REVO.tech - B2B Global Electronics Marketplace的动态

Huge Cost-Savings opportunity on #Nvidia #H100 #GPU PCIe Cards this week Without Long Factory lead times! 900-21010-0300-030 HPE Spare Part Number: P54751-001 We are excited to announce that we have 740 units of the powerful?#Nvidia #H100?in stock and ready for immediate shipping! Key Features: Memory:?80GB HBM2e for handling the most demanding applications Compute Capability:?7,000+ TFLOPS of AI performance Architecture:?Nvidia Hopper, designed for next-generation AI and HPC workloads NVLink:?Supports Nvidia NVLink for high-speed GPU-to-GPU communication Efficiency:?Advanced power management for optimal performance and energy efficiency Ideal For: Artificial Intelligence & Machine Learning Data Science High-Performance Computing (HPC) Deep Learning Rendering Contact us at [email protected] if you require other electronics parts, or if you have your own electronics Excess that you would like to sell! #ampere #electronics #semiconductor #embeddedsystems #iot #semiconductors #electronicmanufacturing #processors #microcontrollers #digitalsignage #nxp #STMicroelectronics #electronicsengineering #components #electroniccomponents #infineon #microcontrollers #microcontroller #cpu #microchip #displays #monitor #kiosks #touchscreen #LCD #displaytechnology #openframe #lcddisplay #samsungdisplay #lg #philips #monitors #monitor #itad #ithardware #excessinventory #excess #excessstock #surplus #wholesalers #wholesaler #wholesale #wholesaledistribution #wholesaledistributor #gpu #ssd #nvidia #nvidiartx #Switches #GPU #DGX #Datacenter #A100 #H100 920-24387-2540-0R0 920-24387-2540-000 920-24387-2540-1R0 920-24387-2540-100 965-2G520-0100-0R0 965-2G520-0100-000 965-2G520-0101-000 965-2G520-0101-0R0 965-2G520-0131-400 965-2G520-0131-4R0 935-23087-0021-300 935-23087-0001-000 935-23087-0101-000 935-23087-0101-0R0 935-23087-0031-400 935-23087-0131-400 935-23087-0131-4R0 935-24287-0000-000 935-24287-0100-000 935-24287-0100-0R0 935-24287-0001-100 935-24287-0001-000 935-24287-0010-300 900-21010-0120-030 900-21010-0020-000 900-21010-0020-001 900-21010-0100-030 900-21010-0000-000 900-21010-0000-001 900-21010-0100-031

New Inventory Alert: #NVIDIA #A2 Half-Length #GPUs Available Now! We are excited to announce the availability of NVIDIA A2 16GB GPUs, specifically designed for compact, high-performance computing environments. These half-length GPUs are ideal for AI inference, machine learning, and data center applications where space and power efficiency are crucial. ? Model Number: 699-2G179-0220-200 ? Memory: 16GB GDDR6 ? Architecture: Ampere ? Form Factor: Half-Length, Single-Slot ? Power Consumption: 60W TDP ? Cooling: Passive Key Features: ? Compact Design: Half-length, single-slot form factor makes the NVIDIA A2 ideal for deployment in dense server environments or edge computing devices. ? Efficient Performance: Powered by the NVIDIA Ampere architecture, this GPU offers robust performance for AI inference and machine learning workloads while maintaining a low power draw of just 60W. ? High Memory Capacity: Equipped with 16GB of GDDR6 memory, the A2 is capable of handling complex models and large datasets with ease. ? Passive Cooling: Designed with passive cooling, making it perfect for environments where airflow and space are limited. ? Versatile Applications: Suitable for use in AI inference, virtual desktops, and other compute-intensive tasks in space-constrained environments. Contact us at [email protected] for more details or to place an order. If you have your own electronics excess that you’d like to sell, we’re here to help! #AI #DeepLearning #HPC #H200 #GH200 #Server #Servers #HDD #NVMe #ITAD #DDR5 #DDR4 #Jetson #Transceivers #excessinventory #electronics #semiconductor #embeddedsystems #iot #semiconductors #electronicmanufacturing #processors #microcontrollers #excessstock #surplus #wholesalers #wholesaler #wholesale #gpu #nvidia #nvidiartx #Switches #GPU #DGX #A100 #H100 #itad #IThardware 900-2G179-0120-001 900-2G179-0120-000 900-2G179-0020-000 900-2G179-0020-001 900-2G179-0120-100 900-2G179-0120-101 900-2G179-0020-100 900-2G179-0020-101

The fabless chip design model has transformed the semiconductor industry. Previously, chip design and manufacturing were often handled by the same company. Now, fabless companies like AMD, NVIDIA, and Qualcomm focus solely on chip design while outsourcing production to specialised foundries such as TSMC and SMIC. In contrast, companies like Intel and Samsung follow the Integrated Device Manufacturer (IDM) model, combining design and manufacturing. This division of labor has spurred innovation, enabling the rapid advancement of technology in various sectors, from smartphones to supercomputers. Breakdown of main players within this space below: 1 $NVDA: Leads in AI and GPU markets, specializing in chips for high-performance computing, artificial intelligence, and cloud infrastructure. Market Cap : $3.43T 2. $TSM: The cornerstone of advanced semiconductor manufacturing with a focus on leading-edge process technology, enabling the fabless model for companies like AMD, NVIDIA, and Qualcomm. Market Cap: $835.76B (26.71T TWD) 3. $AVGO: Dominates in network and connectivity chips, specializing in products critical for data centers, telecom, and infrastructure applications. Market Cap: $826.65B 4. $AMD: A rising competitor in CPUs and GPUs, specializing in high-performance, energy-efficient designs for gaming, data centers, and high-performance computing. Market Cap: $240.15B 5. $QCOM: A key player in 5G and mobile technology, focusing on mobile and wireless chipsets, with a strong patent portfolio for the telecom and mobile industries. Market Cap: $187.32B 6. $INTC: Established in CPU production with a focus on x86 architecture, leveraging an IDM model (Integrated Device Manufacturer) but facing strong competition in a rapidly evolving market. Market Cap: $95.47B

?? Meet Sohu: The Fastest AI Chip of All Time ?? Imagine processing over 500,000 tokens per second with the Llama 70B model. Sohu makes it possible, offering unprecedented capabilities that surpass traditional GPUs. One 8xSohu server can replace 160 H100s. ?? Specialization at Its Best Sohu is the first specialized chip (ASIC) designed for transformer models, delivering unmatched performance. While it can’t run CNNs, LSTMs, or other AI models, its focus on transformers gives it a significant edge. Why Specialized Chips are the Future: ?? Sohu is >10x faster and cheaper than NVIDIA’s next-gen Blackwell (B200) GPUs. ?? One Sohu server processes 500,000 Llama 70B tokens per second—20x more than an H100 server, and 10x more than a B200 server. ?? Benchmarked in FP8 without sparsity, using 8x model parallelism with 2048 input/128 output lengths. Figures are from TensorRT-LLM 0.10.08 (latest version), and 8xB200 figures are estimated. ?? The Limitations of GPUs GPUs are becoming larger, not better. Compute density has only improved by ~15% over the past four years. Upcoming GPUs (like NVIDIA B200, AMD MI300X, Intel Gaudi 3, AWS Trainium2) are merely doubling chips to boost performance. ?? With Moore’s law slowing, specialization is the path forward. Economics of Scale: ?? AI models now cost $1B+ to train and generate $10B+ in inference. At this scale, even a 1% improvement can justify a $50-100M custom chip project. ?? ASICs are 10-100x faster than GPUs. This was evident with bitcoin miners in 2014, and it’s now transforming AI. The Hardware Lottery: Transformers dominate because they run fastest and cheapest on current hardware. AI labs have optimized transformers extensively, making them the most efficient choice for large-scale models. The Future: As Sohu and other ASICs enter the market, a shift is inevitable. Future transformer killers will need to outperform transformers on GPUs and custom chips alike. When they do, we’ll be ready to build the next generation of ASICs. ?? Get Ready for the Sohu Revolution. #Innovation #ArtificialIntelligence #Chip

AMD Unveils AI-Focused Processor Lineup AMD has launched its latest lineup of AI-optimized processors at Advancing AI 2024, targeting the booming data center and AI chip market. Key Features: ?? ?Ryzen AI PRO 300: 40% better performance than Intel's Core Ultra chips for enterprise AI PCs ?? ?Instinct MI325X AI accelerator: 1.8x higher memory capacity, 1.3x more bandwidth than Nvidia's H200 GPU ?? ?EPYC 5th Gen CPUs: "World's best for enterprise, AI, and cloud" (CEO Lisa Su) ?? ?Annual AI chip releases planned: MI350X (2025), MI400 (2026) Impact: ?? ?AMD aggressively competes with Nvidia and Intel in the AI chip market ?? ?Expands possibilities for AI adoption in enterprise, cloud, and data centers ?? ?Addresses growing demand for powerful AI processors Article: https://lnkd.in/gy2bacSp How will AMD's new AI-focused processors impact your business or projects? Will you leverage these advancements? #AMD #AIProcessors #DataCenter #ArtificialIntelligence #TechInnovation #ChipWar

Intel has announced a new AI chip called the 6th generation "Xeon6" server CPU AI chip at the Computex Conference in Taipei, Taiwan. This chip is aimed at challenging competitors Nvidia and AMD in the AI sector. In addition, Intel unveiled the "Gaudi 3 Accelerator Kit," an AI training chipset that offers cost effectiveness compared to Nvidia's H100. Intel's CEO, Pat Gelsinger, highlighted the improved performance and lower power consumption of the Xeon6 chip. This announcement comes as Intel seeks to regain market share in the data center sector, where it currently lags behind AMD. Gelsinger also outlined a strategy to position Intel as a serious competitor to Nvidia by offering superior cost effectiveness with their AI accelerators. Additionally, Intel announced a next-generation processor for laptops called "Lunar Lake," which boasts enhanced performance and reduced power consumption. The products will be manufactured using TSMC's 3-nm process. https://lnkd.in/g45CEnxv

Hello to the smart machines The race for AI chip dominance Intel and AMD are already leading the charge on AI chips for PCs. Intel has come together with Microsoft to define what an AI PC is. In addition to needing a CPU, GPU, and NPU, these PCs should be able to run Microsoft Copilot locally and feature a physical Microsoft Copilot key on the keyboard. Looking at AMD, the company announced a new series of AI processors in April 2024, with its next generation Ryzen chips expected to power models for many of the leading PC brands. And of course, the Star of the AI revolution, Nvidia has unveiled new GPUs designed to enhance generative AI capabilities namely b200 gpu series, as well as releasing Chat with RTX, an application that makes it easier to run a large language model on a Windows PC. Google also wants to lead this revolution and announced Trillium, its sixth generation of Tensor processors.CEO Sundar Pichai just announced new Trillium chips, coming later this year, that are 4.7 times faster than their predecessors. Meanwhile, Apple unveiled an iPad Pro powered by its M4 chip capable of running artificial intelligence (AI) applications locally. Apple said the new M4 chip is "more powerful than any neural processing unit in any AI PC today." last but not least, Qualcomm has been working on advanced AI-focused Snapdragon processors and has now become a serious alternative to AI PCs based on chips using Intel architecture.

NVIDIA Blackwell's High Power Consumption Drives Cooling Demands; Liquid Cooling Penetration Expected to Reach 10% by Late 2024, Says TrendForce Corporation ?? With the growing demand for high-speed #computing, more effective cooling solutions for #AI #servers are gaining significant attention. TrendForce Corporation's latest report on AI servers reveals that #NVIDIA is set to launch its next-generation Blackwell platform by the end of 2024. Major CSPs are expected to start building AI server #datacenters based on this new platform, potentially driving the penetration rate of liquid cooling solutions to 10%. TrendForce reports that the NVIDIA #Blackwell platform will officially launch in 2025, replacing the current Hopper platform and becoming the dominant solution for NVIDIA's high-end #GPUs, accounting for nearly 83% of all high-end products. High-performance AI server models like the B200 and GB200 are designed for maximum efficiency, with individual GPUs consuming over 1,000W. HGX models will house 8 GPUs each, while NVL models will support 36 or 72 GPUs per rack, significantly boosting the growth of the liquid cooling supply chain for AI servers. TrendForce highlights the increasing TDP of #server #chips, with the B200 chip's TDP reaching 1,000W, making traditional air cooling solutions inadequate. The TDP of the GB200 NVL36 and NVL72 complete rack systems is projected to reach 70kW and nearly 140kW, respectively, necessitating advanced liquid cooling solutions for effective heat management. Thanks again to TrendForce Corporation for the full article with more background and insights via the link below ?????? https://lnkd.in/eVkkM7Av #semiconductorindustry #semiconductors #semiconductormanufacturing #technology #chip #chips #artificialintelligence #tsmc #icdesign #usa #it #taiwan #advancedpackaging #computer #computing #innovation #cpu

#latesttechnews #intelprocessors #intel #processors #lunarlake #tech #latesttechnology #technology ?? After much anticipation, Intel has finally introduced its Core Ultra 200V series processors, also known as Lunar Lake mobile processors. The Core Ultra mobile processors (Series 2) are completely redesigned for efficiency. Intel says the new Lunar Lake processors will break the myth that x86 can’t be as efficient as ARM processors such as Qualcomm’s Snapdragon X series chipsets. ?? Lunar Lake processors have a frequency range between 3.5GHz to 5.1GHz. And the default TDP is 30W (including memory) with a turbo power boost of up to 37W across all SKUs. Fanless laptops may have a TDP between 8W to 17W. ?? The Xe2 GPU is also quite powerful packing eight 2nd-gen Xe cores with Ray Tracing and XeSS (Super Sampling) support for AI upscaling. The new GPU delivers up to 31% better GPU performance than Meteor Lake; 68% better than SD X Elite GPU, and 16% better than AMD HX 370 GPU. The GPU can also perform up to 67 trillion operations per second (TOPS) to process AI workloads. ?? Finally, the NPU can perform up to 48 TOPS and it even beats the Snapdragon X Elite NPU in the UL Procyon AI Computer Vision test. In the Geekbench AI test, Intel’s NPU consistently delivers better AI performance than Snapdragon X Elite’s Hexagon NPU in many data types. You also get Wi-Fi 7 and Bluetooth 5.4 support with Lunar Lake processors https://lnkd.in/dxApqwyv

NVIDIA’s H200 vs. AMD’s MI300X AI chip giants NVIDIA and AMD have been under heated competition for a couple of years. NVIDIA, though controls the lion’s share of the market for AI computing, had been challenged by AMD while the latter launched Instinct MI300X GPU, claiming the product to be the fastest AI chip in the world, which beats NVIDIA’s H200 GPUs. AMD’s MI300X: More TRANSISTORS, More MEMORY CAPACITY, More ADVANCED PACKAGING…With a HIGHER COST ?? NVIDIA’s H200 is implemented using TSMC’s N4 node with 80 billion transistors. On the other hand, AMD’s MI300X is built with 153 billion transistors, featuring TSMC’s 5nm process. Number of transistors in the logic compute die and the total die size/total cost are roughly proportional. AMD’s MI300X, equipped with nearly twice the number of transistors compared to NVIDIA’s H200, therefore, is said to cost twice as much of the latter in this respect. NVIDIA’S 80% MARGIN: HIGH At FIRST GLANCE, BUT ACTUALLY JUSTIFIABLE?? the H200 outperforms the MI300X by over 40%. This means that if AMD wants to maintain a similar cost/performance ratio NVIDIA claims the demand for Hopper remains strong, while Blackwell chips will potentially generate billions of dollars in revenue in the fourth quarter. AMD’s Instinct MI300 series, on the other hand, has emerged as a primary growth driver, as it is expected to generate more than USD 4.5 billion in sales this year. Credit:https://lnkd.in/gmYN2CvM