Things in the OR world I am looking forward to in 2009 – Part I: Gurobi’s New Software

Already the first month of 2009 is over.  I am finally catching up…    So, this post comes a little late (I started writing this series of posts in 2008!)  but finally here it is.

A lot happened on the software side of operations research in 2008, one of the most notable being ILOG’s acquisition by IBM. The purchase was finalized early January 2009.  But there is a new company in town: Gurobi Optimization and everyone is excited.

With a management team composed of CPLEX’s founder and brain power Robert Bixby, Ed Rothberg and Zonghao Gu, expectations are high.

Version 1.0 of the Gurobi suite of optimization products will be released in April, 2009, prior to the INFORMS Practice meeting.

They will be in Phoenix, AZ at the pre-conference workshop on Sunday, April 26th.

Looks like one of the strongest points of Gurobi’s optimization software will be its capability of taking advantage of multiple cores and processors:

The Gurobi MIP solver will include shared memory parallelism, capable of simultaneously exploiting any number of processors and cores per processor. The implementation will be deterministic: two separate runs on the same model will produce identical solution paths.

No official software yet but the first results comparing CPLEX, Gurobi and Mosek came 2 days ago:  Take a look at the initial results.  For the test set 1 (selected problems from MIPLIB) CPLEX beats Gurobi 8 to 5.  But for the test set 4, especially for the neos problems, Gurobi beats CPLEX 17 to 8 with 2 ties and 1 problem unsolved by both software using a single thread (if I didn’t miscount!).  At a first glance, the performance of Gurobi relative to CPLEX using a single thread vs. 4 threads improves only slightly.  I can’t wait to see the new results in April.

On another note, my first reaction when I first heard of Gurobi was “what an interesting name, Gurobi”… It looks like a combination of the first two letters of the co-founders’ last names Gu (Zonghao Gu) Ro (Edward Rothberg) Bi (Robert Bixby). Just a guess.


One Response to Things in the OR world I am looking forward to in 2009 – Part I: Gurobi’s New Software

  1. Although Gurobi Optimization plans to offer the Gurobi Solver directly (with its own APIs and tool set) starting in April, it’s available in a full commercial release this month (February) as embedded in several modeling systems (AIMMS, Frontline, GAMS, MPL). We have it available for download right now (15 day free trial with no other restrictions) at You can run the Mittleman benchmarks yourself, or solve your own models in MPS, LP and OSIL format using a little program GurobiEval that we include. You can create models from scratch, and solve them with Gurobi, either in Excel with our modeling system, Risk Solver Platform, or in C, C++, C#, VB6, VB.NET, Java or MATLAB using our Solver Platform SDK.

    You’re right that, on the Mittleman problems, the relative performance of the Gurobi Solver improves only modestly on 4-cores vs. 1-core. I understand from Gurobi that this is due to many factors, including the fact that Gurobi can solve a number of these problems at the root node via cuts without ever branching — and of course its parallelism comes fully into play only when it does start branching.

    We have a selection of 50 of the MIPLIB problems (some old, some new) that seem to be very hard to parallelize. On these problems, the Gurobi Solver’s performance on 4-cores vs. 1-core is much more impressive, with improvements on 42 of the 50 problems. We’ve seen even better relative performance on some customers models in our internal library.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: