<p>U ovoj disertaciji su prikazani rezultati kompleksnih istraživanja uticaja<br />kosmičkog zračenja na nuklearne spektre na nivou mora.<br />Utvrđivanje specifičnih vremenskih intervala u vremenskom spektru,<br />kojima odgovaraju tačno određene grupe događaja indukovane<br />kosmičkim mionima, omogućava adekvatno odbacivanje ometajućih<br />događaja u željenim energetskim oblastima detektovanih spektara, pri<br />antikoincidentnom režimu rada ultraniskofonskih sistema. U prvom<br />eksperimentu prikazanom u ovoj disertaciji,&nbsp; istraživana je mogućnost<br />vremenskog razlaganja detektovanih događaja,&nbsp; pomoću koincidentnog<br />sistema u čijem sklopu se nalazi HPGe detektor i&nbsp; plastični scintilator, u<br />povr&scaron;inskoj laboratoriji. Ustanovljeno je da se promptni i zakasneli<br />koincidentni događaji između plastičnog &ldquo;veto&rdquo; detektora i<br />germanijumskog detektora mogu jasno razdvojiti u dve grupe, za oko<br />100 ns. Dodatno, zakočno zračenje i&nbsp; anihilacioni događaji mogu se<br />razdvojiti u vremenu od (n,n&rsquo;) događaja, iako svi ovi događaji pripadaju<br />grupi zakasnelih događaja. Takođe, registrovani su&nbsp; i&nbsp; značajno zakasneli<br />anihilacioni događaji, koji nastaju usled raspada zaustavljenih pozitivnih<br />miona.<br />Drugi eksperiment prikazan u ovoj disertaciji baziran je na ultra-niskofonskom HPGe spektrometru relativne efikasnosti 100%.<br />Dodatkom dva plastična scintilatora i brzo-sporog koincidentnog kola,<br />istraživani su koincidentni događaji između plastičnih scintilatora i<br />HPGe spektrometra. Ovaj spektrometarski sistem MIREDO <em>(Muon<br />Induced Rare Event Dynamic Observatory</em>) prvenstveno je namenjen<br />proučavanju procesa indukovanih kosmičkim mionima u različitim<br />materijalima. Analiza ovakvih interakcija može biti od značaja za ultra-niskofonske eksperimente. Rezultati dobijeni za tri ispitivana materijala,<br />pakovana u<em> Marinelli</em> sud, prezentovani su i diskutovani.<br />U trećem eksperimentu prikazanom u ovoj disertaciji ispitan je<br />potencijalni uticaj solarnih&nbsp; neutrina&nbsp; na izmerenu brzinu&nbsp; radioaktivnog<br />raspada,&nbsp; merenjem varijacija u brzini brojanja<br />³H&nbsp; metodom tečnog&nbsp;scintilacionog brojanja. Kori&scaron;ćenjem sofisticiranog tečnog scintilacionog&nbsp;spektrometra&nbsp;<em> Quantulus</em>&nbsp; ustanovljeno je da na merenje<br />visokoenergetskog dela&nbsp;³H spektra može značajno uticati nestabilnost<br />instrumenta. Oscilatorni karakter izmerenog visokoenergetskog dela<br />³H&nbsp;spektra&nbsp; je registrovan, ali sa veoma malom amplitudom (manjom od<br />0.5%), koja se ne može jednostavno objasniti samo nestabilno&scaron;ću<br />instrumenta.&nbsp; Kada je meren ukupan&nbsp;³H spektar, nisu nađene značajne<br />varijacije u brzini brojanja.<br />Već duže vreme je poznato da je niskoenergetsko gama zračenje<br />kontinualne distribucije prisutno na otvorenom prostoru, u vazduhu na<br />povr&scaron;ini Zemlje. U prethodnim istraživanjima pretpostavljano je da&nbsp; ovo<br />zračenje potiče skoro isključivo od gama fotona koji su emitovani usled<br />prirodne radioaktivnosti i potom rasejani u nazad od strane vazduha<br />iznad zemlje. U četvrtom eksperimentu prikazanom u ovoj disertaciji<br />pokazano je da&nbsp; je&nbsp; ovo zračenje (u energetskom regionu 30 keV-300<br />keV), sa maksimumom na oko&nbsp; 90 keV, u značajnoj meri&nbsp; proizvedeno<br />kosmičkim zračenjem, sa fluksom fotona od oko 3000 m^-2s^-1. Takođe,<br />ustanovljeno je da dozama op&scaron;te populacije doprinosi ovo sveprisutno<br />niskoenergetsko gama zračenja kosmičkog porekla, zajedno sa<br />odgovarajućim fluksom niskoenergetskih elektrona i da ove komponente<br />ukupnih doza indukovanih kosmičkim zračenjem na nivou mora nisu<br />zanemarljive.</p> / <p>In this dissertation results of complex research on cosmic-ray impact on&nbsp;nuclear spectra at sea level are shown.&nbsp; The appropriate selection of&nbsp;coincidence time interval &nbsp;in low-background experiments that are based&nbsp;on the rejection of anticoincidence background events is very important&nbsp;for reducing the influence of cosmic-ray muons on acquired spectral&nbsp;data. In&nbsp; the first&nbsp; experiment&nbsp; presented in this dissertation,&nbsp; performed by&nbsp;the coincidence system of an HPGe detector and a plastic detect or&nbsp; in a&nbsp;surface laboratory, the time resolution of the detected events is explored.&nbsp;It is&nbsp; found that the prompt and delayed coincidence events between a&nbsp;plastic veto detector and a &nbsp;Ge detector can be sharply divided for&nbsp;approximately 100 ns in two groups. In addition, the bremsstrahlung&nbsp;and annihilation events can&nbsp; be time-resolved from the (n,n&rsquo;) events,&nbsp;although all of these events belong to the group of delayed events. Also,&nbsp;substantially delayed annihilation events, which are caused by the&nbsp;<br />decays of stopped positive muons, were detected.<br />The second experiment shown in this dissertation is based on the 100%&nbsp;relative efficiency ultra-low-background HPGe spectrometer. With the&nbsp;addition of two plastic scintillators and a fast-slow coincidence circuit,&nbsp;the coincidence events between the plastic detectors and the HPGe&nbsp;spectrometer have been investigated. This&nbsp; MIREDO (Muon Induced&nbsp;Rare Event Dynamic Observatory) spectrometer system is primarily&nbsp;<br />developed for the study of cosmic muon induced processes in different&nbsp;materials. Exploration of such interactions can be important for ultra-low&nbsp;background experiments. Results derived for three samples, placed in a&nbsp;Marinelli beaker, are presented and discussed.<br />In third experiment shown in this dissertation,&nbsp; the potential influence of&nbsp;solar neutrinos on measured decay rate is investigated by&nbsp; the liquid&nbsp;scintillation measurement of the count rate variations of&nbsp;³H. Making use of the sophisticated Quantulus liquid scintillation spectrometer, it is&nbsp;found that the&nbsp; measurement of the high-energy tail of&nbsp;³H spectrum may be significantly influenced by instrumental&nbsp; instability.&nbsp; The oscillatory behavior of measured high-energy tail of ³H spectrum&nbsp; is registered, but&nbsp;with very small amplitude (less than 0.5%), which cannot&nbsp; be easily&nbsp;<br />explained only by instrumental instability. When the total&nbsp;³H spectrum was measured, no significant variations in the count rate were found.&nbsp;<br />For a long time, it has been known that low-energy continuous gamma&nbsp;radiation is present in open air at the Earth&rsquo;s surface. In previous&nbsp;investigations it was assumed that this radiation is produced almost&nbsp;exclusively by gamma photons emitted due to the natural radioactivity,&nbsp;which are backscattered by air above ground.&nbsp; In the fourth experiment&nbsp;presented in this dissertation, it is&nbsp; shown&nbsp; that significant amount of this&nbsp;<br />radiation (related to energy region 30&nbsp; keV-300&nbsp; keV) that peaks at about&nbsp;90 keV, is produced by cosmic-rays, with the photon flux of about 3000&nbsp;m^-2s^-1.&nbsp; Also, it is found&nbsp; that the contribution of this omnipresent low&nbsp;energy gamma radiation of cosmic-ray origin, including the&nbsp;corresponding low-energy electron flux, to the doses of general&nbsp;<br />population are non-negligible components of overall doses induced by&nbsp;cosmic rays near sea level.</p>

The estimation of the worst-case execution time (WCET) of a task is a problem that concerns the field of embedded systems and, especially, real-time systems. Estimating a safe WCET for single-core architectures without speculative mechanisms is a challenging task and an active research topic. However, the advent of advanced hardware mechanisms, which often lack predictability, complicates the current WCET analysis methods. The field of Embedded Systems has high safety considerations and is, therefore, conservative with speculative mechanisms. However, nowadays, even safety-critical applications move to the direction of multiprocessor systems. In a multiprocessor system, each task that runs on a processing unit might affect the execution time of the tasks running on different processing units. In shared-memory symmetric multiprocessor systems, this interference occurs through the shared memory and the common bus. The presence of private caches introduces cachecoherence issues that result in further dependencies between the tasks. The purpose of this thesis is twofold: (1) to evaluate the feasibility of an existing one-pass WCET analysis method with an integrated cache analysis and (2) to design and implement a cachebased multiprocessor WCET analysis by extending the singlecore method. The single-core analysis is part of the KTH’s Timing Analysis (KTA) tool. The WCET analysis of KTA uses Abstract Search-based WCET Analysis, an one-pass technique that is based on abstract interpretation. The evaluation of the feasibility of this analysis includes the integration of microarchitecture features, such as cache and pipeline, into KTA. These features are necessary for extending the analysis for hardware models of modern embedded systems. The multiprocessor analysis of this work uses the single-core analysis in two stages to estimate the WCET of a task running under the presence of temporally and spatially interfering tasks. The first phase records the memory accesses of all the temporally interfering tasks, and the second phase uses this information to perform the multiprocessor WCET analysis. The multiprocessor analysis assumes the presence of private caches and a shared communication bus and implements the MESI protocol to maintain cache coherence. / Uppskattning av längsta exekveringstid (eng. worst-case execution time eller WCET) är ett problem som angår inbyggda system och i synnerhet realtidssystem. Att uppskatta en säker WCET för enkelkärniga system utan spekulativa mekanismer är en utmanande uppgift och ett aktuellt forskningsämne. Tillkomsten av avancerade hårdvarumekanismer, som ofta saknar förutsägbarhet, komplicerar ytterligare de nuvarande analysmetoderna för WCET. Inom fältet för inbyggda system ställs höga säkerhetskrav. Således antas en konservativ inställning till nya spekulativa mekanismer. Trotts detta går säkerhetskritiska system mer och mer i riktning mot multiprocessorsystem. I multiprocessorsystem påverkas en process som exekveras på en processorenhet av processer som exekveras på andra processorenheter. I symmetriska multiprocessorsystem med delade minnen påträffas denna interferens i det delade minnet och den gemensamma bussen. Privata minnen introducerar cache-koherens problem som resulterar i ytterligare beroende mellan processerna. Syftet med detta examensarbete är tvåfaldigt: (1) att utvärdera en befintlig analysmetod för WCET efter integrering av en lågnivå analys och (2) att designa och implementera en cache-baserad flerkärnig WCET-analys genom att utvidga denna enkelkärniga metod. Den enkelkärniga metoden är implementerad i KTH’s Timing Analysis (KTA), ett verktyg för tidsanalys. KTA genomför en så-kallad Abstrakt Sök-baserad Metod som är baserad på Abstrakt Interpretation. Utvärderingen av denna analys innefattar integrering av mikroarkitektur mekanismer, såsom cache-minne och pipeline, i KTA. Dessa mekanismer är nödvändiga för att utvidga analysen till att omfatta de hårdvarumodeller som används idag inom fältet för inbyggda system. Den flerkärniga WCET-analysen genomförs i två steg och uppskattar WCET av en process som körs i närvaron av olika tids och rumsligt störande processer. Första steget registrerar minnesåtkomst för alla tids störande processer, medans andra steget använder sig av första stegets information för att utföra den flerkärniga WCET-analysen. Den flerkärniga analysen förutsätter ett system med privata cache-minnen och en gemensamm buss som implementerar MESI protokolen för att upprätthålla cache-koherens.

Graphics Processing Units (GPUs) are emerging as the most popular accelerator for many applications, powering the core of Machine Learning applications and many computing-intensive workloads. GPUs have typically been consideredas accelerators, with Central Processing Units (CPUs) in charge of the mainapplication logic, data movement, and network connectivity. In these architectures,input and output data of network-based GPU-accelerated application typically traverse the CPU, and the Operating System network stack multiple times, getting copied across the system main memory. These increase application latency and require expensive CPU cycles, reducing the power efficiency of systems, and increasing the overall response times. These inefficiencies become of higher importance in latency-bounded deployments, or with high throughput, where copy times could easily inflate the response time of modern GPUs. The main contribution of this dissertation is towards a GPU-centric network architecture, allowing GPUs to initiate network transfers without the intervention of CPUs. We focus on commodity hardware, using NVIDIA GPUs and Remote Direct Memory Access over Converged Ethernet (RoCE) to realize this architecture, removing the need of highly homogeneous clusters and ad-hoc designed network architecture, as it is required by many other similar approaches. By porting some rdma-core posting routines to GPU runtime, we can saturate a 100-Gbps link without any CPU cycle, reducing the overall system response time, while increasing the power efficiency and improving the application throughput.The second contribution concerns the analysis of Clockwork, a State-of-The-Art inference serving system, showing the limitations imposed by controller-centric, CPU-mediated architectures. We then propose an alternative architecture to this system based on an RDMA transport, and we study some performance gains that such a system would introduce. An integral component of an inference system is to account and track user flows,and distribute them across multiple worker nodes. Our third contribution aims to understand the challenges of Connection Tracking applications running at 100Gbps, in the context of a Stateful Load Balancer running on commodity hardware. / <p>QC 20240315</p>