HPC Architectures

HPC Rankings & Benchmarks:

Report on the Fujitsu Fugaku system

Fujitsu achieves over 1,000 petaflops on new HPL-AI benchmark

Summit achieves 445 petaflops on new 'HPL-AI' benchmark

K computer leads Graph500 ranking for ninth consecutive time

June 2020 Topo500 list    June 2020 Green500 list

Shadow over June 2019 Top500 list    Top500 and Green500 merge

TOP500 Supercomputer Sites November 2019    HPCG Nov 2019    HPC Challenge Benchmark Nov 2018 

The Green500 List Nov 2019    The Graph 500  BFS List June 2019    The Graph 500 SSSP List June 2019

LINPACK's 'companion metric' gains traction    SPEC releases major new CPU benchmark suite

Deep500 - researchers tackle an HPC benchmark for Deep Learning

Fujitsu-developed deep learning technology sets record on ABCI

How do you measure a supercomputer's speed?    Linpack, exascale & the Top500

Supercomputer slowdown: world's fastest system sees no challengers

TOP500 reanalysis shows 'nothing wrong with Moore's Law'

SGI's Pangea recognized as industry's top commercial supercomputer on TOP500 list

Top problems with the TOP500   Council on Competitiveness defends TOP500 usefulness

TOP500: the missing puzzle pieces    Architectual surprises underpin new HPC benchmark results

The new realism: software runs slowly on supercomputers

What's the true hourly cost of outsourcing

HPC Lists we'd like to see    Too big to FLOP?

Sandia-developed benchmark re-ranks top computers

A conversation with ISC Fellow Jack Dongarra

Toward a new metric for ranking high performance computing systems

Focus sharpens on new HPC benchmark     New HPC benchmark delivers promising results

A path to energy efficient HPC data centers

Tackling the power and energy wall for future HPC systems

How can supercomputers survive a drought?

IEEE group seeks to reinvent computing as scaling stalls

IDC says Lenovo strengthens, IBM stumbles in Q1 2015 ranking

ACM Gordon Bell Prize recognizes top accomplishments in running science apps on HPC

STREAM benchmark author McCalpin traces system balance trends

Searching for dark energy in the HPC universe    Tackling the co-design puzzle - new thinking is needed

HPC iron, soft, data, peaople - it takes an ecosystem

How ASCI revolutionized the world of HPC and advanced modeling & simulation

Towards Exascale:

The national HPC race

Pondering the future of HPC architectures

The exascale computing race isn't about bragging rights

Exascale computing: the space race of our time

The path to an energy-efficient exascale supercomputer

Exascale processor proliferation? How many is too many?

The accelerator wall: a new problem for a post-Moore's Law world

Exascale in Europe    Ecosystem for future European exascale supercomputer    

EU funds 20M euro ARM + FPGA exascale project

China maintains lead in race to exascale    China establishes seventh National Supercomputing Center

Analysis of the characteristics and development trends of next-gen supercomputers (outside of Japan)

Exascale: a race to the future of HPC    Doug Kothe on the race to build exascale applications

US Exascale Computing Project releases software technology project report 

Will the real exascale race please stand up    HPC architecture in the 2020's     

Gordon Bell reflects on race to build fastest computer

Moving just to exascale or preparing for sustained scaling

Extreme-scale requirements to push the frontiers of deep learning

IEEE Conference Keynoters Lay Out Path to Exascale Computing

Exascale Computing   On the Road to Exascale   The Bumpy Road to Exascale

Tackling power and resiience at exascale    Paving the roadmap to exascale    The exascale revolution

World-leading HPC centers partner to form accelerated computing institute

Advancing modular supercomputing with DEEP and DEEP-ER architectures

Improving energy efficiency and exploiting parallelism with Processor-in-Memory and Near-Data Processing
China sets ambitious goal to reach exascale by 2020    Sugon announces roadmap to exascale computing

China Flexes High-Tech Muscles    China wins new bragging rights in supercomputers

India plots 3-ohase indigenous supercomputing starategy

New Japanese supercomputing project targets exascale

US moves exascale goalposts, targets 2021 delivery   

US supercomputing leaders tackle the China question   

The vital importance of HPC to US competitiveness    IBM promises a 200 Petaflop supercomputer by 2018

Top scientist urges 'ambitious' U.S. exascale supercomputer plan    US DoE exascale ECP program update

Six exascale PathForward vendors selected: DoE providing $258M

DOE exascale plan gets support with caveats    DoE to fund exascale resilience research

DOE primes pump for exascale supercomputers    DOE commissions extreme computing study

New US national HPC strategy is bold, important and more daunting than the US moonshot

Exascale computing - the view from Argonne   Argonne's Peter Beckman on the Race to Exascale    

Inside Japan's future exascale ARM supercomputer    Japan's post-K computer hits 1-2 year speed bump

Post-K supercomputer named after Japan's tallest peak

European Project Will Design Energy-Efficient Exascale Machine

EU projects unite on heterogenous ARM-based exascale prototype    Latest ARM-HPC developments

European consortium to develop exascale architecture  ExaNeSt: 10M computers into a single supercomputer

Leap to the extreme scale could break science boundaries       

Charming exascale power problems    Exascale: Power Is Not the Problem!   Powering Up Exascale

HPC evolves from the commodity cluster    

Can our computers continue to get smaller and more powerful?

Trends in HPC and Machine Learning drive evolution of cooling solutions 

Solving the Sky: new computers resolve clouds while keeping cool

Cray, Fujitsu both bringing Fujitsu AFX64-based supercomputers to market in 2020

Intel and Cray to build the next generation US supercomputer    Intel: the future of HPC has arrived

Intel launches Cascade Lake Xeons with up to 56 cores 

Intel Skylake: Xeon goes from chip to platform 

Momentum builds for Intel's HPC Scalable System Framework

IBM first to 7nm process with working transistors

IBM's new open Linux Power servers take on x86 in HPC

IBM's first OpenPOWER server targets HPC workloads

IBM begins Power9 rollout with backing frim DOE, Google

AMD's exscale strategy hinges on heterogeneity

AMD expands exascale vision at IEEE HPC symposium

Qualcomm takes first step in uphill battle  with Intel over server market

KNUPATH Hermosa chip expected in first half of 2017

Fujitsu launches M12 servers; emphasizes commitment to Sparc

Fujitsu presents post-K CPU specs

Oracle layoffs reportedly hit SPARC and Solaris hard

Number Crunching, Data Crunching and Energy Efficiency: the HPC Hat Trick

Revisiting Supercomputer Architectures   End-users need to design exascale computers

Trends in High-Performance Computing   Future challenges within HPC   Facing the future    

Supercomputers Let Up on Speed     A New Generation of Smarter, Not Faster, Supercomputers

IEEE members on suoercomputing predictions, challenges    Will we ever reach Zettaflops? 

HPC: A Key Technology for Europe   How Cool are Supercomputers?     Petaflop in a box  

Working towards higher performance    Taking a Disruptive Approach to Exascale    

MPI milestone achieved on the path to  exascale

Engineering codes to meet the exascale era       Cray Chapel new computing language

UCLA engineers develop new energy-efficient computer memory using magnetic materials

Exascale Research: Preparing for the Post-Moore Era     IT in the next decade

HPC programming in the age of multicore: one man's view    

Hybrid Computing (CPUs/GPUs/FPGAs):

Why 2016 is the most important year in HPC in over two decades    

Modular HPC goes mainstream    PRACE Supercomuting MOOC

Accelerating the development of HPC    Gearing up for the cluster of tomorrow

Heterogeneous: the benefits to performance and power consumption

Better, faster systems with heterogenous designs    Accelerated computing: a tipping point for HPC

Will 2015 be the year of the FPGA?    Revving up HPC workloads with FPGAs  

A blending of strengths    The rise of GPU compute   

Redefining the role of the CPU in the era of CPU-GPU integration

When dense matrix representations beat sparse    Densifying assumed-sparse tensors

Stampede foreshadows heterogeneous supercomputing

Are supercomputing's elite turning backs on accelerators? 

Massively multithreaded computer systems  

NCSA Director: GPU is the future of supercomputing   The GPU Computing Era

GPUs versus CPUs: Apples and Oranges?   Top 10 Objections to GPU Computing Reconsidered

Supercomputing with Tesla GPUs   Cray Unveils Its First GPU Supercomputer  

From Multithreaded Programming to GPU Computing    The heterogenous programming jungle

Performance Analysis of a Hybrid MPI/CUDA Implementation of the NAS-LU Benchmark

Accelerating Computational Science Symposium 2012    Making sense of when to use FPGAs

Overcoming hurdles to parallelisation   GPUs and the future of Parallel Computing

NVIDIA Eyes Post-CUDA Era of GPU Computing   NVIDIA Opens Up CUDA Compiler

NVidia unveils 1.3 teraflop GPU for supercomputing   NVidia steers roadmap around GPU bottlenecks

Summary of 2015 NVidia GPU growth

AMD refreshes roadmap, transitions back to HPC

GlobalFoundaries: 7 nm chips coming in 2018, EUV in 2019

Intel will ship Knights Corner chip in 2012    Intel fires the next salvo in supercomputer battle

AMD announces FLOP-Monster GPU card

Intel Atom architecture in a handheld form factor: Medfield

OpenCL gains ground on CUDA    A Tale of Two GPU Computing Models   Computer games help HPC

HPC debugging redefined    

Ten Ways to Fool the Masses When Giving Performance Results on GPUs

Engineers Boost Computer Processor Performance By Over 20 Percent

NVidia pokes holes in Intel's Manycore story

AMD: The Integration Revolution    GPUs add up for ARM chips in HPC

Commodity memory for a high performance world    Energy-efficient supercomputing

A comparison of heterogeneous and manycore programming models

Why CTOs should reconsider FPGAs

Manycore/Multicore Computing:

Parallelism has crossed a threshold    World's first 1,000 processor chip 

Toward a converged exascale-big data software stack

Adapteva unveills 64-core chip    Smarter multicore chips     

Intel brings manycore X86 to market with Knights Corner

Intel unveils 72-core X86 Knights Landing CPU for exascale computing

HP scientists envision 10-teraflop many-core chip   

Fujitsu targets 100 petaflop supercomputing

China makes its own supercomputing cores

We need more than Multi-core  Chip simulation for the 1000-core era 

Multicore Is Bad News For Supercomputers

Qualcomm targets Intel datacenter dominance with 10nm ARM-based server chip

ARM unveils Neoverse N1 platform with up to 128-cores

ARM Processors Set to Challenge x86 On Its Own Turf   

ARM Discloses Technical Details of 64-bit Architecture 

Enabling the ARM ecosysteem for HPC    Bolstering the ARM case for HPC workloads

Atos previews exascale-focused ARM-based Bull Sequena X1310

Migrating high performance applications to the Armv8-A architecture 

ARM doubles down on AI

Arm Yourselves for Exascale, Part 1   Arm yourselves for Exascale, Part 2

Russia launches ARM chip effort to end X86 dependence

China's expanding efforts to win in microchips

DARPA continues investment in post Moore's technologies

A new breed of heterogenous computing 

NVIDIA Tegra Processors Blaze the Way for ARM in Supercomputing

Attack of the Killer Micros Redux

Remapping Computer Circuitry to Avert Impending Bottlenecks

Studying the Impact of Hardware Prefetching and Bandwidth Partitioning in Chip-Multiprocessors

A new golden age for computer architecture

Computing Models for FPGA-Accelerators

Moving Scientific Codes to MultiCore Microprocessor CPUs   Core Scalability in Volume Rendering

Playing with NAG’s Library for SMP & multicore (Or parallel programming made easy)p

IOLanes: Advancing the Scalability and Performance of I/O Subsystems in Multicore Platforms

IBM Specs Out Blue Gene/Q Chip   Processor Whispers: About 16 and 17-core processors

IBM launches POWER8 Systems, OpenPOWER roadmap

IBM Power8 Systems   OpenPower systems coming in mid-2015

OpenPower9 ecosystem gains steam, new customers and more

Lenovo aims high for systems expansion      Lenovo gets serious about HPC

Intel Scales Up Cores and Memory with New Westmere EX CPUs  

Intel Adds New Dimension to Transistor Making  

Intel Xeon E5 Debuts on TOP500; First Intel MIC Co-processor Hits 1 Teraflop

Adapteva Builds Manycore Processor That Will Deliver 70 Gigaflops/Watt

Tilera Releases 16-Core and 36-Core Processors for 64-bit Computing

Cluster Numbers: An Automated Benchmarking Tool for HPC Clusters

China Preps Godson Chip for Supercomputing Duty   China mulls national CPU architecture spec

China could become supercomputing partner   China's Dark Horse Supercomputing Chip: FeiTeng

China reports breakthrough in developing 3nm transistors 

Homemade CPUs on the way for local supercomputers

Designer of Microprocessor-Memory Chip Aims to Topple Memory and Power Walls  

Application Specific Processors:
MPI versus OpenMP:
Shared-Memory & In-Memory Computing:
New Memory Technologies:    

Vectors, Mixed Precision:
Spaceborn Computing: