
Cerebras Announces Capability to Train Largest Models Easily
$1,500.00
Authors: Alex Norton and Thomas Sorensen
Publication Date: September 2022
Length: 1 pages
In mid-June of 2022, Cerebras Systems announced a new feature that allows users to train some of the largest AI models in the world within a single CS-2 machine using a simplified software support scheme. The announcement highlights multiple capabilities that Cerebras sees as their competitive advantages over other companies. Notable examples cited include the ability to accommodate an entire training model within the memory, through Cerebras’ Weight Streaming software on the Wafer Scale Engine (WSE), instead of splitting it across processors, as well as the ability for users to manipulate a few inputs within the software scheme and GUI to choose the scale of model desired for training (i.e., GPT-3 13B, GPT-3XL 1.3B). Cerebras claims that this advancement can cut down the setup of large model training runs from months to minutes, with the Cerebras software managing much of the initial setup.
Related Products
EU Commits Over €100 Million to Deploy Six Quantum Computers in 2023
Bob Sorensen & Tom Sorensen
The European High-Performance Computing Joint Undertaking (EuroHPC JU) recently announced the selection of six sites to host what could be the first major round of EU government-sponsored quantum computer (QC) procurements. The new QC systems will be integrated into leading EU-based HPC sites including the IT4Innovations National Supercomputing Centre in Czechia, the Barcelona Supercomputer Center in Spain, and Cineca in Italy, as well as key HPC sites in France, Germany, and Poland. The QC systems will be networked to support EU-wide access by academic, commercial and government research facilities. Half of the 2023 €100 million procurement budget will come from the EU and the remainder from the 17 countries participating in the EuroHPC JU. The QC hardware and software for this effort will draw exclusively on EU technology developed under EU-funded quantum initiatives, related national programs, and private investments.
December 2022 | Uncategorized
NOAA and Microsoft Announce Cloud Computing Collaboration to Advance Climate-Ready Nation Mission
Jaclyn Ludema and Mark Nossokoff
The US National Oceanic and Atmospheric Administration (NOAA) and Microsoft have entered into a Cooperative Research and Development Agreement (CRADA), formalizing NOAA's commitment to using Microsoft Azure cloud computing resources in the pursuit of NOAA's mission to build a Climate[1]Ready Nation by 2030. Several initiatives are envisioned whereby NOAA scientists and engineers will work with Microsoft experts to leverage Azure's machine learning and HPC capabilities: ▪ Fast-tracking innovative contributions to NOAA Earth Prediction Innovation Center (EPIC) earth systems modeling and research ▪ Applying machine learning capabilities to improve models supporting air quality, smoke, and particulate pollution forecasts, as well as relevant NOAA climate models ▪ Accelerating NOAA Fisheries' survey and observations data collection and management ▪ Creating new ocean observations cataloging efforts ▪ Designing resilient and accessible weather modeling and forecasting that can incorporate external data sources with NOAA enterprise data
December 2022 | Uncategorized