RISC-V
PulseAugur coverage of RISC-V — every cluster mentioning RISC-V across labs, papers, and developer communities, ranked by signal.
- developed by University of California, Berkeley 100%
- instance of jalapeño 90%
- acquired by Tenstorrent 90%
- competes with Arm Holdings 70%
- used by Qualcomm 70%
- acquired Qualcomm 70%
- used by Tenstorrent 70%
- instance of central processing unit 70%
- used by system on a chip 70%
- uses Qualcomm 70%
- used by Arm Holdings 60%
- affiliated with Arm Holdings 50%
10 day(s) with sentiment data
-
Banana Pi launches RISC-V AI computer module matching Raspberry Pi CM4 size
The Banana Pi BPI-CM6 is a new RISC-V computer module designed to be the same size as the Raspberry Pi CM4. It offers 2.0 TOPs of AI processing power, includes 4GB of DDR RAM, and 16GB of eMMC storage. The module is cur…
-
OpenAI accelerates custom AI chip design with AI assistance
OpenAI is developing its own custom AI chips, codenamed 'Jalapeño,' with the first generation aiming for deployment by the end of the year. AI is significantly accelerating the design process, reducing the time for cert…
-
CapsuleMotion optimizes VCE energy use and accuracy
Researchers have developed CapsuleMotion, a novel system designed to optimize energy consumption and improve the efficiency of Video Capsule Endoscopy (VCE). This patient-specific dynamic capsule behavior predicts motio…
-
Tomb Raider playable on $25, 1-watt microcontroller
A retro gaming enthusiast has successfully run the classic game Tomb Raider on an ESP32-P4 microcontroller, a low-power, inexpensive development board. The dual-core 400 MHz RISC-V processor, consuming about 1 watt, was…
-
d-Matrix adopts NVIDIA NVLink Fusion for AI inference deployment
AI inference chipmaker d-Matrix has announced its adoption of NVIDIA's NVLink Fusion technology for its upcoming Raptor XPUs. This integration will allow d-Matrix to connect its custom silicon directly into NVIDIA's AI …
-
Intel Processor Trace detailed as powerful new debugging tool
Brendan Gregg's article "Your Intellectual Fly Is Open" discusses the capabilities and implications of Intel Processor Trace (Intel PT). Gregg explains how Intel PT can be used for deep system tracing, similar to dtrace…
-
Haiku OS Beta 6 launches with hardware virtualization and Firefox support
Haiku OS has released its sixth beta, introducing significant updates including a hardware-accelerated hypervisor and enhanced web browser support. The WebPositive browser now utilizes the WebKit rendering engine for im…
-
New RISC-V Chip Aims to Solve Video Generation's "Storage Wall" Bottleneck
Zhongke Tongliang has launched the SmarCo GC3, a novel AI chip designed for video generation tasks. This chip utilizes a RISC-V dataflow architecture, aiming to overcome the "storage wall" bottleneck inherent in traditi…
-
AI agents solve complex problems but lack oversight, leading to costly errors
AI agents are demonstrating advanced capabilities, with Google's Antigravity framework successfully solving complex math problems and contributing to open-source projects. However, this progress is shadowed by significa…
-
LLMs struggle with RISC-V accuracy, author offers debugging fixes
A technical article discusses how Large Language Models (LLMs) often provide incorrect information about the RISC-V instruction set architecture. The author suggests that presenting LLMs with complex debugging tasks, si…
-
AI could halve chip design time but won't replace engineers, says Cadence exec
Cadence executive Teng Jinqing believes AI can significantly accelerate chip design, potentially halving the time required for certain tasks by automating repetitive processes and improving engineer productivity. Howeve…
-
AI Agents Spark CPU Market Resurgence, Igniting Five-Year Investment Window
The AI inference era, particularly with the rise of Agents, is driving renewed interest and investment in CPUs, shifting them from a supporting role to a more central one. This surge in demand, coupled with price increa…
-
CAS Flux launches OmniMate BOX for AI digital employees
CAS Flux (SMARCO) has launched the OmniMate BOX, a lightweight AI digital employee solution designed to upgrade existing screens into interactive terminals. This system utilizes a split architecture, connecting a local …
-
OpenAI's Jalapeño chip to excel at inference, Nvidia Groq 3 benchmarks revealed
The Register's AI section highlights several developments, including OpenAI's upcoming Jalapeño chip, which is expected to excel at inference tasks. Additionally, benchmarks for Nvidia's Groq 3 LPU suggest potential for…
-
NVIDIA Groq 3 LPX inference chip enters full production, targets agentic AI
NVIDIA has announced that its Groq 3 LPX inference chip has entered full production, reportedly achieving 3,400 tokens per second for agentic AI workloads. While NVIDIA claims this is four times faster than Cerebras, th…
-
AI advances in satellites, computing, and music disclosure
Orbitworks, an Abu Dhabi-based company, has sold capacity on its Altair satellite to CNES, marking a step in the UAE's space industry expansion. Meanwhile, the Karnataka government in India has launched KEO, an AI-ready…
-
Mistral AI integrates Chinese AI model, highlighting EU's tech sovereignty challenge · 2 sources tracked
European AI company Mistral AI is facing scrutiny for its decision to integrate China's GLM-5.2 model into its European-controlled AI infrastructure. This move, aimed at fostering European tech sovereignty by focusing o…
-
Chestnut OS: New OS Built for Machine Learning Workloads
Chestnut OS is a new operating system designed specifically for machine learning workloads. It supports a range of architectures including ESP32, ARM, and RISC-V. The OS was developed for Beluga Navigation Systems and i…
-
AMD, Google, Cerebras advance AI hardware; Stripe invests in AI gateways
The Register's AI section covers several distinct developments, including AMD's progress in AI energy efficiency, Google's strategic moves in the AI chip market involving Marvell and Broadcom, and Cerebras's CS-4 system…
-
LLM agents create adaptive hardware Trojans to test detector weaknesses
Researchers have developed TrojanGYM, a novel framework that utilizes multiple large language models to create adaptive hardware Trojans. These Trojans are designed to bypass existing learning-based detectors by generat…