Sebastian Pokutta's Blog

First benchmark results for the new (and still unreleased) solver of Gurobi (the new company of Gu, Rothberg, and Bixby) have been published yesterday by Hans Mittelmann and compared to CPLEX 11.2. The results look pretty decent and it looks like the gurobi code is very much comparable to the one of CPLEX in terms of performance.

From the Intechne Blog >>Less is more<<:

• Gurobi solves MIPs faster on single-processor machines: It beats or equals CPLEX’s performance 55% of the time. (A third solver – MOSEK – is also compared. It wins or ties for first 5% of the time.)
• Gurobi parallelizes robustly: Having used it in a number of projects over the years, I can attest that CPLEX’s parallel branch and bound is a very solid piece of code. In this one-on-one comparison on a 4-processor machine, Gurobi wins or ties CPLEX 58% of the time.
• Aggregate solution times are less on Gurobi: Considering only problems on which at least one integer solution was found, on a single-CPU system Gurobi runs through the test-set in 25% less time than CPLEX, as measured using the geometric mean of times. On 4-CPU systems, the improvement is still considerable – 15%.
• Gurobi is good at finding integer feasible solutions: CPLEX fails to identify a single integer feasible solution on 2 instances in either mode. Gurobi fails on one instance in single-threaded mode. It finds at least one integer solution on all test-set problems when using four processors.
• CPLEX takes better advantage of parallelization: CPLEX’s speed-up in going from one to four processors is 40%, whereas Gurobi only manages 30%.

Also check out the post on Michael Trick’s Blog.

UPDATE: A trial version of the Gurobi solver is available on http://www.solver.com/gurobi. For details, see the comment of Daniel H. Fylstra in the comment section of this post.

UPDATE 05/20/2009: The standalone version of Gurobi is available now. (see also here)

Did you ever wonder how much merchants are charged when you use your credit card? The website True Cost of Credit provides you with these numbers for major credit card providers – the numbers are for the US though. Given that the fees are not insignificant (somewhere between 2-4% in most cases and can go as high as 18% for e.g. gums) one could easily imagine that these fees eat up a lot of the profit margin or in turn are at least partially passed on to the customer.

I just found the Mendeley service online. They offer a software for free to manage your research library including nifty features such as automatic meta data extraction and much more. I am going to check it out and let you know what I think about it. Let’s see if this sofware can reduce the hassle of not knowing where exactly *that* paper is. Stay tuned!

I am going to Zaragoza, Spain to talk about “Computational methods in Production Planning and Logistics” at the MIT-ZARAGOZA Speaker Series this Wednesday (Jan, 28th).

Abstract:

The impact of modern computational methods from operations research, analytics, and computer algebra when solving supply chain management, logistics, and planning problems has been enormous. Many supply chains could hardly be operated without the support of sophisticated, computerized optimization and planning engines nowadays.

I would like to present two special production and planning problems which have been successfully solved with computational methods. The first one is a stowage optimization problem for inland vessels which has been tackled with integer programming. The challenge for the transportation of inland vessels on auxiliary routes comprises low bridge heights and reduced water depths. In order to ensure save passage of bridges of critical heights load masters are forced to load the vessels with a reduced number of containers in order to reduce the ship’s maximal height over waterline which effectively reduces the ship capacity. We present a solution to calculate optimal stowage plan that restores the ship’s full capacity. The second problem is a production problem from the oil industry that has been addressed with methods from computer algebra. The available data that describes the oil production process is noisy and of low quality. Nonetheless, understanding the subsurface fluid dynamics that guide the production process is essential for optimizing production and keeping the field in a healthy state. The presented solution concept starts solely from noisy measured data and returns a (polynomial) model that approximates the production process. The derived models can be used for monitoring and control purposes in a second step.