Erudite Asia’s Post

Huida's next-generation AI chip was revealed to be delayed by three months before mass production. Nvidia's next-generation Blackwell architecture AI flagship chip, the GB200, will face a delay of at least three months before mass production due to design defects. An undisclosed Microsoft employee indicated that Nvidia informed Microsoft of the shipment delays last week. This delay affects major tech companies, including Google, which ordered over 400,000 chips with a value exceeding $10 billion, and Meta, which also placed a $10 billion order. The GB200 chip includes two interconnected Blackwell GPUs and a Grace CPU. The design flaw was identified by TSMC engineers during preparations for mass production, resulting in the need for design adjustments and new trial production. Initially expected to begin mass production in the third quarter and shipments in the fourth quarter, the timeline has now shifted to mass production in the fourth quarter and shipments postponed to the first quarter of the following year. #AI #Taiwan https://lnkd.in/gYXSTeqg

It's cool to see the Enterprise AI Infrastructure brand leader survey results from IT Brand Pulse featured in Adam Zaki's recent article on 5 AI trends CFOs must know! Dell Technologies NetApp NVIDIA AMD Samsung Semiconductor Intel Corporation #enterpriseai #ai #gpus #cpus #storage #networking

Intel's Gaudi 3: The Powerful AI Chip Shaking Up the Tech World! #AIcomputingmarket #Amperearchitecture #autonomousvehicles. #competitioninAIcomputingmarket #deeplearning #energyefficiency #finance #Gaudi3impact #Google #healthcare #IntelGaudi3 #IntelsmoveintoAIcomputing #Microsoft #naturallanguageprocessing #NVIDIAdominance #performanceperwatt #PyTorch #softwarecompatibility #softwareframeworks #techindustryreputation #TensorFlow

Return to leadership trend continues - 1) Nov 2023: Lead industry on 1st AIPC industry wide CPU, Meteor Lake 2) Sept 2024: quantum leap in industry performance, battery, performance/watt, graphics and AIPC workloads vs ARM variants and AMD - at IFA last month with our Lunar Lake AIPC processor launch 3) Sept 2024: Amazon and US Government award Intel Foundry multi-billion contracts using Intels 18A process 4) and now - per below, industry-analysts highlighting Intel doubling performance in our latest gen Enterprise AI portfolio in Xeon 6 and Gaudi 3 Bold vision - Four nodes in five years. Momentum Inside!

In today's fast-paced business environment, companies are eager to integrate generative AI into their operations in order to introduce new services to the market. This growing demand places significant pressure on data center infrastructure. While training large language models presents one set of challenges, the ability to deliver real-time services powered by these models presents another entirely.

With Gaudi 3, Intel Can Sell AI Accelerators To The PyTorch Masses

Our Instinct MI300X accelerators already power leading AI models from OpenAI, Meta and Hugging Face. With the introduction of the Instinct MI325X AI chip, we are redefining performance for the most demanding AI workloads. Learn more from Quartz: https://bit.ly/4841mkS

Cerebras Systems launched Wafer Scale Engine 3 (WSE-3), an AI wafer-scale chip that can accommodate 4 trillion transistors; has 900,000 AI cores; 44GB on-chip SRAM; based on a 5-nanometer architecture. The WSE-3 can be used to train AI models up to 24 trillion parameters delivering a peak performance of 125 FP16 PetaFLOPS. The chip requires 97% less code to train LLM as compared to current GPUs. The WSE-3 design has been improved significantly compared to its predecessor WSE-2 which was based on 7-nanometer architecture with a transistor count of 2.6 trillion only. The chip would power Cerebras's CS-3 supercomputer and used in government, cloud, and medical applications. #aichips #semiconductor Links: https://lnkd.in/gQS_RRFa https://lnkd.in/gMgCtSac https://lnkd.in/g7AKPf7Q https://lnkd.in/gSuET9Tw https://lnkd.in/gDjggB6B

Damo: GPU upgrades and expanded applications promote AI development The development of AI technology is being promoted by GPU design upgrades, expanded product applications, and increased supply chain participation, according to a Morgan Stanley report. The report reveals specific plans and developments by Huida and Gigabyte, and points to the potential market size of cloud AI semiconductors reaching US$100 billion by 2024. Additionally, TSMC and SK Hynix have formed an AI semiconductor alliance, indicating TSMC's leading position in advanced packaging of AI chips. #AI #Taiwan https://lnkd.in/gQkAtPi7

TSMC’s Q3 is not very prosperous as AI-enabled Apple-eating mature process wafer fabs TSMC's third quarter performance may not meet expectations due to competition from China's mature processes and low prices. TSMC's 3 and 5 nm production capacity is nearly full, with high demand for server GPUs. Other wafer foundries like UMC, Advanced Semiconductor, Power Semiconductor, and Mosilicon are benefiting from customer orders but facing challenges due to price competition and conservative customer behavior. #Semiconductors #Taiwan https://lnkd.in/g3g9bN9G