About Me

Monday, December 16, 2013

STRATUS: Is there a future for emerging researchers in NZ? Surviving the postdoc apocalypse

Postdocalypse:
  • Science Exchange: Place to outsource scientific experiments. A way for labs to get funding for the work they like doing.
  • PhD can set you up for startups: Working on something useful, working on data, hardworking, passion.
  • Highly unlikely that postdocs end up being in a long term academic career.
  • Observation: Most researchers are old. Are we in prime position to replace them when they retire?
  • Industry and academic research being disconnected is the wrong mindset. Both tries to answer questions. Just don't lose your skills.
  • Learning to write grants is important!
  • UniServices can give the opportunity for a staff member to pause their teaching for a year and pursue their startup.
  • Overseas positions are good for character development.
  • Use your network to secure your postdoc.
  • Use the postdoc to learn what you want to do and what you don't want to do.
  • Be open minded and look beyond your mentors (e.g., lecturers, colleagues, etc).
  • Might need to step down to step up (in the ranks, pay, etc).
  • Your PhD should have trained you to plan a few years in advance. Periodically evaluate your goals.
  • Money is not why you want to be in research.
  • Tell your story positively.

Wednesday, April 20, 2011

Leadership

Speaker: Roger Branch
  • Taking account. Initiative rather than being reactive.
  • Don't withhold information.
  • Performance standards.
  • Empowering.
  • Plan, Do, Check, Act.
  • Cannot abdicate responsibility.
  • Discretionary effort.
  • Climate.
  • Input.
  • Feedback.
  • Expectations.
  • Delegating within competence.
  • Shared vision
  • Appreciation.
  • Framework.
  • End result.
  • Deadline.
  • Restatement and commitment.
  • Discuss initial ideas.
  • Communication != Understanding != Agreement != Action != Results

Why? (read up) How? (read down)
  • Vision
  • Mission
  • Objective
  • Task

Monday, September 27, 2010

Priv Bradoo

Priv Bradoo
  • PhD on a gene discovery
  • "Business was for losers"
  • Am I empowered?
  • Old business school 301
  • PGSA 3rd board meeting
  • 3 to 13
  • Ironed out in retrospect
  • New set of challenges
  • Ecosystem has to be so entrenched
  • Infused into individuals
  • Could've been a scientist and could've been entrepreneurial.
  • If you don't ask then you'll die not knowing.
  • Opportunities exists but you have to put in the effort.
  • Hold onto your core.
  • Going to clean tech.
  • Success is your ability is have a fully rational understanding of the challenges but an irrational understanding if how to overcome it.
  • US didn't make sense for Lanzatech. China didn't work out.
  • Changed to water clean tech.
  • Singularity University
  • Entrepreneurship is a mindset not a job
  • Goal is to look inside; that there's more than one way.
  • Dream big then dream bigger.
  • Fail is subliminal. Ask how many times they have failed. If they haven't failed, then don't hire them.
  • Learn.
  • Let 1000 flowers bloom.
  • Branding
  • Non-for-profit seed fund for businesses.
  • Spark counted as work experience.

Thursday, September 9, 2010

Derek Handley

Hyperfactory:
  • 2001
  • Mobile advertising communications.
  • Agencies didn't know how big or how fast the mobile advertising would get.
  • Realized that the agencies would start to competing against each other so decided to go straight to the client.
  • Connect brands with their consumers using mobile SMS.
  • Business plan becomes a bog scribble.
  • Believed in the vision.
  • Checking-in to make site you're going towards the goal.
  • Using New Zealand as a test bed doesn't really work.
  • Get a foot in with Saachi & Saachi through a friend.
  • Dad had to mortgage his house for a quarter of a million dollars.
  • Prepared to take the risk to see the rewards?
  • Trouble: you have money. Tend to spend it too fast. Lose focus. Went to raise money from the market too late because of the global financial crisis.
  • Trim down the business.
  • When looking for capital, they got an offer to sell the business to a media company.
  • Took one year to wind back and correct the misalignment and different visions and bad hires.
  • Unwritten values hampered the hiring process.
  • Global benchmarking.
  • Time. Start early. The later you start, the more barriers that will come up.
  • If the customer wants customized products then you end up being a customized services business.

Wednesday, August 18, 2010

Concurrent Python using transactional memory

Speaker: Fuad Tabba

"Parallelism is hard"
  • Figure out the parallelism in the application.
  • Figure out the required synchronisation. How do you protect the critical sections? Race conditions, deadlocks, livelocks. Locks take away the parallelism you had hoped to achieve.
  • Locks have inherent overhead.
Python uses one "Global Interpreter Lock" to protect shared resources.

Transactional memory: Atomicity, Consistency, Isolation, Durability.
The use of transactional memory is to abstract away critical sections so that parallel programs can become easier to write.

In the Sun Rock Processor (which has limited hardware support for transactional memory), there are two caches:
  • L1 Cache tracks memory locations that have been read and written to.
  • Write buffer stores tentative writes (uncommitted transactions).
Cache coherence protocol is used to detect transaction conflicts. The physical size of the caches limit the size of the transactions. When a transaction is committed, the new values have to be propagated to each processor. Transactions can abort/fail for unspecified reasons.

Monday, May 10, 2010

SGI

  • Simulation examples (Hardship)
  • Bandwidth of copper wire doesn't keep up with bandwidth of processors.
  • Latency of conducts not improving.
  • Fibre optics not very fast. Only air core; but increases available bandwidth.
  • mpi_reduce >> mpi_waitall, mpi_reduce, mpi_barrier, mpi_recv.
  • Ethernet chip to offload communication.
  • Memory chip speeds have not increased.
  • Bandwidth/Latency.
  • Linux -> Computer frontend -> Matrix of computing resources.

Experience with and potential of hardware transactional memory

  • Rock processor
  • Everyone needs to do concurrent programming. Not just OS or VM developers.
  • Essence of TM: ability to access multiple memory locations in an atomic transaction, without specifying how atomicity is achieved.
  • Using one lock: not scalable but mimicks transactions. Parallel critical sections.
  • Finer grained locking.
  • Best effort hardware transactional memory - Hardware can abort. Problems raised to the software level.
  • Abort feedback important.
  • Speculative haredware features gave subtle bugs.
  • Lock elision.
  • Doubled ended queues.