skip to primary navigationskip to content


What do some of our users have to say about CamGrid?

"In much of our work, we require significant high-throughput capabilities. In some cases we have to run simulations over many different parameters, such as temperature, or run repeated runs for high statistical accuracy. These requirements are best me t by grid computing infrastructures, and CamGrid is ideal for this. In particular, CamGrid has an excellent set of compute resources with good compute performance and memory (considerably better than other current campus grids). We are using CamGrid for atomistic simulations of the behaviour of materials and minerals, for exploring the interactions between atoms within materials, for computations of the interactions between pollutants and minerals, and for inverse modelling of neutron scattering data. All these tasks require significant high-throughput capability without which the projects would simply be impossible to run. CamGrid provides us with the on-demand compute capacity to make these projects viable. Even ownership of a local cluster would not be sufficient; CamGrid offers us capabilities that significantly exceed what a cluster would provide." 

-- Prof. Martin Dove, Department of Earth Sciences


"I am a population geneticist who collaborates closely with mathematicians to develop new methods of analysing large genetic datasets. Our recent work has focused on Markov chain Monte Carlo methods, which allow the extraction of often very faint signals of historical population mixing and changes in size, but which are highly computer intensive. In a typical study we construct a 20 x 20 population matrix and each comparison takes 12-24 hours to run - in practice requiring in excess of 1 year's dedicated CPU time. Without CamGrid this research would simply not be feasible. Over the last year I have used close on a decade of CPU time! This research has the potential to revolutionise our knowledge of human evolution." 

-- Prof. Bill Amos, Department of Zoology


"Here in the high energy physics group we have been using cosmoligical constraints on dark matter and particle physics data to forecast likely signals in the forthcoming Large Hadron Collider experiment at CERN, Switzerland. The experiment may be able to produce dark matter particles, but identifying them will be like finding the proverbial needle in a haystack. Our research allows us to find strategies to do this. CamGrid was an invaluable tool allowing us to reliably sample the large parameter space in a reasonable amount of time. A half-year's worth of CPU running was collected in a week." 

-- Dr. Ben Allanach, Department of Applied Mathematics and Theoretical Physics


"In the high energy physics (HEP) group we have been exploring a well defined class of string theory models that has distinct low energy implications that could be tested at the large hadron collider (LHC) that is about to start running at CERN, Geneva. In previous work we were able to calculate the masses of supersymmetric particles at the string scale (which is much higher than the TeV scale to be probed at LHC). In order to make contact with the lower energy scale, a well defined prescription has to be followed by using the renormalization group equations to compute the spectrum of supersymmetric particles at low energies. For this there are codes (like SOFTSUSY) that computes the masses at LHC energies. Furthermore we also need event generators (like PYTHIA) and detector simulators (like PGS) in order to go all the way to the measurable quantities at the LHC. All these codes require a heavy amount of computer power. CamGrid was essential in order for us to be able to run the different codes in real time. We are finishing a long article (with B. Allanach, J. Conlon, C.-H. Kom, and K. Suruliz) that illustrates all the computer power that was used (simulating for instance one year of LHC running). We acknowledge CamGrid for invaluable help." 

-- Prof. Fernando Quevedo, Department of Applied Mathematics and Theoretical Physics


"As an evolutionary biologist addressing issues of eukaryotic cellular origins and complexity, computational resources are exceedingly important. The CamGrid system has allowed me to analyse my data much more quickly than have other systems to which I have had access. Analyses that would take a week on other servers are finished overnight on CamGrid. The system has also allowed me to expand my research in size and scope. As an example, I recently ran phylogenetic analyses on three datasets each consisting of 45 taxa and approximately 36,000 amino acid positions. These datasets were bootstrapped 100X and analysed using a protein maximum-likelihood method. On CamGrid, they ran concurrently and were all completed in less than 5 days. This analysis simply would not have been possible with any of the computational resources that I previously have had available. CamGrid has greatly enhanced my research into protist genomics, eukaryotic relationships and the evolution of membrane-trafficking." 

-- Dr. Joel Dacks, Department of Pathology


"For a detailed model of protein movements in bacterial cells, I needed to run simulations that took a couple of weeks each. Without access to the processors on CamGrid, it would have taken a couple of years to get enough results for a publication." 

-- Dr. Karen Lipkow, Department of Biochemistry


"Our research is generally concerned with understanding quantum mechanical effects in semiconductor nanostructure devices. We have a substantial experimental program to investigate this which is supported by theoretical and computational work, mainly dealing with numerically solving the time-dependent Schrodinger equation. This is an extremely computationally intensive task, and it would have been impossible to perform this research without the resources provided by CamGrid." 

-- Dr. Crispin Barnes, Semiconductor Physics Group


"We employ microarray based comparative genomic hybridisation to investigate genomic imbalances in cervical carcinoma. Reliably segmenting and calling significant clones from arrays across large data sets is proving a bioinformatics challenge to which numerous CPU intensive mathematical models have been applied. To ensure no loss of biological data and also validate the performance of several favored models (Heterogeneous Hidden Markov Model [HMM]), a biological weighted HMM (BioHMM), a non-parametric change-point method (DNAcopy), and a Gaussian approach (GLAD) we have turned to CamGrid in order to carry out our data processing. This strategy would have proven impractical without CamGrid's high throughput capabilities. Moreover, we are extremely grateful to Mole and CamGrid for making this service available to end users without the need for extensive IT knowledge." 

-- Dr. Ian Roberts, MRC Cancer Cell Unit