DB-Engines Releases the April 2024 Database Ranking
This month's DB-Engines Ranking includes 419 products, one more than last month. DB-Engines again showed broad declines this month, with most database products seeing score decreases. Oracle led the overall list with a sharp increase of 13.21 points. PostgreSQL, the database of the year, ranked second with a gain of 10.15 points. The top 10 rankings remained unchanged from last month.
_Source: https://www.modb.pro/db/1775436711576801280_
Tencent Cloud Announces a Comprehensive Cloud Storage Upgrade for AIGC Scenarios
On April 8, Tencent Cloud announced a comprehensive upgrade of its cloud storage solutions for AIGC scenarios, providing comprehensive and efficient cloud storage support across the full lifecycle of AI large model data collection, cleaning, training, inference, and data governance.
According to data, Tencent Cloud's AIGC cloud storage solution can double both data cleaning and training efficiency for large models, cutting the required time in half. Tencent Cloud said this is the first cloud storage solution in China to achieve fully self-developed storage engines.
_Source: https://www.niutoushe.com/lives/tengxunyunxuanbuyuncunchujiejuefanganmianxiangaigcchangjingquanmianshengji_
Baidu AI Cloud Hosts Its First Global Ecosystem Conference
On April 9, Baidu AI Cloud held its first GENERATE Global Ecosystem Conference in Chengdu. At the conference, Baidu AI Cloud joined partners to discuss a new cloud computing ecosystem in the era of large models. It also formulated differentiated ecosystem strategies for three target markets: leading markets, value markets, and high-potential markets, and released a series of partner market benefits. Around the large model technology stack, it also released product benefits and support plans across compute power, models, application development, and application sales.
_Source: https://www.niutoushe.com/lives/baiduzhinengyunjubanshoujiequanqiushengtaidahuimianxiangsanleishichangfabushengtaihuobanquanyi_
Oracle Releases DTrace 2.0.0-1.14
DTrace 2.0.0-1.14 was released in the first week of April. The new version is based on BPF and other Linux kernel tracing capabilities, and is fully implemented as a user-space application. It can be used for tracing on any Linux kernel that provides BPF-based tracing and BTF type data.
_Source: https://www.oschina.net/news/286806_
IDC FutureScape Releases 2024 Global IT Industry Predictions
The report provides predictions for the future development of the global IT industry in 2024, focusing on the widespread application of artificial intelligence (AI) and its impact on enterprise technology decisions and digital business plans.
_Source: https://www.oschina.net/news/287404_
Adobe Buys Video Assets at USD 2.62 per Minute to Build AI Models
News on April 11: Adobe has begun working to obtain videos to build its own AI text-to-video model, competing with OpenAI's Sora. According to reports, Adobe asked its photographer and artist communities to provide more than 100 short videos showing human actions and emotional expressions, as well as simple anatomical images of feet, hands, or eyes. The video fee averages about USD 2.62 per minute, but may be as high as about USD 7.25 per minute.
_Source: https://www.niutoushe.com/lives/adobeyi262meiyuanfenzhonggoumaishipinziyuanlaigoujianaimoxing_
Elon Musk's xAI Showcases Its First Multimodal Model, Grok-1.5V
News on April 15: After launching the Grok-1.5 large language model in late March, Elon Musk's AI company xAI recently released its first multimodal model, Grok-1.5 Vision. xAI said it will soon invite early testers and existing Grok users to test Grok-1.5 Vision (Grok-1.5V), which can understand not only text but also content in documents, charts, screenshots, and photos.
_Source: https://www.niutoushe.com/lives/masikexaizhanshishougeduomotaimoxinggrok15v_
Google Releases the CodeGemma Code Large Model
Google released CodeGemma, a large code model for programming assistance. CodeGemma is based on Gemma, the lightweight open-source large model Google released in February. For the two Gemma parameter sizes, Gemma 2B and Gemma 7B, Google released CodeGemma 2B and CodeGemma 7B, with both pretrained and instruction-tuned versions.
_Source: https://www.oschina.net/news/286993/google-codegemma_
Intel Releases the Gaudi 3 AI Chip
Intel released its latest AI chip, Gaudi 3, which is expected to become widely available in the third quarter. Companies including Dell, HP, and Supermicro will use the chip. Intel did not provide a price range for Gaudi 3. The company claims the new Gaudi 3 chip delivers, on average, 50% better inference capability and 40% better energy efficiency than NVIDIA's H100 chip, and runs AI models 1.5 times faster than H100. It is expected to be roughly comparable to NVIDIA's latest H200, and may perform better in some areas.
_Source: https://www.oschina.net/news/286960_