I am a specialist in the design and delivery of storage systems for High Performance Computing, focussing mostly on PFSS which is HPE's imlimentation of Spectrum Scale (GPFS) from IBM. I cover the full suite of HPE's high performance storage products include ClusterStor (Lustre) and the two tiering products: CDS and DMF.

My first six months in HPE was a welcome return to a fully hands-on role in the HPC Storage group at HPE. I helped deploy DMF at several customers in the UK and the rest of Europe. I also deployed Cluster Data Services at a high profile UK customer. The biggest project was the first major "HPC in Private Cloud" deployment for a large customer in Germany. This included configuring PBS Pro, Nice DCV and EnginFrame.

I joined IBM in July 2014 to strength the team following IBM's acquisition of Platform Computing the year before. My core role was in the area of Spectrum Scale (formerly known as GPFS), providing technical presales, Partner enablement and training.My secondary role was in HPC cluster job management with Spectrum LSF and Spectrum Conductor.I covered the full set of IBM Spectrum products including Spectrum Archive (LTFS) and Spectrum Discover. My interested extend into the newer fields of Cognitive computing with IBM Watson and Continerization with Kubernetes.

I provided technical support to Pre-sales and also some customer support, particularly for customers in defence and secure government sites.I helped configured and design HPC clusters, including all networking, CPUs and co-processors together with the environmental aspects of power and cooling.The largest system being 'Blackthorn' at AWE which has over 1000 compute nodes and a full bi-sectional bandwidth Infiniband fabric.

I returned back to Quadrics in April 2008 to continue as an expert in parallel applications, interconnects and cluster management.I designed and installed Quadrics' last customer installation at BAe Systems, Warton. The installation went very smoothly, only taking 2 days at the customer's site until the first customer production jobs were able to be run.The main focus was working on the launch of the long-awaited 3rd generation Quadrics interconnect - with its 7-core (!) network ASIC. This new ASIC though did not make it to a commercial release and Quadrics was subsequently closed. Only four members of staff were saved. I was the most senior of the four. These staff members were transferred to Vega plc. a partner company. I led the support of Quadrics remaining key customer sites including systems in Germany, the USA, BAe Systems and UK government.

Customer facing Engineering role.Worked on porting codes to ClearSpeed's SIMD coprocessor, including Molpro and AMBER, all the way down from Fortran/C down to on-chip microcode. Designed and implimented an API for interfacing co-processors with high-level applications.Supported partners at AWE, B.Ae. and a long list of universities.Help run numerous trade shows including ISC in Dresden and Supercomputing in Tampa and Reno.

Presales consulting and technical support of our customers. Developed at ran the technical side of the shows including SC and ISC.During this time I was seconded to provide onsite support at a major UK government site for 3 mission critical systems. This covered all software issues and escalation of hardware issues. I provided monthly and quarterly reports, automated from live cluster monitoring. I also wrote and developed a cluster monitoring GUI written in Perl. This provided visualisation of the clusters' status in real time. In one case, the view of node temperatures provided and alerted that the machine room temperatures were rising quickly, several minutes before the fault in the aircon equipment was detected.

A great 3 years at the time Manchester hosted the UK's fastest computer - a 768-cpu Cray T3E.I provided technical support to users, particuarly in CFD, meterology and oceanography. I produced and delivered many training courses covering topics from "Vector processing on the Fujitsu VPP" to "Java for HPC"I ran the weekly MRCSS seminar series which brought in visiting speakers from both Industry and Academia., I developed live webcast of these reaching viewer in several countries.I ran the 2-week long 'Linux for HPC" course in Summer 2000 bringing in speakers from 4 countries. This breakthough event was at the time when there were only 28 Linux system on the top500 list. The critics said "for Supercomputing, Linux will never catch on". I helped build the world's first IA64 cluster with Myrinet.I also ran the county's highest bandwidth link. This connected Mancehster with RAL in Oxfordshire (It took the form of trucks carrying DLT tapes - great bandwidth, lousy latency)

Fresh out of my Ph.D I was a PostDoc with some lecturing duties then a full time member of the academic staff.I taught courses in Computing, Numerical Methods, Finite Element Analysis and Design as well as pursuing my research career in large non-linear 3D finite element code development using at the time state of the art supercomputers .