SemiPulse | AI-Powered Semiconductor Supply Chain Intelligence & Market Signals

Qualcomm’s HBC Stacks Compute Beneath DRAM To Smash The AI Memory Wall, Claiming 6x The Bandwidth Per Watt Of HBM - Wccftech

0.92

wccftech.com 2026-06-25 Wccftech

Qualcomm is unveiling its breakthrough for the AI data center market called HBC (High-Bandwidth Compute), to break the memory wall. At its Investors Day 2026, Qualcomm unveiled HBC under its Dragonfly brand, an innovative technology that aims to offer a major boost in memory capacity and bandwidth. The HBC architecture deploys a purpose-built & near-memory solution that bonds compute with boosted

Qualcomm AI Data Center Memory Wall HBC LPDDR 3D Integration Bandwidth Per Watt AI Accelerator

Qualcomm Reportedly Helping China’s ByteDance With Custom Chip Design Services, Also Acquires Modular To Accelerate Its Unified Compute Strategy - Wccftech

0.92

wccftech.com 2026-06-25 Wccftech

Qualcomm is expanding its chip-design services to China as it is reportedly in early talks with ByteDance. Custom-chip designs are picking up momentum across the tech industry, and a report by Reuters suggests that Qualcomm might be assisting ByteDance, a leading Chinese firm, with chip-design services. Although the talks between Qualcomm and ByteDance are still early, if they succeed, then that

Qualcomm ByteDance Custom Chip Chip Design Services AI Chips Semiconductor Industry US-China Tech Relations Chip Diversification

After Earning Major Profits From HBM, SK Hynix Now Plans To Prioritize DDR5 “General-Purpose DRAM” Production - Wccftech

0.92

wccftech.com 2026-06-23 Wccftech

SK Hynix is planning to return to general-purpose DRAM, such as DDR5, after securing record revenue from HBM memory. HBM has been the main driver for the big three DRAM manufacturers for the last couple of months due to increased demand from AI firms. This has led to record profits for Samsung, Micron, and SK Hynix. At the same time, these companies stopped focusing production on general-purpose

DRAM production HBM memory DDR5 SK Hynix NVIDIA AI chips memory market semiconductor supply chain

JEDEC Approves SPHBM4 to Break HBM’s Costly Packaging Bottleneck, Retaining HBM4-level Speeds With Standard Packages - Wccftech

0.92

wccftech.com 2026-06-23 Wccftech

SPHBM4 is a new JEDEC standard that aims to solve the high cost and packaging concerns with existing HBM technologies. Almost all AI & HPC accelerators rolling out these days feature some form of HBM memory. The highest-end solutions are leveraging the latest HBM4 designs, and HBM4E is being sampled to the top chipmakers. But as demand continues to rise, and shortages persist in the premium DRAM

HBM JEDEC SPHBM4 High Performance Memory Packaging Technology AI Chips HPC Accelerators DRAM

Micron Deploys Claude To Accelerate AI Factories While Anthropic Gets Backing To Support Its Multi-Year Growth Trajectory - Wccftech

0.92

wccftech.com 2026-06-22 Wccftech

Micron and Anthropic have announced a strategic agreement to scale next-generation AI infrastructure that spans memory and storage for rising compute demands. Press Release: Micron Technology today announced a strategic agreement with Anthropic that spans memory and storage AI architecture design, supply and demand, enterprise adoption of Claude across Micron, and a strategic investment in Anthro

Micron Anthropic AI Infrastructure Memory Technology Storage Technology AI Models High-Bandwidth Memory DRAM

NVIDIA Prepares For Next-Gen GPU Architectures As “Blackwell-Next” Spotted In Linux 7.2 Kernel Patch - Wccftech

0.92

wccftech.com 2026-06-22 Wccftech

The Linux 7.2 Kernel Patches have seen the addition of a next-gen NVIDIA GPU architecture, cleverly listed as "Blackwell-Next". NVIDIA's GPU architectures are not a mystery since the company has been outlining them in roadmaps for years now. We knew that Rubin was going to replace Blackwell years before its launch, and now we know that Feynman will replace Rubin. NVIDIA has only started volume pr

NVIDIA GPU Architecture Linux Kernel Blackwell Rubin Feynman Grace CPU Vera CPU

Exynos 2700’s Innovative Heat Dissipation Solutions Are Key To Compensate For Samsung’s Inferior 2nm Process, Rumor Pits TSMC’s N2P Node As Superior - Wccftech

0.85

wccftech.com 2026-06-21 Wccftech

Samsung has proven that it can develop cutting-edge lithography, with its upcoming Exynos 2700 serving as an excellent example, as the company has been reported to be progressing well with its next flagship SoC. However, as far as its 2nm GAA process goes, a rumor states that when comparing Power, Performance, and Area (PPA) metrics, TSMC’s 2nm N2P is ahead in the race. This may explain why, along

Samsung Exynos 2700 TSMC 2nm process heat dissipation chip performance mobile SoC semiconductor manufacturing

TSMC Accelerates CoPoS Packaging to Replace CoWoS, as Glass Core Substrates Cut Costs 30% and Boost Wafer Utilization Past 90% - Wccftech

0.92

wccftech.com 2026-06-21 Wccftech

TSMC is aggressively working on CoPoS (Panel-Level) packaging to replace CoWoS for growing compute demand as Glass Core Substrates take center stage. The ever-growing AI and compute demand requires next-generation packaging technologies. Intel and TSMC are aggressively working towards that goal, and glass core substrates are going to be a major part of their trajectories moving forward. In a rec

Semiconductor Packaging CoPoS CoWoS Glass Substrate Wafer Utilization AI Chips Advanced Packaging TSMC

TSMC Accelerates CoPoS Packaging to Replace CoWoS, as Glass Core Substrates Cut Costs 30% and Boost Wafer Utilization Past 90% - Wccftech

0.92

wccftech.com 2026-06-21 Wccftech

TSMC is aggressively working on CoPoS (Panel-Level) packaging to replace CoWoS for growing compute demand as Glass Core Substrates take center stage. The ever-growing AI and compute demand requires next-generation packaging technologies. Intel and TSMC are aggressively working towards that goal, and glass core substrates are going to be a major part of their trajectories moving forward. In a rec

Semiconductor Packaging TSMC CoPoS CoWoS Glass Substrate AI Chips Advanced Packaging Wafer Utilization

Open-Source NVIDIA NVK Vulkan Driver Receives DLSS Support In Mesa 26.2 - Wccftech

0.92

wccftech.com 2026-06-21 Wccftech

The latest update brings NVIDIA's AI upscaling technology to the open-source NVK driver stack, improving the gaming experience on Linux. The open-source NVIDIA NVK Vulkan driver has taken a major step forward on Linux as developers merge initial support for NVIDIA's DLSS technology into Mesa 26.2. This enables AI-powered upscaling in compatible Vulkan games running on Linux operating systems. Fo

Open Source Driver NVIDIA DLSS Vulkan Linux Gaming Graphics Driver Mesa GPU

Foxconn Pegs NVIDIA Vera Rubin AI Datacenter at $47 Billion Per Gigawatt, as Power Bills Hit $1.3B Yearly - Wccftech

0.92

wccftech.com 2026-06-21 Wccftech

NVIDIA's Vera Rubin is entering data centers soon, and the rise of Agentic AI also comes with a mammoth rise in costs. The Vera Rubin era is upon us. The first systems are already being shipped to major cloud providers who are validating and testing them before bringing them up on scale. With the full volume production already going on, NVIDIA is eyeing an even bigger success than Blackwell. But

NVIDIA Vera Rubin AI Data Center Foxconn Agentic AI Compute Cost Power Consumption Data Center Construction AI Infrastructure

RTX Spark Is Apple’s And Qualcomm’s Kryptonite In One Crucial Area And These Two Only Have A Year To Get Things Right - Wccftech

0.92

wccftech.com 2026-06-20 Wccftech

The ARM-based ecosystem has been graced by another entrant in the form of NVIDIA’s RTX Spark, and with its arrival in laptops this fall, Qualcomm and Apple should feel a little hint of worry, mainly in one category, because they have had years to address a problem existing in these notebook SoCs and have failed to do so. Well, the clock is ticking for these two companies.

Semiconductor Chip Mobile Processor ARM Architecture GPU Apple Qualcomm NVIDIA

Qualcomm Is Copying Samsung Exynos 2600’s Heat Path Block For Snapdragon 8 Elite Gen 6 Pro, But Has Botched The Implementation - Wccftech

0.85