惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

WordPress大学
WordPress大学
博客园 - 司徒正美
小众软件
小众软件
H
Help Net Security
博客园 - 聂微东
宝玉的分享
宝玉的分享
Jina AI
Jina AI
酷 壳 – CoolShell
酷 壳 – CoolShell
阮一峰的网络日志
阮一峰的网络日志
M
MIT News - Artificial intelligence
博客园 - 【当耐特】
U
Unit 42
大猫的无限游戏
大猫的无限游戏
Apple Machine Learning Research
Apple Machine Learning Research
S
SegmentFault 最新的问题
腾讯CDC
MongoDB | Blog
MongoDB | Blog
云风的 BLOG
云风的 BLOG
J
Java Code Geeks
I
InfoQ
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
Martin Fowler
Martin Fowler
博客园 - 三生石上(FineUI控件)
Vercel News
Vercel News

TechPowerUp

Microsoft Visual Studio Professional 2026 + 15 Coding Courses Just $50 No Surprise Microsoft Office Hikes—Own it for Life for $50 The Witcher 3: Wild Hunt Reaches 65 Million Copies Ahead of Songs of the Past Launch Valve Steam Deck Sells Out in 24 Hours Despite Price Hike 007 First Light Exceeds 1.5 Million Sales in 24 Hours Rockstar Workers Unionize Ahead of GTA VI Launch Following Dismissal of 31 Workers Unknown Worlds Earns $250M Performance Bonus After Stellar Subnautica 2 Launch Dell Technologies Delivers First Quarter Fiscal 2027 Financial Results Marathon Season 2 To Start With Free Week and Plenty of New Content ASRock iBox Fanless Mini PCs Get Intel Panther Lake Upgrade Acer Broadens Portfolio with Two New Laptops Powered by the Latest Snapdragon Processors Qualcomm Introduces Snapdragon C Entry‑Level Processors for Budget Laptops OneXPlayer 3 Gaming Handheld Emerges With Intel Arc G3 Extreme Intel Arc G3 CPU Family Officially Released for Handheld Gaming PCs Samsung Exynos 2600 SoC Annotated, Showcases 3-tier CPU, AMD RDNA 4 iGPU Acer Expands Gaming Portfolio with Predator Atlas 8 Handheld Powered by Intel Silicon Motion Introduces SM2524XT PCIe Gen 5 DRAMless SSD Controller PXN Launches the Vector X Professional-Grade Sim Racing Pedals TP-Link Introduces Archer 8, Its First Wi-Fi 8 Router Platform Synology Announces Availability of New FlashStation FS200T Philips Announces Evnia 32M2N8900P QD-OLED 4K 240 Hz Gaming Monitor Samsung Display Develops First 4K 360 Hz QD-OLED Panel for Monitors LG Display Begins Mass Production of World's First 240 Hz RGB Stripe OLED ZALMAN Intros ZM-STC11 Silicone-based Thermal Paste Stream Deck Becomes the Action Layer for AI, Starting with NVIDIA G-Assist GIGABYTE Debuts New BRIX Mini PC Powered by Panther Lake to Scale Enterprise AI Sharkoon Announces the S25 Series Cases Scythe Intros the Magoroku Dual Fin-stack Air CPU Cooler First Look at ZOTAC's GeForce RTX 50-series 20th Anniversary Edition Graphics Cards ADATA TRUSTA AI Scaler Extended Memory Solution Breaks GPU Limits
Qualcomm HBC Gen 1 Achieves 133 TB/s Bandwidth by Stackin...
by AleksandarK Today, 16:47 Discuss (4 Comments) · 2026-06-25 · via TechPowerUp

Yesterday, Qualcomm announced its High Bandwidth Compute (HBC) product, a memory-compute solution hybrid designed to replace traditional High Bandwidth Memory (HBM) and offer greater performance, efficiency, and throughput. HBC utilizes LPDDR memory, stacking them across multiple layers in a 3D vertical space and interconnecting them with through-silicon vias (TSVs). This approach provides better energy efficiency than traditional HBM, which stacks regular DDR in vertical layers, as the low-power DDR (LPDDR) chips consume less power while delivering similar bandwidth and capacity. Qualcomm has successfully stacked multiple LPDDR layers on top of each other, with the base of the HBC being a compute die that performs near-memory computation and offloads some processing from the main processor.

This technique is similar to what HBM4 memory currently does, with the base die being a logic die to better integrate compute solutions underneath for packet tracing and data preparation for entering and exiting HBM. However, Qualcomm aims to achieve higher efficiency and more optimized data movement, with a 6x increase in bandwidth per watt compared to the current HBM specification, likely HBM4. HBC Gen 1 achieved 133 TB/s of bandwidth on the AI250 accelerator, a 18x increase over the LPDDR5X currently used in the AI200 card from Qualcomm. The company also claims there is a viable roadmap for using HBC across many Qualcomm AI accelerators, with HBC Gen 1 shipping with the AI250 AI accelerator in mid-2027. The next-generation HBC Gen 2 is expected to yield further significant gains in bandwidth.

Currently, it is unclear where Qualcomm plans to source the LPDDR memory, which LPDDR memory version is used in HBC, and where the company plans to package this. Qualcomm is known for manufacturing its Snapdragon SoCs at TSMC, but the compute die could be manufactured by any memory maker partnering for the HBC launch. The most logical choice is Samsung Foundry, which manufactures both LPDDR in-house and offers custom logic manufacturing. However, this has yet to be confirmed by Qualcomm.