🤖 Tesla demonstrates Optimus Gen 2

Tesla has unveiled the Optimus Gen 2 humanoid robot, showcasing improved humanlike dexterity and mobility. The robot features 11 degrees of freedom in its hands with tactile sensors in each finger, a 2-DoF neck, in-house-designed actuators and sensors, and integrated electronics. The enhancements include a 30% increase in walking speed, force/torque sensing, articulated toe sections in its feet, and a 10 kg weight reduction for improved energy efficiency, cost, and safety.

The company plans to initially deploy the robot in its automotive and battery factories. However, amidst the positive news, Tesla is also recalling over 2 million vehicles due to concerns surrounding its Autopilot system. The National Highway Traffic Safety Administration (NHTSA) is investigating safety issues related to Autopilot, and critics argue that Tesla didn't adequately communicate the risks.

Despite the recall, Tesla CEO Elon Musk has released an impressive video of the Gen 2 Optimus robot, emphasizing its capabilities and addressing skepticism. Musk believes the Optimus business could eventually surpass the value of Tesla's car business and full self-driving technology.

⛓ More information...

#news #robots

📼 Andrej Karpathy’s Baby Llama Runs on Samsung Galaxy Watch 4

Andrej Karpathy, from OpenAI, developed Baby Llama as a weekend project to demonstrate the possibility of running the Llama 2 model on edge devices. The project, inspired by Georgi Gerganov's llama.cpp, involves training the Llama 2 model using PyTorch and then saving the weights in a binary file. Karpathy created a compact C file ('run.c') for loading and running the model with single-precision calculations, ensuring a low-memory footprint and independence from external libraries.

The model runs efficiently even on a single M1 laptop without GPUs, with performance enhancements achieved through various compilation flags for code optimization. A Twitter user successfully demonstrated Baby Llama running on a Samsung Galaxy Watch 4.

Karpathy's repository provides the pre-trained model and code, allowing anyone to experiment with Baby Llama on their devices. It's important to note that this project is experimental and not meant for production use, aiming to showcase the viability of running advanced models on low-powered devices with C, a language typically not associated with machine learning due to its GPU incompatibility.

⛓ More information...

#news #artificial_intelligence

⛓ 3 trends for 2024: edge intelligence, RISC-V, & chiplets

This article discusses three major technology trends expected to shape 2024:

➡️ Edge Intelligence Enhancement: The article highlights advancements in deploying large language models (LLMs) on devices with limited memory, particularly focusing on a research paper by Apple. This paper introduces techniques for efficient LLMs operation on such devices, indicating a broader trend of increased machine learning and inference capabilities at the edge of networks.

➡️ RISC-V Architecture Adoption: The formation of Quintauris, a company backed by major tech players, is set to accelerate the commercialization of RISC-V hardware globally. This development suggests a growing interest in RISC-V architecture, especially in automotive, mobile, and IoT applications. The industry is moving towards heterogeneous computing architectures, where RISC-V plays a crucial role in various functions.

➡️ Rise of Chiplets: Chiplets are emerging as a solution to the high demands of machine learning-intensive products. The trend is compared to the evolution of the intellectual property business two decades ago. Startups like Zero ASIC are democratizing chip-making with platforms for chiplet-based design and emulation. The increasing use of chiplets will rely on standardization efforts like the UCIe specification.

⛓ More information...

#news #artificial_intelligence

💻 Sipeed Longan Pi3H – A Raspberry Pi Zero-sized board with gigabit Ethernet, WiFi 6, HDMI, and USB ports

This small but mighty single board computer is powered by a high-performance quad-core processor and offers various connectivity options, including full-size HDMI 2.0 and gigabit Ethernet. In addition to that, the board includes a unique design featuring a carrier board and a replaceable LM3H CPU module containing the SoC, RAM, and eMMC flash.

Key Specifications:
💻 Processor: Allwinner H618 Quad-core Cortex-A53 up to 1.5GHz
💻 GPU: Arm Mali-G31 MP2
💻 System Memory: 2GB or 4GB LPDDR4
💻 Storage: MicroSD slot, optional 32GB eMMC flash
💻 Video & Audio Output: HDMI 2.0a up to 4Kp60
💻 Networking: Gigabit Ethernet, Dual-band Wi-Fi 6, Bluetooth 5.4
💻 USB Ports: 2x USB 2.0 Type-A, 1x USB 2.0 OTG Type-C

Sipeed provides Debian and Android 12 images for the board, with SDKs available on GitHub. Mainline Linux support is expected in Q1 2024. Sipeed has begun pre-orders on Aliexpress, with prices ranging from $23.90 (2GB RAM, no eMMC flash) to $34.90 (4GB RAM and 32GB eMMC flash), and the LM3H module alone starting at $17.90. Shipping is expected to start on February 11.

⛓ More information...

#boards

🔃 pcb2blender - KiCad to Blender 3D model workflow

The pcb2blender tool revolutionizes the way KiCad projects are rendered in Blender. It streamlines the process of creating professional-looking product renders by allowing users to export their KiCad boards as .pcb3d files and then directly import them into Blender.

This innovation eliminates the tedious cleanup work usually involved, enabling users to focus on scene composition and lighting. Notably faster than previous methods, pcb2blender supports rendering work-in-progress boards or those undergoing last-minute revisions.

More information:
☠️ The Hackaday article
😹 The GitHub page

#libraries #tools #3d_modeling #design #electronics

📺 Running Advanced AI on Raspberry Pi: A Journey with GPT-4 and Beyond

In an innovative exploration of AI's reach beyond high-end hardware, this video delves into the feasibility of deploying advanced language models on accessible technology like the Raspberry Pi 5. The presenter, a seasoned enthusiast in small-tech advancements, embarks on a journey to test a variety of language models including Orca, Fi, and the groundbreaking GPT-4.

With a focus on practicality, the video demonstrates how these models, typically associated with powerful computing resources, can operate on a device as compact and affordable as the Raspberry Pi 5. This approach not only democratizes access to cutting-edge AI but also opens up a realm of possibilities for tech enthusiasts and hobbyists.

The video highlights the Raspberry Pi's capacity to handle models with up to 1.7 trillion parameters, a feat that traditionally requires substantial computing power. The presenter's experiments with local, private AI deployments raise intriguing questions about the potential of small-scale, private AI systems.

⛓ https://youtu.be/Y2ldwg8xsgE

#video #raspberry #artificial_intelligence

🔋 Nuclear battery produces power for 50 years without needing to charge

Betavolt, a Chinese startup, has developed an innovative nuclear battery named Betavolt BV100, which is capable of delivering a constant power output of 3V/100 microwatts for over 50 years without needing a recharge. Measuring just 15x15x5 mm, this battery represents a significant advancement in the miniaturization of atomic energy. While 100 microwatts may not seem substantial, Betavolt has plans to scale up the technology to produce a 1-watt nuclear battery by 2025, potentially revolutionizing power supply in devices like smartphones and drones.

The BV100 utilizes radioactive nickel-63, with a half-life of about 100 years, which decays into non-radioactive copper, generating electricity. This process involves a unique structure comprising a nickel-63 sheet placed between diamond semiconductor converters. Despite the use of radioactive material, safety is a key feature of the BV100. Nickel-63 poses minimal external hazard due to its limited range in air and water, and the battery's design ensures the radioactive material is securely encased. This advancement in nuclear battery technology, echoing the early betavoltaic batteries used in space programs and demonstrates significant progress in power density and safety.

⛓ More information...

#news #energy #battery

📰 Researchers Claim First Functioning Graphene-Based Chip

Researchers at Georgia Tech, led by Professor Walt de Heer, have developed the world's first functioning graphene-based semiconductor, potentially revolutionizing electronics with faster traditional computers and new possibilities for quantum computing.

Published in Nature, their research utilizes epitaxial graphene, a carbon crystal bonded to silicon carbide, to create a novel semiconducting material, semiconducting epitaxial graphene (SEC), also known as epigraphene. This material exhibits enhanced electron mobility compared to silicon, enabling transistors to operate at terahertz frequencies, about 10 times faster than current silicon-based transistors.

This new semiconductor exhibits a significant bandgap, a key requirement for efficient transistors, and could mark a paradigm shift in electronics, moving beyond silicon to more efficient and compatible materials like silicon carbide and boron nitride. While promising, the technology still requires further development to fully realize its potential in digital electronics and quantum computing.

⛓ More information...

#news