SearcharxivSearch

arXiv subjects

Alberto Moreno

Publications and source records attributed to Alberto Moreno.

2 recordsLinked to original sources

EPAC: The Last Dance

This paper presents EPAC, a RISC-V-based accelerator chip developed within the European Processor Initiative (EPI) as part of a multi-year, multi-partner effort to build a European HPC processor ecosystem. EPAC is implemented in GlobalFoundries 22FDX (GF22FDX) technology, covers an area of 27 sq mm with approximately 0.3 billion transistors, and integrates three distinct RISC-V compute tiles targeting different workload classes: VEC, a vector processing tile for double-precision HPC workloads; STX, a many-core tile optimized for stencil and machine learning computations; and VRP, a variable-precision tile for iterative numerical solvers requiring extended floating-point formats. All tiles are connected through a Coherent Hub Interface (CHI) based network-on-chip with a distributed L2 cache system and communicate with external memory via a SerDes link. The chip was taped out in GF22FDX technology and successfully brought up, with all major IP blocks validated. This paper describes the architecture of each tile and the uncore infrastructure, the integration and physical implementation process, and the board-level bring-up activities. It also reflects on the engineering and coordination lessons learned from a full chip design effort distributed across academic and industrial partners in Europe.

cs.AR

Challenges in assessing Fire Weather changes in a warming climate

The Canadian Fire Weather Index (FWI) is widely used to assess wildfire danger and relies on meteorological data at local noon. However, climate models often provide only daily aggregated data, which poses a challenge for accurate FWI calculations in climate change studies. Here, we examine how using daily approximations for FWI95d -- the annual count of extreme fire weather days exceeding the 95th percentile of local daily FWI values -- compares to the standard noon-based approach for the period 1980--2023. Our findings reveal that FWI95d calculated with noon-specific data increased globally by approximately 65\%, corresponding to 11.66 additional extreme fire weather days over 44 years. In contrast, daily approximations tend to overestimate these trends by 5--10\%, with combinations involving minimum relative humidity showing the largest divergences. Globally, up to 15 million km$^2$, particularly in the western United States, southern Africa, and parts of Asia, exhibit significant overestimations. Among our daily approximation methods, the least biased proxy is the one that uses daily mean data for all variables. We recommend (i) prioritizing the inclusion of sub-daily meteorological data in future climate model intercomparison projects to enhance FWI accuracy, and (ii) adopting daily mean approximations as the least-biased alternative if noon-specific data are unavailable.

physics.ao-ph