Concurrent computing
Executing several computations during overlapping time periods
Follow Concurrent computing on Notably News to receive short updates to your email — rarely!
We include updates on Supercomputer, Actor model, Coroutine, Prefix sum, Data transfer object, Hyper-threading, Computer cluster, Simultaneous multithreading, Cache coherence, Cooperative multitasking, Pthreads, Stackless Python, ROCm, DLL injection, Shared memory, ZeroMQ ... and more.
| 2024 |
Virtual thread
Java resolved the previous limitation of virtual threads being pinned to OS threads in Java 24, allowing more flexible threading capabilities.
|
| 2024 |
Supercomputer
El Capitan becomes the world's top supercomputer, located at Lawrence Livermore National Laboratory, with 1,051,392 CPU cores and 9,988,224 GPU cores, utilizing HPE Cray EX255a architecture and AMD technologies
|
| 2024 |
Supercomputer
Alps supercomputer at CSCS Swiss National Supercomputing Centre becomes operational, featuring 748,800 ARM Neoverse V2 cores and 1,372,800 Nvidia Hopper H100 GPU cores
|
| December 10 2024 |
Actor model
Release of Goblins-Racket and Goblins-Guile, Actor frameworks for Racket and Guile Scheme under Apache 2.0 license
|
| April 30 2024 |
Actor model
Release of Cloud Haskell, an Actor framework for Haskell under BSD license
|
| March 4 2024 |
Actor model
Release of Abstractor, an Actor library for Java under Apache 2.0 license
|
| January 2 2024 |
Actor model
Release of Otavia, an Actor library for Scala under Apache 2.0 license
|
| 2023 |
Virtual thread
Java introduced virtual threads with Java 21, initially with a limitation that code using synchronized blocks or native calls would be pinned to its carrier OS thread.
|
| 2023 |
Supercomputer
Eagle supercomputer by Microsoft becomes fourth most powerful, using NDv5 architecture with 2,073,600 total cores and Ubuntu 22.04 LTS
|
| 2023 |
Supercomputer
Aurora supercomputer at Argonne National Laboratory ranks third, with 1,104,896 CPU cores and 8,159,232 GPU cores using Intel Xeon Max processors
|
| 2022 |
Supercomputer
Exascale supercomputers capable of performing over 1018 FLOPS were developed, representing a significant leap in computational power.
|
| 2022 |
Supercomputer
Frontier supercomputer at Oak Ridge National Laboratory ranks second globally, featuring 614,656 CPU cores and 8,451,520 GPU cores with HPE Cray EX235a system
|
| 2021 |
Virtual thread
Consumer grade computers typically offered a parallelism of tens of concurrent execution units, highlighting the context for virtual thread development and performance optimization.
|
| 2021 |
Virtual thread
Consumer-grade computers demonstrated the capability to support tens of millions of preemptive tasks and events using virtual threads, significantly improving computational efficiency compared to traditional operating system threads.
|
| 2021 |
Structured concurrency
Swift programming language adopts structured concurrency as a core feature.
|
| 2021 |
Structured concurrency
A draft proposal is published to add structured concurrency to Java, signaling potential widespread adoption.
|
| 2020 |
Supercomputer
Fugaku supercomputer at Riken Center for Computational Science remains significant, with 7,630,848 Fujitsu A64FX CPU cores using Tofu interconnect D
|
| December 8 2020 |
Intel Parallel Studio
Intel replaced Intel Parallel Studio XE with oneAPI toolkits, marking a significant shift in their development tool ecosystem.
|
| November 2020 |
ROCm
Navi 2X GPU series released, introducing RDNA 2 microarchitecture
|
| January 2020 |
Supercomputer
Supercomputers used to simulate and find potential compounds to stop COVID-19 spread, running extensive parallel CPU simulations.
|
| 2019 |
Hyper-threading
Security experts broadly recommended disabling hyper-threading on all devices due to a set of discovered vulnerabilities.
|
| 2019 |
Hyper-threading
Intel temporarily removed hyper-threading from mainstream Core i7 desktop processors, keeping it only in high-end Core i9 parts and Pentium Gold CPUs. Intel also began recommending disabling hyper-threading due to new CPU vulnerability attacks.
|
| December 16 2019 |
Intel Parallel Studio
Intel released Intel Parallel Studio XE 2020, their latest version before transitioning to a new platform.
|
| July 2019 |
ROCm
Navi 1X GPU series released, marking the first generation of RDNA microarchitecture
|
| 2018 |
Virtual thread
Go programming language introduced virtual threads support with version 1.18, enabling the creation of up to 15 million virtual threads on a consumer-grade computer with approximately 350,000 virtual threads per gigabyte of main memory.
|
| 2018 |
Supercomputer
Lenovo achieved a significant milestone by becoming the world's largest provider of supercomputers, producing 117 units in the TOP500 list.
|
| 2018 |
Concurrent hash table
Mega-KV key-value store system achieved a record throughput of up to 888 million key-value operations per second, with optimizations from Nvidia and Oak Ridge National Lab using GPU acceleration and cuckoo hashing.
|
| 2018 |
Hyper-threading
OpenBSD operating system disabled hyper-threading to prevent potential data leakage caused by Foreshadow/L1TF vulnerabilities.
|
| 2018 |
Hyper-threading
ARM produced simultaneous multithreading (SMT) cores for the first time.
|
| September 12 2018 |
Intel Parallel Studio
Intel released Intel Parallel Studio XE 2019, keeping pace with their annual release strategy.
|
| June 2018 |
Supercomputer
All combined supercomputers on the TOP500 list collectively broke the 1 exaFLOPS performance mark for the first time.
|
| 2017 |
Hyper-threading
A bug was revealed in Intel's Skylake and Kaby Lake processors that could cause data loss in their hyper-threading implementation. Microcode updates were later released to address the issue.
|
| 2017 |
Structured concurrency
Nathaniel J. Smith refines structured concurrency by introducing the 'nursery pattern' in his Python implementation called Trio.
|
| November 2017 |
Supercomputer
All of the world's fastest 500 supercomputers began running on Linux-based operating systems.
|
| September 12 2017 |
Intel Parallel Studio
Intel released Intel Parallel Studio XE 2018, maintaining their software development tool progression.
|
| June 2017 |
ROCm
Vega GPU series released, introducing GCN 5th generation microarchitecture and representing the last GCN generation before RDNA
|
| 2016 |
Structured concurrency
Martin Sústrik formulates the concept of structured concurrency by developing the C library libdill, using goroutines as a starting point.
|
| 2016 |
ROCm
AMD launched ROCm as part of the Boltzmann Initiative, building upon previous AMD GPU software stacks like Close to Metal (CTM) and Stream.
|
| September 6 2016 |
Intel Parallel Studio
Intel released Intel Parallel Studio XE 2017, continuing their annual update cycle.
|
| June 2016 |
ROCm
Arctic Islands/Polaris GPU series released, utilizing GCN 4th generation microarchitecture
|
| January 12 2016 |
Vert.x
Tim Fox stepped down as the lead of the Vert.x project, with Julien Viet, a long-time contributor, taking his place.
|
| August 25 2015 |
Intel Parallel Studio
Intel released Intel Parallel Studio XE 2016, further evolving their development tools.
|
| June 2015 |
ROCm
Volcanic Islands GPU series released, featuring GCN 3rd generation microarchitecture
|
| August 26 2014 |
Intel Parallel Studio
Intel released Intel Parallel Studio XE 2015, advancing their parallel computing software suite.
|
| May 2014 |
Vert.x
Vert.x won the 'Most Innovative Java Technology' award at the JAX Innovation awards.
|
| 2013 |
Hyper-threading
Intel dropped simultaneous multithreading (SMT) in the Silvermont processor cores, preferring out-of-order execution for better performance and power efficiency.
|
| September 2013 |
ROCm
Sea Islands GPU series released, representing GCN 2nd generation microarchitecture
|
|
We are only showing the most recent entries for this topic. |
|
This contents of the box above is based on material from the Wikipedia articles Supercomputer, Light Weight Kernel Threads, Concurrent hash table, JCSP, Vert.x, Actor model, ROCm, Structured concurrency, Intel Parallel Studio, Coroutine, Hyper-threading & Virtual thread, which are released under the Creative Commons Attribution-ShareAlike 4.0 International License.