune 500 companies now use its products, indicating a sea change in the way companies approach AI integration.

The company continues to innovate continuously, and one of the latest developments is the introduction of GPT-4o mini, a more efficient version of OpenAI’s most powerful model. This version aligns with the growing industry trend towards creating smaller, optimized language models without compromising performance.

In addition, OpenAI recently unveiled an early prototype of SearchGPT, an AI-powered search function that could potentially rival global search engines.

Google and more powerful AI-enabled GPUs

Google announced a major update to its Cloud Run service, introducing support for NVIDIA L4 GPUs, a major improvement for users of its AI offerings.

This enhancement is designed to make advanced AI capabilities more accessible to various applications, transforming the functionality of everyday software.

This enables substantial computational power without the burden of maintaining a constant and costly GPU infrastructure, which is crucial to enabling the wider adoption of more powerful AI systems. In addition, the service’s ability to scale down to zero during periods of inactivity ensures that users are not charged when the service is not in use, providing significant cost savings.

Cerebras vs Nvidia

AI hardware startup Cerebras has created an artificial intelligence inference solution that could become a major rival to Nvidia’s enterprise GPUs.

Cerebras Inference is based on the company’s Wafer-Scale engine and promises to deliver astonishing performance. The tool has achieved speeds of 1,800 tokens per second for Llama 3.1 8B and 450 tokens per second for Llama 3.1 70B. Cerebras claims that these speeds are not only faster than the usual hyper-scale cloud products that require Nvidia GPUs to generate these systems but are also more cost-effective.

However, despite the potential performance advantages, Cerebras faces significant challenges in the enterprise market, such as Nvidia’s large offering. The choice between Nvidia and alternatives such as Cerebras depends on several factors, including scale of operations and available capital. Smaller companies are likely to choose Nvidia, as it offers established solutions. At the same time, larger companies with more capital may opt for the latter to increase efficiency and save costs.

As the AI hardware market continues to evolve, Cerebras will also face competition from specialized cloud providers, hyper scalers such as Microsoft, AWS, and Google, and dedicated inference providers such as Groq. The balance between performance, cost, and ease of deployment will likely determine companies’ decisions to adopt new inference technologies.

Microsoft Designer

Microsoft’s app is now available to the general public, with features that help you create and edit in a way you’ve never seen before. With the help of AI, it promises to allow you to express yourself and bring creative ideas to life in seconds.

Just describe what you want to see and Designer can create it for you. It is available through Copilot in the most popular Microsoft 365 apps on the web and PC.

It features prompt templates to help drive the creative process, create custom stickers, images, invitations, change styles, generate collages, replace backgrounds, etc.

Imagen 3 of Google DeepMind

Google has launched Image 3, its highest-quality text-to-image model, capable of generating images with better detail, richer lighting, and fewer artifacts, making them almost impossible to distinguish from a real photograph.

This model helps generate a wide range of visual styles and capture small details from longer prompts. To be even more useful, it will be available in multiple versions, each optimized for different types of tasks, from generating quick sketches to high-resolution images.

Advanced tools such as inpainting and outpainting, along with style transfer options, allow for limitless creativity. Designers and creatives now have more control over the final product than ever before.

 

And so much for the most important launches and novelties of these summer months. If you don’t want to miss anything from the sector, subscribe to our newsletter!

Elena Canorea
Author
Elena Canorea
Communications Lead