Search

Abstract As the exascale era approaches, the increasing capacity of high-performance computing (HPC) systems with targeted power and energy budget goals introduces significant challenges in reliability. Silent data corruptions (SDCs), or silent errors, are one of the major sources that corrupt the execution results of HPC applications without being detected. In this work, we explore a set of novel SDC detectors – by leveraging epsilon-insensitive support vector machine regression – to detect SDCs that occur in HPC applications. The key contributions are threefold. (1) Our exploration takes temporal, spatial, and spatiotemporal features into account and analyzes different detectors based on different features. (2) We provide an in-depth study on the detection ability and performance with different parameters, and we optimize the detection range carefully. (3) Experiments with eight real-world HPC applications show that support-vector-machine-based detectors can achieve detection sensitivity (i.e., recall) up to 99% yet suffer a less than 1% false positive rate for most cases. Our detectors incur low performance overhead, 5% on average, for all benchmarks studied in this work. [ABSTRACT FROM AUTHOR]

Cassava (Manihot esculenta Crantz) is a key crop for starch and biofuels production. This study focuses on the polysaccharide composition and saccharification efficiency in cassava pulp through genome-wide association studies (GWAS), targeting the improvement of root characteristics for industrial use. We analyzed 135 partially inbred lines population, performing monosaccharide composition and saccharification analyses to reveal substantial variability in storage root biomass. Among 33 traits examined, 128 significant SNPs were associated with 23 biomass traits, highlighting a complex genetic architecture. Saccharification potential varied from 39 to 95 nmol Glu mg−1 h−1, with high broad-sense heritability for saccharification and several monosaccharide traits, indicating a strong genetic control. Our findings revealed that cassava pulp comprises similar proportions of pectin, hemicellulose, and cellulose in all genotypes. Correlation analysis showed significant associations between cellulose content and saccharification, suggesting that enhancing these traits can improve bioconversion efficiency. Negative correlations with glucose and glucuronic acid in hemicellulose and pectin fractions imply these components may inhibit saccharification. We identified 118 candidate genes associated with 21 traits, with many involved in stress responses affecting cell wall composition. This study verified 12 key candidate genes through sequence and expression analysis, including MANES_07G081200 , a YTH domain-containing protein associated with saccharification. Several stress-response genes, such as MANES_04G118600 and MANES_09G174600 , were linked to monosaccharide traits, suggesting that adaptive stress pathways influence biomass characteristics. This study provides insights into the genetic determinants of cassava pulp's saccharification and polysaccharide composition, aiding breeding efforts to develop cassava varieties optimized for industrial applications. • GWAS identifies 128 significant SNPs linked to 23 cassava biomass traits based on cassava pulp. • High heritability found for saccharification and several monosaccharide traits. • Negative impact of glucose and glucuronic acid on saccharification of cassava pulp. • 118 candidate genes identified, many involved in stress responses affecting cell wall composition. [ABSTRACT FROM AUTHOR]

Valorizing rice straw could mitigate the detrimental health and environmental consequences of disposal through field burning. An essential step for achieving this is reducing straw recalcitrance to digestion by either enzymes or animals to facilitate uses as fertilizer, animal feed, or conversion into fuels and chemicals. In the present work, we developed and characterized a Philippine rice diversity panel to explore the chemical basis of biomass recalcitrance. We used high throughput phenotyping of straw samples from the panel to identify chemical compounds that confer recalcitrance. We determined the saccharification potential, silica, ferulic, and p-coumaric acid content in each rice accession, as well as the chemical fingerprint of biomass composition using pyrolysis followed by GC/MS. Multivariate analysis of the phenotypic data allowed us to characterize the relationship between biomass components and straw saccharification establishing that Si, ferulic acid and coumaric acid are inversely correlated with saccharification. PCA analysis showed that pyrolysis products derived from lignin constitute the largest proportion of compounds inversely correlated with saccharification. • A rice association panel was characterized for straw quality. • Pyr-GC/MS fingerprinting of straw shows the compositional diversity in the panel. • Silica, coumaric acid and ferulic acid are inversely correlated to saccharification. • We identified pyrolysis peaks associated with saccharification potential. [ABSTRACT FROM AUTHOR]

This laboratory scale study aims to demonstrate the effectiveness of thermochemical and biological saccharification of Miscanthus giganteus (MG) for generation of fermentable saccharides and its subsequent fermentation into solvents i.e. acetone, ethanol and butanol (ABE) using Clostridium acetobutylicum ATCC 824. Saccharide hydrolysates were derived from MG by thermochemical (water, acid and alkali at 130 °C) and biological saccharification (Fibrobacter succinogenes S85) processes and were subjected to batch fermentation for 120 h using C. acetobutylicum ATCC 824. At the end of fermentation of thermochemically-derived hydrolysates, 742 g m−3 of saccharides from water treatment, 9572 g m−3 of saccharides from acid treatment and 4054 g m−3 of saccharides from alkali treatment were fermented and yielded 0.045, 0.0069 and 0.01 g g−1 of total solvents, respectively. Similarly, at the end of fermentation of biological hydrolysate (using F. succinogenes), 2504 g m−3 of saccharides was fermented and yielded 0.091 g g−1 of total solvents. The highest yield of total solvents was achieved by water (thermochemical) and biological saccharification of MG using C. acetobutylicum. Whereas, acid and alkali-treated hydrolysates showed lower yields of solvents presumably due to production of inhibitory compounds during saccharification. Compared to thermochemical saccharification, biological saccharification using F. succinogenes is a promising approach since it yielded the highest amount of solvents whilst being eco-friendly. Our future studies will focus on optimisation of biological saccharification (using F. succinogenes) and sequential co-culture fermentation (using C. acetobutylicum). The development of alternative consolidated bioprocessing approach using biological saccharification will contribute towards making lignocellulosic biofuels a reality. • High lignocellulosic Acetone-Butanol-Ethanol (ABE) production cost, led us to explore simple/new saccharification methods. • Thermochemical and biological pre-treatment were used to obtain fermentable saccharides and fermented by C. acetobutylicum. • Highest yield of total solvents achieved by thermochemical and biological saccharification, was 0.045 g g−1 and 0.091 g g−1. • Biological pre-treatments seem promising, but further optimisation to make this commercially attractive is required. [ABSTRACT FROM AUTHOR]

Sugarcane bagasse represents one of the best potential feedstocks for the production of second generation bioethanol. The most efficient method to produce fermentable sugars is by enzymatic hydrolysis, assisted by thermochemical pretreatments. Previous research was focused on conventional heating pretreatment and the pretreated biomass residue characteristics. In this work, microwave energy is applied to facilitate sodium hydroxide (NaOH) and sulphuric acid (H 2 SO 4 ) pretreatments on sugarcane bagasse and the efficiency of sugar production was evaluated on the soluble sugars released during pretreatment. The results show that microwave assisted pretreatment was more efficient than conventional heating pretreatment and it gave rise to 4 times higher reducing sugar release by using 5.7 times less pretreatment time. It is highlighted that enrichment of xylose and glucose can be tuned by changing pretreatment media (NaOH/H 2 SO 4 ) and holding time. SEM study shows significant delignification effect of NaOH pretreatment, suggesting a possible improved enzymatic hydrolysis process. However, severe acid conditions should be avoided (long holding time or high acid concentration) under microwave heating conditions. It led to biomass carbonization, reducing sugar production and forming ‘humins’. Overall, in comparison with conventional pretreatment, microwave assisted pretreatment removed significant amount of hemicellulose and lignin and led to high amount of sugar production during pretreatment process, suggesting microwave heating pretreatment is an effective and efficient pretreatment method. [ABSTRACT FROM AUTHOR]

• Simultaneous optimisation of pretreatment and saccharification using design tool. • High-resolution Design of Experiments maximizes sugar release in sugarcane bagasse. • Fractional Factorial Design was combined with a Central Composite Orthogonal design. • Design of Experiments monitored by compositional analysis of biomass. To maximize the sugar release from sugarcane bagasse, a high-resolution Fractional Factorial Design (FFD) was combined with a Central Composite Orthogonal (CCO) design to simultaneously evaluate a wide range of variables for alkaline pretreatment (NaOH: 0.1–1 mol/L, temperature: 100–220 °C, and time: 20–80 min) and enzymatic saccharification (enzyme loading: 2.5–17.5%, and reaction volume: 550–850 µL). A total of 46 experimental conditions were evaluated and the maximum sugar yield (423 mg/g) was obtained after 18 h enzymatic hydrolysis under optimized conditions (0.25 mol/L NaOH at 202 °C for 40 min, with 12.5% of enzyme loading). Biomass compositional analyses showed that the pretreatments strongly removed lignin (up to 70%), silica (up to 80%) and promoted cellulose enrichment (25–110%). This robust design of experiments resulted in maximizing enzymatic hydrolysis efficiency of sugarcane bagasse and further indicated that this combined approach is versatile for other lignocellulosic biomasses. [ABSTRACT FROM AUTHOR]

• Methane yields of 5 selected CAM plants were equivalent to that of maize. • Agave angustifolia Haw. produced the highest methane yield of all CAM substrates. • Lower lignin in CAM plants as compared to maize resulted in a faster VFA production. • Rumen fluid could play an important role in the hydrolysis step. • CAM plants could represent other viable new feedstocks for biogas production. In this work, five Crassulacean Acid Metabolism (CAM) species from the five different genera (Agave , Ananas , Euphorbia , Kalanchoe , and Opuntia) were selected as alternative feedstocks and their biochemical methane potentials (BMP) were investigated. Batch assays were performed using sludge and rumen fluid as inocula under uncontrolled pH and at mesophilic temperature (39 °C). Mean methane yields from the CAM plants inoculated with AD sludge ranged from 281 to 382 ml/gVS. These values were not significantly different from the methane yield obtained from maize, a feedstock for biomethane and volatile fatty acid (VFA), suggesting that CAM plants may be viable as bioenergy crops on poor-quality soils in areas with low rainfall that are unsuitable for cultivation of food crops. [ABSTRACT FROM AUTHOR]