Ali has his Masters in Computer Engineering from UIUC. He believes that the problems faced in the field of big data are best addressed with intimate knowledge of all layers of the computing stack. In this vein, he has pursued a wide variety of fields including computer architecture, distributed systems, system software, GPGPU, many-core architectures, and machine learning. At Flux7 Labs, he leads the team on the technical front managing a variety of consulting projects. In these projects he has dealt with Hadoop, distributed database design, and innovative network communication technologies. Before this Ali was doing performance analysis and simulation at ARM. His focus there lay in techniques to get data on and off the CPU on high-end ARM systems including Cortex-A57 and CCI-400. He also made significant contributions to the development of the ARMv7 Virtualization Extensions and Large Physical Address Extension. He was also a part of the performance team on Intel Atom.
CTO and Co-founder @ Flux7 Labs is a solutions company in the cloud and DevOps space. Our expertise is in helping the operations teams develop scalable resilient solutions while enabling increased productivity for the developers. Having solved the same problems multiple times we bring a set of best practices, reference architectures and customizable solutions, and expertise across a wide-range of platforms. In addition, we are constantly investing in research to provide our clients with innovative solutions. The majority of our engagements are on the cloud, in particular Amazon Web Services (AWS).
As the CTO of Flux7 Labs I handle the technical team on three fronts. The first is directly in client engagements. I both supervise other employees and do hands-on work. In addition I manage the relationship with the client to ensure their needs are met. My second impact is in managing our company's research, ensuring we are creating value for our customers and showcase our technical excellence. Our blog (www.flux7.com/blogs) contains some of the interesting information coming out of this effort. My third area of impact in Flux7 Labs is in technology development. We are always developing technology solutions to improve the flow for our customers. Some of these are for internal workflow like our Docker-based developer workflows. We have packaged the most directly beneficial services in our attune packages for AWS and Docker. These packages are designed to get you from 0-60 in a very short amount of time.
Solutions:
Cloud migration, developer workflows, scalable and resilient systems, performance analysis
Technologies, including:
Analytics: Hadoop, Storm, ElasticSearch, Spark
Databases: Cassandra, CouchDB, MongoDB, DynamoDB
Code Management: Jenkins, Gerrit, Red Mine, Git, Subversion
Management: Chef, OpenStack, Puppet, Ansible, Salt, Docker, Rundeck
Frameworks: Ruby on Rails, Python Django , Wordpress From April 2013 to Present (2 years 9 months) Modeling and Performance Engineer at ARM @ Performance Modeling
- Modeled, correlated, and analyzed the performance of Level 2 cache and memory bandwidth on Cortex-A15 and Cortex-A57.
- Scope of work included the TLBs, the prefetcher, and the coherence logic.
- Conducted and designed performance Experiments.
- Created methodology for memory throughput investigations adopted by the rest of the team.
- Identified bottlenecks and worked with RTL designers to improve the design.
- Instrument RTL events for enabling performance debugging. The instrumented units included decode, dispatch, reorder buffer, issue queues, loadstore and level2 cache.
- Led investigation of latencies and throughput to better understand memory hierarchy and improve performance.
- Led to major improvements to the prefetcher performance.
- Maintained performance regressions on RTL and debugged performance drift.
Development of Fastsim Functional Simulator
- Developed new architectural features and implementation defined features for the Cortex-A15 Fastsim model.
- Developed a harness and plugins for running Fastsim in lockstep with the RTL for the validation team.
- Maintained the simulator for incoming requests from validation and product release of the simulator
- Gained comprehensive command over the ARMv7 architecture encompassing the application instructions, system level instructions, virtualization extensions, implementation defined features and unpredictable cases.
- Used my software experience along with the knowledge of the ARM architecture to help RTL designers better understand use-cases and requirements for various instructions.
Gained familiarity with writing and debugging JIT'ted code. From September 2008 to April 2013 (4 years 8 months) Austin, Texas AreaTeaching Assistant @ Head Teaching Assistant (Spring 2008)
- In addition to regular teaching assistant responsibilities, led a team of 8 teaching assistants
- Instituted several notable changes to the teaching of the course and roles and responsibilities of TAs
- Started one-on-one code reviews with students
- Created discussion section worksheets to reinforce lectures
Teaching Assistant (Fall 2006 and Spring 2007)
- Lectured 20-40 students in discussion sections
- Held office and lab hours
- Developed and graded homework problems, programming labs, and exams. From August 2006 to May 2008 (1 year 10 months) Urbana-Champaign, Illinois AreaPerformance Modeling in Intel Atom Team @ - Developed infrastructure for benchmark regressions.
- Developed hyper threaded workloads.
- Profiled instructions for estimating speedup of certain features.
- Automated measurement of power usage data. From May 2007 to December 2007 (8 months)
Master of Science, Computer Engineering @ University of Illinois at Urbana-ChampaignBachelor of Science with High Honors, Computer Engineering @ The University of Texas at AustinIntermediate of Science- Highschool equivalent, Pre-Engineering @ Government College University, Lahore Ali Hussain is skilled in: Distributed Systems, Performance Analysis, Machine Learning, Cloud Computing, Databases, ARM, Computer Architecture, Low-power Design, Assembly Language, Validation, Binary Translation, SystemC, Hadoop, Computer Graphics, Twitter Storm