Ismael de Esteban Fernández

Today we are announcing an integration with Anyscale. Anyscale is another great AI platform that hosts open weight LLM models such as LLama and Mistral.

5

Testing new AI models with #ollama in a small laptop. https://lnkd.in/deNkm9xT results are pretty different, lots of testing has to be done before adopting one of them.

1

Excited to be hosting a Kafka meetup in Madrid on October 1st where seasoned engineers will share their Kafka horror stories and lessons learned from when things went wrong. Whether you're a Kafka enthusiast or just curious about streaming systems, this event is perfect for learning how to manage failures and optimize your data infrastructure. 🕒 Event Details: Date: October 1st Time: 6:30-9:00 PM Location: Celonis Office, Pl. de Manuel Gómez-Moreno, 2, Tetuán, 28020 Madrid, Spain Talks: 1. How to protect yourself from Kafka Connector storm? Speaker: Rafael Garcia, Platform Engineer at Adidas It’s 3pm on a Friday, and Kafka clients start dropping off while the infrastructure appears stable. How do you analyze and tackle this? Topics covered include root cause analysis, blast radius, key metrics, and remediations. 2. Taming silos and stale data with Streaming Data Lakes Speakers: Francisco Morillo (AWS) & Adam Noonan (Lenses.io) Learn how to solve data silo and stale data issues using Kafka, Flink, and Iceberg to ensure your data is fresh and actionable for real-time analytics. Register in the comments below! #Kafka #StreamingData #DataLakes

15

2 comentarios

Embedded engineers Nathalia Rodríguez and Francisco Javier Molinero Chasco tackled the world of embedded AI and how it will reshape the future. Companies like Inventive are already investing in this tech, while others argue that AI isn't as intelligent as it seems. Even more importantly, a few companies already provide AI-powered cybersecurity solutions, like Palo Alto Networks Trellix, and Zscaler. What do you think about this? Is AI changing the world as we know it, or is it not as relevant as it seems? Full episode on Youtube: https://lnkd.in/dyEiY4e7 Check out our website for more news about the embedded systems industry: https://techteamz.io/ #embeddedai #embeddedsystems #techpodcast #techteamztalks #techteamzpodcast #debate #ainews #artificialintelligence

23

2 comentarios

The first RISC Zero proof has been verified on Starknet testnet! Using RISC Zero's zkVM, developers can easily leverage ZKPs in Rust, unlocking: ✅ Verifiable off-chain execution ✅ Enhanced developer expressivity

24

🚀Embark on a journey into the cutting-edge realm of embedded development with Rust! 💡 Our experts delve deep into how this modern language is reshaping the landscape. Discover: 1️⃣ The essence of embedded development. 2️⃣ Unveiling the concept of legacy technology. 3️⃣ Is Rust challenging the dominance of legacy technologies in embedded systems? Join us at Sencury as we explore the dawn of a new era in embedded development. Dive in, read 📰, rate ⭐️, and share your thoughts ✏️! https://lnkd.in/dsqpeQkG #RustEmbedded #EmbeddedSystems #ProgrammingLanguages #SafetyFirst #Concurrency #PerformanceDriven #ModernDevelopment #CodeReliability #EmbeddedTech #InnovationInCoding #FutureOfEmbedded #TechEvolution #RustLang #CodingCommunity #SoftwareDevelopment #EmbeddedWorld #EfficientCoding #SecurityInCoding #IoTDevelopment #NextGenEmbedded #C++

1

New job opening in Flow. https://lnkd.in/d9v__9HM #LLVM #FlowComputing #PPU #CPUTechnology #CPUtech #ParallelProcessing #SuperCPU #Computing #Seriousperformance #Deeptech

3

How do CPU caches work? CPU caches are a critical component in modern computer architectures, designed to reduce the time it takes for the CPU to access data from main memory (RAM). The key ideas behind caches are locality and reducing memory access latency. 👉 What Are CPU Caches? A CPU cache is a smaller, faster type of memory located on or near the CPU. Its purpose is to store frequently accessed or recently used data and instructions, allowing the CPU to retrieve them more quickly than if they were in the main memory. Caches are generally organized in multiple levels, with each level providing different benefits. 👉How Do CPU Caches Work? CPU caches operate by storing copies of data from main memory. When the CPU needs to access data, it first checks whether the data is in the cache. This is called a "cache hit." If the data is in the cache, it can be accessed quickly. If the data is not in the cache (a "cache miss"), the CPU must fetch it from main memory, which is much slower. 👉Cache Hierarchy 1️⃣ L1 Cache: The fastest and smallest cache, usually split into separate instruction and data caches. It's located closest to the CPU cores. 2️⃣ L2 Cache: Larger and slightly slower than L1, often shared among multiple cores. 3️⃣ L3 Cache: The largest and slowest of the three, typically shared across all cores in a CPU. 👉Cache Lines and Blocks Caches work with fixed-size blocks of data, known as cache lines. When a cache miss occurs, an entire cache line is loaded from memory, not just the specific byte or word that the CPU needs. This helps optimize data retrieval by taking advantage of spatial locality. 👉Types of Locality 1️⃣ Spatial Locality: If one memory address is accessed, nearby addresses are likely to be accessed soon. This is why cache lines are typically 64 bytes or larger. 2️⃣ Temporal Locality: If a memory address is accessed, it's likely to be accessed again soon. This justifies keeping frequently used data in the cache. 👉Cache Replacement Policies When a cache is full, new data replacing old data requires a "replacement policy." Common policies include: 1️⃣ Least Recently Used (LRU): Evicts the least recently used cache line. Random: Randomly evicts a cache line. 2️⃣ First-In, First-Out (FIFO): Removes the oldest cache line. 👉Why Are CPU Caches Important? Caches are vital because they significantly improve CPU performance by reducing memory access latency. Without caches, every memory access would require fetching data from much slower main memory, creating a bottleneck. 👉Caches help to: 1️⃣ Improve Speed: Provide faster data access compared to main memory. 2️⃣ Enhance Throughput: Reduce CPU idle time while waiting for data. 3️⃣ Minimize Power Consumption: Faster data access reduces CPU idle time, conserving energy. Credit : Gopal Chakraborty #cpu #systemdesign #performance #embedded #bsp #arm #embeddedc #embeddedsystem #learning #linux #kernel

71

1 comentario

RISC Zero is coming to Base Our Universal Verifier will deploy on Base, enabling devs to compute without limits. This integration gives Base the power of verifiable off-chain execution - converting real-world apps into on-chain dApps. The future is verified. 🟡🔵

8

Learn how to simplify camera calibration with real and synthetic data to enhance the accuracy of your AI-powered multi-camera tracking system. 📷 https://nvda.ws/3MuH0r1

280

4 comentarios

"Leveraging HPC Infrastructure for #AI Workloads – The Role of #Kubernetes in AI and HPC" by Jordi Blasco Pallares (CTO / HPCNow! – Do IT Now Alliance) 🔴 Watch the video here: ⏪▶️⏩ https://lnkd.in/gWbuNRQW Abstract: As the demand for AI workloads surges, there’s a pressing need to adapt existing High-Performance Computing infrastructures to accommodate the requirements of this new emerging user community. While many HPC users already run AI workloads, the influx of new AI users with little exposure to HPC, demands a more accessible and interactive platform. While Kubernetes is emerging as a solution for AI workloads, significant challenges arise as soon as the AI workloads need to scale. Addressing the needs of this growing user base goes beyond standard Kubernetes solutions. This talk delves into the challenges and solutions to integrating AI workloads with HPC infrastructures. It discusses leveraging existing HPC solutions, such as Warewulf for provisioning Kubernetes alongside Slurm clusters, enabling Slurm to balance the resources effectively for both Kubernetes and traditional HPC workloads based on the load. This talk will also introduce the work done by HPCNow! to enhance performance and efficiency for AI workloads on Kubernetes. Looking towards the future, the talk also examines the development of more mature and suitable solutions for AI and HPC on Kubernetes. #HPCKP24 #HPCKP #HPC

35

🌎 Automate Your Bilingual Resume Creation Just published a repo that generates English & Spanish resumes using GitHub Actions - no more Overleaf issues! (Ps. The resume is made with LaTeX) https://lnkd.in/ekzDfqxd #GitHubActions #MultilingualResumes #Automation

6

Are you maximizing the potential of your #embedded systems? In our latest blog post, we explore the strategic integration of Rust with C++. From handling complex operations to improving memory safety, discover how combining both languages can lead to more robust and secure systems. Dive into the world of Rust and C++ integration for a deeper understanding of their potential in embedded development: https://lnkd.in/eCdS5h64 #RustLang #SecureCoding Ferrous Systems GmbH

8

This is excellent and indeed good news. A welcome sight in our current age 🥳 The Linux kernel’s achievement of hard real-time capabilities is a significant milestone for critical control systems which sit behind grid modernisation. One of the key advantages here is the use of open-source technology, which brings immense value to critical infrastructure. Open-source solutions like Linux RTOS help avoid reliance on monolithic, proprietary vendor ecosystems, instead creating public assets that foster innovation, competition, and choice. This is especially important in grid modernisation, an expensive yet essential process for improving resilience and enabling energy diversification. Grid control systems are dominated by a small group of vendors offering either proprietary or pseudo-open solutions. Linux RTOS introduces new opportunities to reduce costs and drive innovation, making the process more accessible and sustainable. Modernising the grid and diversifying energy generation is vital for reducing energy costs, decreasing reliance on fossil fuels, and protecting against global economic shocks. What might this mean for digital buildings? Microgeneration is another piece in the puzzle of diversification, however, it’s full capability won’t be realised unless the generated energy can actually be used where it is needed. Other digital buildings systems with real time control are load shedding and power management systems. These systems are currently rather expensive, and is usually value engineered. There isn’t really any competition on the market, as per the grid control systems, so, Linux RT should drive more innovation and bring down cost. While this new capability won’t single-handedly solve all these challenges, it represents an important yet unseen-to-the- untrained-eye piece of the puzzle — it’ll enable more effective solutions to emerge in this complex transformation. A step in the right direction, one which made do a little tap dance 🕺 #DigitalTransformation

3

"Leveraging #HPC Infrastructure for #AI Workloads – The Role of #Kubernetes in AI and HPC" by Jordi Blasco Pallares (CTO / HPCNow! – Do IT Now Alliance) Abstract: As the demand for AI workloads surges, there’s a pressing need to adapt existing High-Performance Computing infrastructures to accommodate the requirements of this new emerging user community. While many HPC users already run AI workloads, the influx of new AI users with little exposure to HPC, demands a more accessible and interactive platform. While Kubernetes is emerging as a solution for AI workloads, significant challenges arise as soon as the AI workloads need to scale. Addressing the needs of this growing user base goes beyond standard Kubernetes solutions. This talk delves into the challenges and solutions to integrating AI workloads with HPC infrastructures. It discusses leveraging existing HPC solutions, such as Warewulf for provisioning Kubernetes alongside Slurm clusters, enabling Slurm to balance the resources effectively for both Kubernetes and traditional HPC workloads based on the load. This talk will also introduce the work done by HPCNow! to enhance performance and efficiency for AI workloads on Kubernetes. Looking towards the future, the talk also examines the development of more mature and suitable solutions for AI and HPC on Kubernetes. #HPCKP24 #HPCKP

30

1 comentario

🚀 How long will it take for AI to code independently? Let's see it at work on one of the most crucial and difficult problems in Operating Systems design: the scheduler! 🖥️ Before we dive into how AI generates and optimizes scheduler code, let's have a look at the exciting new Linux developments that allowed such a test. ✨ Andrea Righi is doing amazing and fun things with sched_ext, sched_scx, and scx_rustland. With this new Linux kernel framework, developers can experiment with custom scheduling policies without requiring a system reboot or risking kernel crashes. 💡 This is possible because sched_scx uses BPF (Berkeley Packet Filter), an isolated virtual machine inside the kernel that safely executes custom code. Even if a BPF-based scheduler misbehaves, it won’t crash the system, allowing developers to make dynamic, real-time adjustments to scheduling policies without any downtime! 🔄 And here’s where AI comes in. Andrea ran an experiment where he used ChatGPT-4 to help generate and improve scheduling policies written in Rust. 🧠 The AI worked on a simple FIFO scheduler, and through iterations, it optimized the code for specific workloads, reducing execution times and enhancing performance. Though AI isn’t ready to fully replace human developers just yet, it’s a powerful tool for lowering the barrier to entry in kernel development and helping with code refinement! 🛠️✨ Check out more details from Andrea’s experiment in his blog! #AI #LinuxKernel #Scheduling

10

4 comentarios

Effortless Integration for Smarter Insights: CloudEagle.ai Meets Bitbucket Integrating CloudEagle.ai with Bitbucket takes just minutes and gives you: ✔ Simplified App Tracking: No more manual reports, get the app usage data you need, when you need it. ✔ Centralized Control: Manage your Bitbucket security from the CloudEagle dashboard. ✔ Proactive Monitoring: Catch unusual login activity before it becomes a problem. Ready for a smarter, more secure development workflow? Book a demo today - https://lnkd.in/gXAhChrt #saas #saasmanagement #toolstuesday Atlassian #jira #toolstuesday

5

Learn about the latest W&B SDK improvements, including over 80% faster performance when logging experiments in parallel, higher reliability tracking large-scale experiments, and greater memory and compute usage efficiency. Get the complete details along with benchmarks in our latest blog: https://lnkd.in/gjzxj6P3

10

1 comentario

Have you already checked out Ara2? It's the first fully open-source vector processor supporting the RISC-V "V" 1.0 ISA. Designed to target data-parallel workloads, Ara2 brings cutting-edge  performance and energy efficiency to the table. In this paper, we present a novel open-source vector architecture  and  characterize its performance and power efficiency on a broad set of applications and workload conditions on single- and multi-core configurations.  Here are some key highlights from our work: Þ   Scalable Design: Ara2 supports the RISC-V "V" 1.0 ISA and comes  in different configurations, from 2 to 16 parallel processing units. Þ   Streamlined Microarchitecture: We show that Ara2 can be  physically implemented in all its configurations, thanks to ad-hoc  microarchitectural optimizations to its interconnects. Þ   Thorough Characterization: We study Ara2's performance on  multiple  kernels from machine learning, DSP, and linear algebra for  different  problem sizes. Ara2 reaches peaks of more than 97% of its maximum performance on real workloads. Þ   Single- and Multi-Core Solution: We study the performance/power trade-off of vector architectures of variable processing capabilities when running applications on single- and multi-core architectures.  Check out our paper here https://lnkd.in/e8t9FBpU  for a technical deep dive and more insights!

22