Tagged articles
heterogeneous computing
65 articles · Page 1 of 1
This article explains the fundamental architectural differences between CPUs and GPUs, detailing why AI workloads rely on GPU parallelism, comparing core counts, memory bandwidth, and CUDA ecosystem dominance, while covering GPU types, key specifications, and future heterogeneous computing trends.
AICPUCUDA
0 likes · 10 min read

This article details Alibaba Cloud ACK's fault self-healing system for heterogeneous compute clusters, covering its Kubernetes-native architecture, multi-layer detection and event conversion, orchestrated repair workflows with stepwise authorization, rate limiting and circuit breaking, full observability, and practical deployment strategies to maximize effective compute throughput.
ACKGPU clustersKubernetes
0 likes · 64 min read

The article analyzes why heterogeneous CPU-accelerator architecture is essential for AI for Science (AI4S), detailing ten frontier applications on Sugon 8000's 100,000-card cluster that demonstrate superior performance in weather forecasting, molecular dynamics, quantum-supercomputing collaboration, and rocket plume simulation, overcoming traditional CPU-only bottlenecks.
AI4SHigh-Performance ComputingSugon 8000
0 likes · 15 min read

H3C's Wang Feng details a full-stack super-node architecture addressing AI infrastructure challenges with 128 GPUs per cabinet via full liquid cooling, 96% network utilization through optical-electrical parallel links, 200GB/s storage bandwidth, 30-day stable training at 100k nodes, three-layer security, and AI-driven operations cutting deployment time 60% and fault diagnosis 90%.
AI infrastructureAIOpsH3C
0 likes · 18 min read

Bank of America research projects the server CPU market will grow from $35B to $211B by 2030, with ARM architectures capturing 47% combined share as Intel falls to 22%, while AI accelerator deployments surge from 13 GW to 56 GW with Nvidia's share diluted by hyperscaler custom chips like Google TPU, AWS Trainium, and Microsoft Maia.
AI AcceleratorsARMBank of America
0 likes · 8 min read

The article argues that AI agents waste compute by using large language models for low-entropy decisions, and introduces Jev, a lightweight decision model that handles tool routing, code search, and log triage in milliseconds, enabling a System 1/System 2 architecture where Jev filters noise before Codex performs deep reasoning, a pattern mirrored by Glean's enterprise search stack.
AI agentsAgent ArchitectureCodex
0 likes · 16 min read

AMD’s upcoming Zen 6 family may introduce a low‑power Zen 6 LP core built by stitching proven Zen 5, Zen 4, and Zen 3 modules—retaining the full Zen 6 ISA while shrinking cache and FPU—to boost thin‑laptop and handheld battery life, though details remain unconfirmed.
AMDCPU architectureZen 6
0 likes · 6 min read

The article explains HaiGuang’s Agent‑to‑Token open computing architecture, detailing how a dual‑chip CPU + DCU design enables end‑to‑end AI agent workflows, token generation, and secure, high‑performance compute through heterogeneous collaboration, open interconnects, and a full software stack.
AI computeAgentCPU
0 likes · 12 min read

The article analyzes how ultra‑large AI compute clusters are evolving, highlighting ten major trends—from memory‑wall challenges and near‑memory computing to heterogeneous accelerators, mixed‑precision designs, diverse high‑bandwidth networking, optical interconnects, and the critical role of talent development—in order to understand the technical and ecosystem shifts driving next‑generation supercomputing.
AI TalentAI computehardware architecture
0 likes · 14 min read

As AI compute infrastructure evolves into a networked utility, the key challenge shifts from acquiring hardware to organizing distributed resources, data, models, and business needs through intelligent scheduling, governance, and cost-aware orchestration across edge, regional, and national layers.
AI DeploymentCost Optimizationcompute infrastructure
0 likes · 14 min read

The article analyzes how AI infrastructure competition is moving from single accelerators to rack‑level, full‑stack, heterogeneous cloud platforms, detailing AMD Helios integration, multi‑vendor strategies, workload‑driven design, and the implications for procurement, operations, and risk management.
AI infrastructureAMD Helioscloud AI
0 likes · 20 min read

The article details how 无问芯穹’s Agentic Infra strategy uses a cross‑cluster heterogeneous PD‑separation architecture (PDD) to cut first‑token latency by 51.5% and token cost by 37.5%, explains the bandwidth bottleneck of KV‑Cache transfer, introduces Decode‑side RadixCache and three‑stage handoff mechanisms, and shows a 37.5% BCR improvement that translates into roughly ten‑fold inference cost reduction.
Agentic InfraCross-ClusterLLM inference
0 likes · 25 min read

JD announced the donation of its Oxygen xLLM inference engine to the OpenAtom Open‑Source Foundation, detailing its service‑engine decoupled architecture, performance breakthroughs across e‑commerce, power and public‑safety workloads, and a roadmap to expand the open‑source AI ecosystem.
AI infrastructureEngineering IntelligenceLarge Model Inference
0 likes · 8 min read

On June 24, 2026 JD announced the donation of its Oxygen xLLM large‑model inference engine to the OpenAtom Open Source Foundation, detailing its service‑engine decoupled architecture, performance breakthroughs, heterogeneous chip support, and real‑world gains in e‑commerce, power‑grid and public‑safety applications while outlining a roadmap for broader ecosystem co‑building and standards leadership.
AI infrastructureEngineering IntelligenceLarge Model Inference
0 likes · 7 min read

The article analyzes how the immutable physical cost of each AI token and the exponential rise in inference demand outpace hardware improvements, driving a shift toward heterogeneous compute architectures, disaggregation, and ultimately an inference foundry model exemplified by NVIDIA's rapid acquisition of Groq.
AI inferenceFoundryGroq
0 likes · 26 min read

The article examines how hidden CPU‑GPU coordination issues can cripple AI inference performance, illustrates a real‑world XPU migration case where a kernel lock in the halolet component throttled throughput, and shows how Btune 2.0’s automated latency analysis and AI agent automatically pinpoint and resolve such bottlenecks.
AI performanceBtune 2.0CPU-GPU bottleneck
0 likes · 10 min read

The new consumer‑grade robot from VeilBlue delivers a thousand‑fold compute boost over previous models, matching Nvidia's Jetson AGX Thor while costing only one‑tenth, thanks to a six‑chip heterogeneous edge cluster, human‑surpassing perception, and safety‑first design validated in real homes.
AI hardwareEmbodied AIRobotics
0 likes · 14 min read

The article analyzes the severe inference bottlenecks of large language models, CNNs, and recommendation systems and presents a suite of research‑driven accelerations—including token‑level pipeline parallelism (HPipe), KV‑cache clustering (ClusterAttn), quantization (QoKV), heterogeneous edge frameworks (DeepZoning, PICO), delay‑aware edge‑cloud scheduling (DECC), and operator choreography (RACE)—validated on real‑world industrial workloads.
AI inferenceModel Accelerationedge AI
0 likes · 16 min read

The article argues that traditional storage‑centric computer science curricula are becoming obsolete as AI‑driven, compute‑centric paradigms dominate hardware, data‑center operations, and software ecosystems, urging universities and students to rapidly adopt new teaching focus and skills.
AI hardwareCUDAassociative memory
0 likes · 10 min read

This article explains the fundamentals of high‑performance computing (HPC), covering serial and parallel processing, the roles of CPUs and GPUs, system architectures, FLOPS metrics, current supercomputer capabilities, and the scale needed to reach the next exa‑FLOPS era.
CPUFLOPSGPU
0 likes · 7 min read

This article explains why migrating AI applications from Nvidia GPUs to domestic graphics cards is urgent, outlines the technical challenges, and introduces JoyScale’s zero‑perception migration stack that enables end‑to‑end hardware, software, and model adaptation for reliable, high‑performance AI deployment.
AI migrationJoyScaledomestic GPUs
0 likes · 11 min read

This article explains why migrating AI applications from Nvidia GPUs to domestic Chinese accelerators is urgent, outlines the technical challenges, and presents JD Cloud's JoyScale zero‑perception migration stack with hardware, software, model, and inference optimizations for real‑world scenarios.
AI migrationJoyScaledomestic GPUs
0 likes · 10 min read

NVLink Fusion, unveiled at Computex 2025, extends NVIDIA’s NVLink technology to enable high‑bandwidth, low‑latency connections between CPUs and GPUs or third‑party accelerators, offering up to 900 GB/s bandwidth, flexible heterogeneous configurations, ecosystem expansion, performance gains for AI training and inference, and potential cost reductions.
AICPUGPU
0 likes · 12 min read

This article provides an in‑depth technical analysis of Huawei’s Ascend AI accelerator architecture, detailing its heterogeneous compute units, memory hierarchy, task scheduling, programming model, compiler optimizations, and the capabilities of the Ascend 310 and 910 chips, while also discussing future challenges and market competition.
AI acceleratorAI hardwareHBM
0 likes · 14 min read

The article details Baidu Baige 4.0’s four‑layer AI infrastructure—hardware, cluster components, training‑inference acceleration, and platform tools—highlighting its heterogeneous computing, high‑performance networking, fault‑tolerant communication library, and optimizations that boost large‑model training and inference efficiency.
AI infrastructureheterogeneous computinghigh performance networking
0 likes · 17 min read

The article reviews NVIDIA's newly released H20 AI accelerator for China, compares its performance and pricing with domestic chips, outlines the expanding Chinese AI chip ecosystem—including Huawei, Cambricon, HaiGuang, Alibaba, ByteDance, and Baidu—while highlighting market size growth, multi‑chip heterogeneity strategies, and the strong demand forecast through 2024.
AI chipsAI computeChina
0 likes · 8 min read

The article examines how the surge of generative AI (AIGC) is fueling rapid growth in AI server demand, detailing the emerging AIGC ecosystem, server hardware composition, model scaling, heterogeneous computing, training vs. inference workloads, market size forecasts, and the competitive landscape of AI server manufacturers.
AI infrastructureAI serversGPU
0 likes · 15 min read

With AI training demands outgrowing single‑chip GPU clusters, this article explains how to construct and speed up heterogeneous AI clusters—combining GPUs, Kunlun, and Ascend chips—by addressing interconnect, distributed parallel strategies, and specialized acceleration suites to achieve high MFU and efficient large‑model training.
AI clusteringGPU accelerationMFU
0 likes · 15 min read

The article analyzes how the surge in AIGC and large language models is reshaping the AI server market, detailing hardware composition, the rise of heterogeneous computing, GPU advantages, demand calculations for models like GPT‑3, and the competitive landscape driving rapid industry growth.
AI serversAIGCGPU computing
0 likes · 16 min read

Senior big data engineer Zhu Yaogai from HuoLala shares the team’s three‑year journey, detailing background challenges, the construction of a multi‑layer big‑data infrastructure, solutions for cost efficiency, operational automation, heterogeneous computing, and future plans, illustrating how high cost‑effectiveness, operational efficiency, and analytical performance drive their evolution.
Data Infrastructureautomationcloud native
0 likes · 11 min read

The article details JD Retail Advertising's three‑stage evolution of its online model system—deep‑learning era, large‑model era, and power‑collaboration era—highlighting heterogeneous computing optimizations, platform and system capabilities, distributed graph computing, online learning, and dynamic power allocation to dramatically improve algorithm iteration speed and model performance.
AIAdvertisingdistributed graph
0 likes · 13 min read

The article analyzes the shift from x86 to ARM in server hardware, compares CISC and RISC instruction sets, reviews domestic Chinese CPU projects, examines GPU/NPU/FPGA accelerators, and proposes concrete ARM instruction‑set, compiler, and operating‑system standards for the telecom industry.
ARMCPU architectureCompiler
0 likes · 14 min read

An in‑depth overview of AI server components, market growth, AIGC‑driven demand, heterogeneous computing architectures, major vendors, and future trends, highlighting hardware composition, cost breakdown, competitive rankings, and the impact of GPU, CPU, and emerging AI accelerators on the industry.
AI serversCPUGPU
0 likes · 14 min read

In a keynote at Baidu Cloud Intelligence Conference, Baidu's VP Hou Zhenyu explains how the rise of large AI models is reshaping cloud infrastructure, spawning new architectures, services, and computing paradigms that converge mobile, deep learning, and cloud into an AI‑native era.
AI NativeMaaScloud computing
0 likes · 11 min read

The article explains the various AI‑capable chips such as CPUs, GPUs, FPGAs, NPUs, and TPUs, compares their performance and efficiency, describes heterogeneous CPU+xPU solutions, and provides market share data while highlighting the growing adoption of specialized AI accelerators.
AI accelerationAI chipsCPU
0 likes · 7 min read

The article provides a comprehensive analysis of the AI server market, detailing server hardware components, cost distribution, logical architecture, firmware, rapid market growth, competitive landscape, AI-driven heterogeneous computing, and future industry trends, while highlighting key vendors and deployment configurations.
AI serversGPUcloud providers
0 likes · 10 min read

Tencent Cloud’s Hybrid vCPU research, presented at KVM Forum 2023, outlines a three‑stage roadmap from homogeneous cores to mixed x86, ARM, and RISC‑V CPUs, detailing how virtualizing heterogeneous topologies, frequencies, caches, and PMU features can boost VM performance, security, live‑migration flexibility, and data‑center utilization.
Hybrid CPUKVMLive Migration
0 likes · 25 min read

This article presents a comprehensive overview of PaddlePaddle's automatic parallel design for heterogeneous scenarios, covering background motivation, architectural principles, key implementation details, practical usage interfaces, and future outlook, while illustrating concepts with detailed diagrams and examples.
AI FrameworksPaddlePaddleautomatic parallelism
0 likes · 19 min read

At the GTC conference, the talk presents Alibaba Cloud’s heterogeneous computing platform and introduces the Open Deep Learning API (ODLA), then details how CUTLASS‑based operator fusion dramatically accelerates attention and MLP layers in large‑scale recommendation models, achieving multi‑fold performance gains in production.
CutlassGPU computingdeep learning
0 likes · 5 min read


21CTOOct 1, 2022 · Cloud Computing
Intel’s recent announcements, featuring Linus Torvalds, the open‑source OneAPI ecosystem, SYCL‑based cross‑architecture tools, and the new Intel Developer Cloud, illustrate how the company is driving a unified, hardware‑agnostic platform for AI, accelerated computing, and secure cloud workloads.
AIDeveloper CloudIntel
0 likes · 10 min read

The article explains how massive parallel pixel calculations demand GPUs, whose high cost and inflexibility are solved by Tencent Cloud’s elastic, virtualized GPU services—including vGPU, qGPU, TACO abstraction, and spot instances—delivering up to 16 EFLOPS for AI, scientific, graphics, and video workloads.
GPU computingTencent Cloudcloud GPU
0 likes · 5 min read

This article surveys the rapid growth of data‑center capacity, the rise of AI and big‑data workloads, and how emerging accelerators such as GPUs, DPUs, SmartNICs and heterogeneous CPU designs from Intel, AMD, Arm and Apple are reshaping server hardware and driving a new wave of performance and efficiency competition.
AICPUGPU
0 likes · 12 min read

The article surveys GPU container virtualization for AI heterogeneous computing, detailing utilization challenges, historical architectures, various virtualization methods, Baidu's dual-engine user- and kernel-space design with isolation and scheduling features, performance benefits, best‑practice scenarios, and deployment guidance, concluding with a technical Q&A.
AI computingContainerizationGPU virtualization
0 likes · 30 min read

This article explains the fundamentals of high‑performance computing, covering serial and parallel processing, heterogeneous CPU‑GPU architectures, FLOPS measurement levels, key terminology, and why HPC is essential for scientific and engineering simulations, while also noting market reports and resource links.
FLOPSHPCHigh-Performance Computing
0 likes · 6 min read

This article thoroughly explains libunifex’s sender/receiver model for C++ asynchronous programming, covering its design goals, module structure, pipeline composition, key functions like schedule, then, sync_wait, and the connect/start mechanisms, while demonstrating practical examples and integration with C++20 coroutines and cancellation support.
Asynchronous ProgrammingC++Cancellation
0 likes · 16 min read

The article analyzes how the slowdown of Moore's Law drives a shift from general‑purpose CPUs to specialized XPU, FPGA, DSA and ASIC designs, highlights heterogeneous chiplet integration, edge‑server growth, and the emerging importance of software, algorithms and architecture in boosting performance and efficiency.
AIoTCPUChiplet
0 likes · 15 min read

This article introduces the fundamentals of heterogeneous computing on mobile GPUs, explains OpenCL concepts and its programming model, and provides a step‑by‑step example of adding two arrays with complete OpenCL code for Android devices.
AndroidC++GPU computing
0 likes · 9 min read

As mobile CPUs plateau, developers increasingly use OpenCL to harness Android GPUs like Qualcomm Adreno and Huawei Mali for heterogeneous computing, leveraging its platform, execution, and memory models to write portable kernels—illustrated by a simple array‑addition example that demonstrates device initialization, kernel creation, buffer management, and parallel execution.
AndroidGPU computingOpenCL
0 likes · 8 min read

The report analyzes how the rapid growth of cloud computing previously drove server demand, and how the emergence of 5G‑enabled edge computing is now reshaping server architectures, hardware needs, deployment models, and operational challenges, forecasting a significant increase in server volume for edge scenarios.
5GEdge Computingcloud computing
0 likes · 16 min read

The article summarizes Dr. Tang Shan's presentation on AI domain‑specific architectures, covering their background, the explosion of diverse AI hardware designs, and the significant software‑stack challenges that arise from fragmented tools and the need for full‑stack solutions.
AIDSAchip architecture
0 likes · 14 min read

The article analyzes how compute has evolved from performance‑driven Moore’s‑law scaling to a comprehensive innovation era driven by cloud democratization, heterogeneous accelerators, edge and memory‑centric architectures, outlining IDC’s ten future‑compute characteristics such as flexible deployment, AI‑enabled operations, multi‑cloud, security and next‑gen interconnects.
Edge Computingheterogeneous computing
0 likes · 17 min read

Baidu’s elastic nearline computing architecture inserts an asynchronous, resource‑adaptive layer between online and offline processing, dynamically harvesting idle CPU, GPU and Kunlun XPU capacity to pre‑compute complex recommendation and search policies, enabling peak‑shifting, valley‑filling, higher timeliness and significant business growth at low cost.
PaaS resource schedulingPeak Shavingcloud computing
0 likes · 18 min read

The article examines the rapid growth of data‑center workloads, explains why heterogeneous accelerators such as CPUs, GPUs, FPGAs and ASICs are needed, outlines evaluation standards, compares their compute performance and power efficiency, and discusses practical deployment cases and future trends.
ASICCPUFPGA
0 likes · 22 min read

The article explains how the fourth paradigm's distributed machine learning framework GDBT tackles the massive data, high‑dimensional features, and real‑time requirements of modern recommendation systems by leveraging heterogeneous computing, parameter servers, RDMA networking, and optimized workloads.
GDBTRDMAdistributed ML
0 likes · 18 min read

The article provides a comprehensive overview of heterogeneous computing, detailing its definition, real‑world system examples, performance advantages, key programming frameworks such as OpenCL and CUDA, industry trends like SOC integration, and a comparative analysis of CPUs, GPUs, FPGAs and ASICs.
CPUCUDAFPGA
0 likes · 9 min read

This article explains how Alibaba Cloud builds a high‑performance AI infrastructure by combining advanced hardware such as Shenlong servers, GPUs, FPGAs, NPUs, and custom interconnects like RDMA, together with virtualization, FPGA‑as‑a‑Service, AIACC, and resource‑pooling technologies to deliver scalable, cost‑effective AI services.
AI hardwareAlibaba CloudFPGA as a Service
0 likes · 20 min read

This article analyzes how system architects can select the optimal heterogeneous compute cores—CPU, GPU, FPGA, ASIC, VPU, or TPU—for edge AI deployments, weighing performance, size, weight, power, and cost to maximize inference efficiency and security.
AI edge computingASICCPU
0 likes · 7 min read

This article explains heterogeneous computing and compares major processing units—CPU, GPU, FPGA, ASIC, and NPU—highlighting their architectures, strengths, and typical use cases, especially in deep‑learning and AI workloads.
ASICCPUFPGA
0 likes · 10 min read

The article examines the rapid growth of information system energy consumption, defines green computing, outlines its research focus—especially on data centers, cloud computing and servers—and analyzes a range of energy‑saving technologies such as DVFS, heterogeneous computing, liquid cooling, rack‑level servers, advanced power supply, backplane cooling and fluorocarbon pump AC systems.
DVFSenergy efficiencygreen computing
0 likes · 13 min read

The article explains why 5G forces multimedia standardization, outlines the challenges of emerging formats like 8K, details the VVC (H.266) compression breakthrough, and describes Tencent's role in shaping future heterogeneous‑computing‑based multimedia architectures for cloud gaming and beyond.
5G8KMultimedia Standards
0 likes · 13 min read

The article explains how heterogeneous parallel computing—distributing tasks across CPUs, GPUs, FPGAs and other accelerators—has become essential after Moore’s law plateau, detailing its principles, hardware and software perspectives, classification of architectures, processing stages, user‑guided versus compiler‑guided methods, and its relevance to AI, cloud and industry workloads.
CPUFPGAGPU
0 likes · 15 min read

The article explains how heterogeneous parallel computing—leveraging CPUs, GPUs, FPGAs and other specialized units—addresses the performance limits of traditional serial programming by distributing tasks across diverse hardware, detailing its concepts, architectures, development models, and relevance to AI and cloud workloads.
CPUFPGAGPU
0 likes · 9 min read

At the 2017 International Workshop on Mathematical Issues in Information Sciences, Alibaba researcher Zhang Weifeng presented a talk on leveraging heterogeneous computing—re‑architected processors, memory‑wall mitigation, and integrated software‑hardware optimization—to dramatically improve machine‑learning performance, highlighting the growing importance of compute resources alongside algorithmic advances.
AIAcceleratorsheterogeneous computing
0 likes · 5 min read

The article explains why general‑purpose CPUs can no longer meet deep‑learning demands due to intrinsic scaling limits and memory‑bandwidth bottlenecks, and surveys how heterogeneous accelerators—GPUs, FPGAs, ASICs and emerging AI processors with high‑bandwidth memory—provide specialized, high‑parallelism, power‑efficient solutions for both cloud and edge workloads.
AI ProcessorsASICCPU
0 likes · 11 min read

The article analyzes the limitations of general‑purpose CPUs for deep‑learning workloads, explains how semiconductor scaling and memory‑bandwidth constraints drive the shift toward specialized heterogeneous processors such as GPUs, FPGAs, and ASICs, and discusses the design trade‑offs of embedded versus cloud AI accelerators.
AIASICCPU
0 likes · 13 min read
