| 1. |
|
|
| 2. |
- Bignert, Anders, et al.
(författare)
-
Comments Concerning the National Swedish Contaminant Monitoring Programme in Marine Biota, 2015
- 2015
-
Rapport (övrigt vetenskapligt/konstnärligt)abstract
- The environmental toxicants examined in this report can be classified into five groups – heavy metals, chlorinated compounds, brominated flame retardants, polyaromatic hydrocarbons and perfluorinated compounds. Each of these contaminants have been examined from various sites for up to six different fish species, in blue mussels, and in guillemot eggs, for varying lengths of time. The following summary examines overall trends, spatial and temporal, for the five groups.Condition and Fat ContentCondition and fat content in different species tended to follow the same pattern at the same sites, with a few exceptions. Most of the fish species generally displayed a decreasing trend in both condition and fat content at most sites examined. Exceptions to this were increases in condition factor seen in cod liver at Fladen, perch muscle at Kvädöfjärden, and for herring at Ängskärsklubb in spring. Also, an increase in fat content was seen during the most recent ten years for herring at Ängskärsklubb in spring. There were also some sites where no log linear trends were seen.Heavy MetalsDue to a change in methods for metal analysis (not mercury) in 2004, values between 2003 and 2007 should be interpreted with care. From 2009 metals are analyzed at ACES, Stockholm University.Generally, higher mercury concentrations are found in the Bothnian Bay, but also from one station in the Northern parts of Baltic Proper, compared to other parts of the Swedish coastline. The time series show varying concentrations over the study period. The longer time series in guillemot egg and spring-caught herring from the southern Bothnian Sea and southern Baltic Proper show significant decreases of mercury. On the other hand, increasing concentrations are seen in e.g., cod muscle, but the concentrations are fairly low compared to measured concentrations in perch from fresh water and coastal sites. In most cases, the mercury concentrations are above the EQSbiota of 20 ng/g wet weight.Lead is generally decreasing over the study period (in time series of sufficient length), supposedly due to the elimination of lead in gasoline. The highest concentrations are seen in the southern part of the Baltic Sea. Elevated lead concentrations between 2003 and 2007 (e.g. Harufjärden) should be viewed with caution (see above regarding change in analysis methods). Lead concentrations are below the suggested target level at all stations.Cadmium concentrations show varying non-linear trends over the monitored period. It is worth noting that despite several measures taken to reduce discharges of cadmium, generally the most recent concentrations are similar to concentrations measured 30 yearsago in the longer time series. Cadmium concentrations in herring and perch are all below the suggested target level of 160 μg/kg wet weight.The reported nickel concentrations show no consistent decreasing trends. Some series begin with two elevated values that exert a strong leverage effect on the regression line and may give a false impression of decreasing trends. Chromium generally shows decreasing concentrations, possibly explained by a shift in analytical method. The essential trace metals, copper and zinc, show no consistent trends during the monitored period.Generally higher concentrations of arsenic and silver are found along the west coast compared to other parts of the Sweadish coast line. However for silver a few stations in the Bothnian Sea and Bothnian Bay show comparable concentrations to the west coast stations.Chlorinated CompoundsGenerally, a decreasing concentrations were observed for all compounds (DDT’s, PCB’s, HCH’s, HCB) in all species examined, with a few exceptions, such as no change in TCDD-equivalents being seen in herring muscle (except at Änskärsklubb where very high concentrations at the beginning of the sampling period were seen and also at the west coast station Fladen). The longer time-series in guillemot also show a marked decrease in TCDD-equivalents from the start in the late 1960s until about 1985 from where no change occurred for many years, however, during the most recent ten years a decrease in the concentration is seen. Concentrations of DDE and CB-118 are for some species and sites still above their respective target levels.The chlorinated compounds generally show higher concentrations in the Bothnian Sea and/or Baltic Proper when compared to the Bothnian Bay and the Swedish west coast.Brominated Flame RetardantsElevated levels of HBCDD are seen in sites from the Baltic Proper, while the investigated PBDEs show higher concentrations in the Bothnian Bay. In addition, lower concentrations of all investigated PBDEs and HBCDD are seen on the Swedish west coast compared to the east coast. Temporally, significant increases in BDE-47, -99 and -100 have been seen in guillemot eggs since the late 1960s until the early 1990s, where concentrations then began to show decreases. Also, the concentration of HBCDD in guillemot eggs shows a decrease during the most recent ten years. For fish and blue mussels, BDE-47, -99, and -153 decreased at some sites and showed no trend at other sites. The concentration of HBCDD in fish and blue mussels showed inconsistent trends. The concentration of HBCDD is below the EQSbiota of 167 μg/kg wet weight for all fish species from all areas, while the concentration of BDE-47 alone is above the EQSbiota for sumPBDE of 0.0085 ng/g wet weight.PAHsOnly blue mussels have been examined for spatial differences in PAH concentrations. Concentration of ΣPAH was found to be higher from Kvädöfjärden in the Baltic Proper compared to stations at the West coast, but individual PAHs showed varying spatial patterns. Over time, acenaphthalene was rarely found above the detection limit. Significant decreasing trends were observed for ΣPAH, chrysene, fluoranthene and pyrene at Fjällbacka; for naphthalene at Kvädöfjärden; and for pyrene at Fladen.All time series where concentrations of various PAHs were compared with the target value based on OSPAR Ecological Assessment Criteria, or EC Environmental Quality Standards were below the target value.PFASsPFHxS and PFOS show a similar spatial pattern, but PFOS concentrations were approximately 25 times higher than PFHxS levels. The distribution of PFOS is quite homogenous along the Swedish coast but with somewhat higher concentrations in the Baltic Proper. PFOS concentrations in guillemot eggs are about 100-200 times higher than in herring liver. An overall increasing concentration of PFOS in guillemot eggs has been observed throughout the whole time period, however, during the most recent ten years, a change of direction is detected. The longer herring time series from Harufjärden, Landsort, and Utlängan show increasing concentrations for PFOS and most carboxylates. For FOSA, on the other hand, decreasing concentrations are seen during the most recent ten years.Organotin compoundsThe majority of the analysed tinorganic compounds showed concentrations below LOQ. However TBT and DPhT showed concentrations above LOQ at all stations with highest reported concentrations in fish from Örefjärden in the northern part of Bothnian Sea.
|
|
| 3. |
- Danielsson, Jakob
(författare)
-
Automatic Characterization and Mitigation of Shared-resource Contention in Multi-core Systems
- 2021
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Multi-core computers are infamous for being hard to use in time-critical systems due to execution-time variations as an effect of shared resource contention. In this thesis, we study the problem of shared resource contention, which occurs when multiple applications executing on different cores do not have exclusive access to of a shared hardware resource. We investigate performance variations of parallel tasks in multi-core systems and present a method to pinpoint the source of the resource contention using hardware performance counters. We investigate mitigation methods for performance variations due to resource contention, including the Jailhouse hypervisor and the cache-partitioning tool PALLOC. We propose a benchmark strategy that quantifies the isolation gained from a specific isolation technique and exemplify this strategy using the Jailhouse hypervisor. We furthermore present and implement solutions for cache-partition allocation during application runtime. Our implementation aims to avoid over-provisioning of cache through pre-runtime estimations of an application's dependency towards the cache and continuous re-partitioning of the cache memory during application runtime.The primary goal of this thesis is to contribute to a process that automates some of the tedious manual testing needed to detect resource contention bottlenecks. The methods we present in this provide a holistic solution for automatic mitigating resource-contention in a multi-core system. First, we evaluate the risk for shared resource contention when several applications execute simultaneously. We then allocate partitions to mitigate resource contention for applications that risk severe performance degradations. We finally present methods that dynamically re-allocate partition space to meet the performance requirements of the running applications.
|
|
| 4. |
- Danielsson, Jakob, et al.
(författare)
-
Automatic Quality of Service Control in Multi-core Systems using Cache Partitioning
- 2021
-
Konferensbidrag (refereegranskat)abstract
- In this paper, we present a last-level cache partitioning controller for multi-core systems. Our objective is to control the Quality of Service (QoS) of applications in multi-core systems by monitoring run-time performance and continuously re-sizing cache partition sizes according to the applications' needs. We discuss two different use-cases; one that promotes application fairness and another one that prioritizes applications according to the system engineers' desired execution behavior. We display the performance drawbacks of maintaining a fair schedule for all system tasks and its performance implications for system applications. We, therefore, implement a second control algorithm that enforces cache partition assignments according to user-defined priorities rather than system fairness. Our experiments reveal that it is possible, with non-instrusive (0.3-0.7\% CPU utilization) cache controlling measures, to increase performance according to setpoints and maintain the QoS for specific applications in an over-saturated system.
|
|
| 5. |
- Danielsson, Jakob, et al.
(författare)
-
LLM-shark -- A Tool for Automatic Resource-boundness Analysis and Cache Partitioning Setup
- 2021
-
Ingår i: 45th IEEE Annual Computers, Software, and Applications Conference, COMPSAC 2021. - 9781665424639 ; , s. 49-58
-
Konferensbidrag (refereegranskat)abstract
- We present LLM-shark, a tool for automatic hardware resource-boundness detection and cache-partitioning. Our tool has three primary objectives: First, it determines the hardware resource-boundness of a given application. Secondly, it estimates the initial cache partition size to ensure that the application performance is conserved and not affected by other processes competing for cache utilization. Thirdly, it continuously monitors that the application performance is maintained over time and, if necessary, change the cache partition size. We demonstrate LLM-shark's functionality through a series of tests using six different applications, including a set of feature detection algorithms and two synthetic applications. Our tests reveal that it is possible to determine an application's resource-boundness using a Pearson-correlation scheme implemented in LLM-shark. We propose a scheme to size cache partitions based on the correlation coefficient applications depending on their resource boundness.
|
|
| 6. |
- Danielsson, Jakob, et al.
(författare)
-
Measurement-based evaluation of data-parallelism for OpenCV feature-detection algorithms
- 2018
-
Ingår i: Staying Smarter in a Smartening World COMPSAC'18. - 9781538626665 ; , s. 701-710
-
Konferensbidrag (refereegranskat)abstract
- We investigate the effects on the execution time, shared cache usage and speed-up gains when using data-partitioned parallelism for the feature detection algorithms available in the OpenCV library. We use a data set of three different images which are scaled to six different sizes to exercise the different cache memories of our test architectures. Our measurements reveal that the algorithms using the default settings of OpenCV behave very differently when using data-partitioned parallelism. Our investigation shows that the executions of the algorithms SURF, Dense and MSER correlate to L3-cache usage and they are therefore not suitable for data-partitioned parallelism on multi-core CPUs. Other algorithms: BRISK, FAST, ORB, HARRIS, GFTT, SimpleBlob and SIFT, do not correlate to L3-cache in the same extent, and they are therefore more suitable for data-partitioned parallelism. Furthermore, the SIFT algorithm provides the most stable speed-up, resulting in an execution between 3 and 3.5 times faster than the original execution time for all image sizes. We also have evaluated the hardware resource usage by measuring the algorithm execution time simultaneously with the L3-cache usage. We have used our measurements to conclude which algorithms are suitable for parallelization on hardware with shared resources.
|
|
| 7. |
- Danielsson, Jakob, et al.
(författare)
-
Modelling Application Cache Behavior using Regression Models
- 2021
-
Konferensbidrag (refereegranskat)abstract
- In this paper, we describe the creation of resource usage forecasts for applications with unknown execution characteristics, by evaluating different regression processes, including autoregressive, multivariate adaptive regression splines, exponential smoothing, etc. We utilize Performance Monitor Units (PMU) and generate hardware resource usage models for the L-2-cache and the L-3-cache using nine different regression processes. The measurement strategy and regression process methodology are general and applicable to any given hardware resource when performance counters are available. We use three benchmark applications: the SIFT feature detection algorithm, a standard matrix multiplication, and a version of Bubblesort. Our evaluation shows that Multi Adaptive Regressive Spline (MARS) models generate the best resource usage forecasts among the considered models, followed by Single Exponential Splines (SES) and Triple Exponential Splines (TES).
|
|
| 8. |
- Danielsson, Jakob, et al.
(författare)
-
Resource Depedency Analysis in Multi-Core Systems
- 2020
-
Ingår i: Proceedings - 2020 IEEE 44th Annual Computers, Software, and Applications Conference, COMPSAC 2020. - : Institute of Electrical and Electronics Engineers Inc.. - 9781728173030 ; , s. 87-94
-
Konferensbidrag (refereegranskat)abstract
- In this paper, we evaluate different methods for statistical determination of application resource dependency in multi-core systems. We measure the performance counters of an application during run-time and create a system resource usage profile. We then use the resource profile to evaluate the application dependency on the specific resource. We discuss and evaluate two methods to process the data, including moving average filter and partitioning the data into smaller segments in order to interpret data for correlation calculations. Our aim with this study is to evaluate and create a generalizeable methods for automatic determination of resource dependencies. The final outcome of the methods used in this study is the answer to the question: 'To what resources is this application dependent on?'. The recommendation of this tool will be used in conjunction with our last-level cache partitioning controller (LLC-PC), to make decision if an application should receive last-level cache partition slices.
|
|
| 9. |
|
|
| 10. |
- Danielsson, Jakob, et al.
(författare)
-
Testing Performance-Isolation in Multi-Core Systems
- 2019
-
Konferensbidrag (refereegranskat)abstract
- In this paper we present a methodology to be used for quantifying the level of performance isolation for a multi-core system. We have devised a test that can be applied to breaches of isolation in different computing resources that may be shared between different cores. We use this test to determine the level of isolation gained by using the Jailhouse hypervisor compared to a regular Linux system in terms of CPU isolation, cache isolation and memory bus isolation. Our measurements show that the Jailhouse hypervisor provides performance isolation of local computing resources such as CPU. We have also evaluated if any isolation could be gained for shared computing resources such as the system wide cache and the memory bus controller. Our tests show no measurable difference in partitioning between a regular Linux system and a Jailhouse partitioned system for shared resources. Using the Jailhouse hypervisor provides only a small noticeable overhead when executing multiple shared-resource intensive tasks on multiple cores, which implies that running Jailhouse in a memory saturated system will not be harmful. However, contention still exist in the memory bus and in the system-wide cache.
|
|
