Arteris’ Post

🏆#EICAccelerator awards RAAAM™ Memory Technologies €5.2M for a unique on-chip memory technology! 📈 The Memory Bottleneck is a major issue in modern Systems-on-a-Chip (SoC), causing on-chip memory to grow and occupy up to 75% of the SoC area. However, standard #SRAM technology is inefficient with space and struggles with scaling in advanced technologies.  🔧#RAAAM’s Gain-Cell Random Access Memory (GCRAM) technology uses only 2-3 transistors to store a bit of data, compared to the 6-8 transistors required by SRAM. This makes GCRAM a more area-efficient and cost-effective solution! 💡Learn more about this high-end innovation: https://lnkd.in/d3aGz2br European Innovation Council and SMEs Executive Agency (EISMEA) #AlienTT #Innovation #Tech #EICAccelerator #RAAAM #MemoryTechnology #SoC

📢 News Alert: VeriSilicon Holdings Co., Ltd., a one-stop custom silicon service and semiconductor IP licensing service provider, has licensed Arteris FlexNoC 5 interconnect IP which offers enhanced cost and design efficiency, scalability and reliability for VeriSilicon’s high-performance data center SoC solutions. “We have a long-standing business relationship with Arteris, which was one of the key factors in our decision to adopt their latest interconnect IP technology,” said Wiseway Wang, senior vice president and general manager of Custom Silicon Platform Division of VeriSilicon. “Arteris’ interconnect IP technologies have been applied in various VeriSilicon’s SoC designs like multimedia, ADAS and data center solutions. The latest FlexNoC 5 interconnect IP empowers us to design data center SoCs with high-performance computing architectures, while effectively addressing timing closure and congestion challenges to ensure on-time delivery.” Read full news release. https://bit.ly/3ML0bgh #semiconductor #innovation #SystemOnChip #NetworkOnChip #Arteris

Let us compare. Directory-based systems excel in scalability, but snooping wins in lower overhead and reduced complexity. Discover the power of Ncore IP's snooping mechanism, reducing coherence overhead while maintaining cache coherency. With efficient snooping, Ncore IP from Arteris simplifies system design, streamlines communication, and facilitates scalability. Experience enhanced performance and seamless cache coherence management. Learn more about Ncore. Unlock the power of Ncore for optimal system performance and scalability! https://bit.ly/3WlxMDo #semiconductor #innovation #SystemOnChip #NetworkOnChip #Arteris

🚀 Emergent Vision Technologies: RDMA and Our Optimized GigE Vision Solutions As the industry discusses #RDMA at the recent #IVSM meeting in Japan, it's essential to evaluate how this technology compares to our proven solutions. At Emergent Vision Technologies, while we are ready for the potential standardization of RDMA, we don't just follow standards—we set them. Optimized GigEVision + GVSP: Our strategy utilizes hardware-based header splitting in off-the-shelf NICs, ensuring compliance with the mature GigE Vision standard while minimizing latency and enhancing performance. This method not only reduces data loss but is also proven in scenarios where RDMA might introduce unnecessary complexity and overhead. Innovating Beyond the Standard: #FlexProc and #FlexTrans: These technologies showcase our commitment to flexibility and control, enabling tailored data management and processing solutions across various system nodes. #GPUDirect: We've long utilized GPUDirect, now offering our customers and partners easy access through our eCapturePro software. All our existing hardware is RDMA-ready, ensuring that if the GigE Vision standard pivots to RDMA, our cameras will seamlessly adapt. However, our goal is to exceed standards, providing systems that are not only compliant but optimized for real-world applications, thus ensuring peak performance and cost-efficiency. 🌟 Why Choose Us? We go beyond typical solutions to offer enhanced capabilities tailored to our customers' specific needs. With over a decade of innovation in high-speed GigE cameras—10+ years with 10GigE, 5+ years with 25GigE, and 3+ years with 100GigE—we are industry leaders in deploying robust (multi-)camera systems. In 2024, we standardly support customers with 20+ camera systems, achieving zero frame drops and low latency, while others struggle to stabilize even a single 25GigE or 10GigE camera. This extensive experience informs our upcoming 5GigE EROS Camera Family, designed for traditional industrial markets where reliability and low latency are paramount, especially in multi-camera setups. We're not just keeping up; we're setting new standards, ensuring your investments today prepare you for the challenges of tomorrow. Stay ahead in machine vision technology with a partner who values innovation and customization. 🔗 Explore UDP, TCP and RDMA for GigE Vision Cameras: https://lnkd.in/gYYASbpt 🔗 Explore our upcoming 5GigE EROS Family: https://lnkd.in/euhfwwBA 🔗 Explore our Plug-In Technology with eCapturePro: https://lnkd.in/ecRhfPYD 🔗 Explore FlexProc & FlexTrans: https://lnkd.in/eHqAf82v #GigEVision #MachineVision #EmergentVisionTech #Innovation #TechnologyLeadership #5GigE #10GigE #25GigE #100GigE

Very well put, Tom: our EMEA super-hero. I will only add that as we released 10G cameras 10+ years ago, 25G cameras 5+ years ago, and 100G cameras 2+ years ago that the MV standards tech committees were doing absolutely nothing for high speed. #rdma is so in its infancy. A plug fest at #ivsm has some claiming a major victory taking a single camera into a computer between one companies software and another companies 10G camera. A very long way to go in the very slow moving machine vision industry for this to formally standardize and mature where many companies adopt and release with full interoperability. We have very mature solutions for example that support one customer in a 500 camera single system. Frankly, doing RDMA is only the start of a very long journey to where we are now. And we keep moving forward with other bleeding edge tech applicable to machine vision and many other industries.

#Cell-#free #massive #multiple-#input #multiple-#output (#CF #mMIMO) network has gained significant attention due to its ability to provide a ubiquitous connectivity over a wide geographical area without having any cell boundary. It serves the #user #equipments (#UEs) by employing multiple #access #points (#APs) connected to a #central #processing #unit (#CPU). In this paper, they focus on designing beam combiners for the #APs under limited availability of the pilots. They argue that the received #signal-#to-#interference #plus #noise (#SINR) can be maximized if received signals at each #AP are beam combined in a way that it aligns with the desired signal and nullifies the interference arising due to the pilot reuse. They use this strategy to improve three popular beam combiners for the #CF #mMIMO network based on the #maximum #ratio (#MR), #generalized #MR (#GMR), and #reduced-#complexity #minimum #mean #square #error (#RCMMSE) criteria, respectively. ----Santosh Kumar Singh, Abhay K. Sah More details can be found at this link: https://lnkd.in/gtT2QhfJ

𝗦𝗘𝗡𝗞𝗢 𝗿𝗲𝗰𝗲𝗶𝘃𝗲𝘀 𝗜𝗻𝗳𝗼𝘀𝘁𝗼𝗻𝗲 𝗘𝘅𝗰𝗲𝗹𝗹𝗲𝗻𝘁 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝗔𝘄𝗮𝗿𝗱 𝗳𝗼𝗿 𝗦𝗡-𝗠𝗧 SENKO, the pioneer and industry leader in Very Small For Factor (VSFF) connectivity has been awarded the 𝙄𝙣𝙛𝙤𝙨𝙩𝙤𝙣𝙚 𝙀𝙭𝙘𝙚𝙡𝙡𝙚𝙣𝙩 𝙏𝙚𝙘𝙝𝙣𝙤𝙡𝙤𝙜𝙮 𝘼𝙬𝙖𝙧𝙙 for its ground-breaking SN-MT connector. More about SN-MT: https://lnkd.in/dn4diuPT The SN-MT is the most compact, low-loss multi-fiber connector in the world with a staggering 2.7 times size reduction compared with legacy MPO connectors. The SN-MT16 is currently the preferred choice for Hyperscale AI Cluster applications where packing-density and size are of paramount importance. With the explosive ramp-up of 800G Dual MPO8 deployments, the SN-MT16 is the ideal choice for consolidating transceivers into one high-density patch panel. Up to 216 transceivers can be connected to a single 1RU patch panel. The SN-MT family can also support 24f and 32f giving operators the perfect building block for future network expansion and evolving data rates.. 𝗦𝗘𝗡𝗞𝗢 𝘞𝘦 𝘷𝘢𝘭𝘶𝘦 𝘦𝘷𝘦𝘳𝘺 𝘤𝘰𝘯𝘯𝘦𝘤𝘵𝘪𝘰𝘯 #senko #fiberoptics #SN-MT

This article focuses on advancing #outdoor #wireless #systems to better support ubiquitous #extended #reality (#XR) applications, and close the gap with current indoor wireless transmission capabilities. The authors propose a #hybrid #knowledge-#data #driven method for channel #semantic acquisition and multi-user beamforming in cell-free massive #multiple-#input #multiple-#output (#MIMO) systems. Specifically, they firstly propose a data-driven #multiple #layer #perceptron (#MLP)-Mixer-based auto-encoder for channel semantic acquisition, where the pilot signals, CSI quantizer for channel semantic embedding, and CSI reconstruction for channel semantic extraction are jointly optimized in an end-to-end manner. ---- Zhen Gao, Shicong Liu, Yu Su, Zhongxiang Li, Dezhi Zheng More details can be found at this link: https://lnkd.in/eH9dyvMn

IonQ unveils thrilling plans for 2024/25! Innovations like #ForteEnterprise and Tempo promise to transform #quantumcomputing with advanced qubit tech and #modulararchitecture. Dive into the details here: https://ow.ly/5Xyx50Sw9mL

Photonic Integration in Next-Gen Optical Networks: 🌐💡 Explore the cutting-edge transformation of optical networks with photonic integration. This technology seamlessly combines various functions on a single photonic chip, enhancing efficiency, speed, and scalability. #PhotonicIntegration #OpticalInnovation 🚀🔗#advancedtechnology