MLDS - All the Webhost You Will Ever Need

The current fall in housing prices shows a wonderful example of how the price of something everyone wants, in this case housing, falling can be as painful or more than if prices rise to fast.

There is a conflict at work here. Home owners and near term home owners want prices to rise over time. That is those who currently own homes want to price of their home to look like an investment. This would mean the price of the house someone is willing to pay for it latter is higher above inflation. Not bad, we have heard for years how real estate is a great investment, but what does this mean for everyone else?

Everyone else is low income, or the very young who wont own their home for many years. This is the fight of housing affordability. From a social point of view affordable housing is very important, governments have tried for years to make housing “more affordable”, but do we really want housing to be more affordable? Governments have also done everything they can to encourage home price rising. Be it tax deduction, home owners credit, zoning etc. How many times have you heard of community bi-laws that were to maintain property value?

High property values are in direct conflict with with affordability of housing. People should remember though that a home is a place to live, provide shelter and security, or at worst, store your stuff. Why should our home be investments? Isn’t the investment in having a place for a family an investment in them?

While once I own a house I would love for it to rise in value quickly such that I could sell it in the future to trade for a lake house or other. Problem is if I have children I also want them to afford their own home.

So what will home prices do? I expect home prices over the long term to rise slightly faster than inflation. The rise will be in lock step with the rise in productivity. That is our average wage rises faster than inflation, and we will commit a standard percent of our wage to housing. Other pressure, as building smaller homes being difficult with rising pressure from government for more permits, water hook ups that are over priced etc. That builders can only build larger houses to keep these coasts under control.

In short prices will only be what people will pay for them. Don’t expect to make big money off your home. Unless you can add value to a property (fix it up, develop etc) expect to get what you paid for it after inflation, on average.

I leave you with the plot of house prices as a multiple of income, you tell me what you think house prices are going to be in the future?

I am going to register the domain for the HPC pod-cast soon, everything is coming together. I need a name that is easy to remember. An acronym would be nice, it is simpler for people to say “Are you listening to this weeks ‘blah’”.

Please put in the comments, or email/message me your ideas. To give you an example of how bad I am at this, the only name I can come up with is “HPC Week”. Thanks

I have two babies, My car a Ford Focus SVT in red and a Suzuki GSXR-600 sport bike. I am going to look at the millage comparason that I have seen on these. Note the millage I get tends to be better than stated from the factory. I think this is because my fraction of freeway driving is above the expected. Lets just say I enjoy them, and I take care of them also.

First some pictures.

The bike gets 50 MPG the car gets about 30 MPG. Now it was expected for the bike to get more but look at the size and weight:

Now the bike is kicking out a lot more power, but it should not need this, but assume that both the car and bike must run at maximum output (peak power) when driving. What would the expected millage of that bike be? I get the feeling that the bike should be a lot better than it should, here are the numbers.

One gallon of regular 87 gas has 132 Mega Joule/Gallon. For those 1 Joule/Second is 1 Watt, and 1 HP is 745 Watts. Thus the bike is 85,675 Watts and the car is 126,650 Watts. Thus the car is consuming 47% more energy than the bike.

Thus if the car is getting 30 MPG shouldn’t the bike be getting 47% more? This is about right, 140% of 30 is 42 MPG. Thus the bike at 50 MPG is about what should be expected, this is because of less engine loss, aerodynamic losses and all the other bits about the bike (um it is a lot lighter

I am sure I got most of my math wrong here, (is 47% more of 63% more?) but I am in the ballpark on this one. My assumption that the bike should be getting better millage than it is was false. It gets right where it should be with current engine technology (these are 2004’s and 2003’s).

Sources:

Bikez 2003 GSX-R
Edmonds 2004 Focus SVT

For a while I have thought this was a no brainier. Why is it that interest from bonds (ignore those right now) and savings/checking accounts are taxed at full income tax rates? While dividends and capital gains are taxed at a much lower rate?

I have thought for a while that it was dumb to have the lowest earning types of savings (best I have right now is 4%, much less at most banks) to be taxed at the highest rate available. Where do people save first? Their 3-6Month emergency fund! Which should always be liquid in 24 hours in some form of Money Market with FDIC. If you want to know how to find these shoot me a message. The first account parents open for their kids is a savings account. Now while most children don’t have income up to the standard deduction, if they did should a child’s first savings be taxed at income tax rates? In the child’s case yes, because the first tax bracket starts at 10%.

I have been saving cash, mostly because I have kept my living arrangements low key, so while I have a “big boy” job my expenses are still “collage kid”. This cash is to be used for a wedding, a house, decent furniture when I get tired of the stuff I have accumulated (mis-matched couches?). The interest I earn on that is taxed at my full income tax rates. That is the little I get paid on my savings is taxed as though it is income. True I could spend it as income, but people not saving (why save, inflation will kill it anyway?) is part of the reason the sky is falling all over the news. So I always have thought that there should be no taxes on savings (bank savings) to encourage people to have some savings around. Today though I changed my mind, here is why.

I found my self falling into the trap 99% of the public and politicians fall into. That is “let us change this one thing, everything else will stay the same”. This is dumb on my part. The current rates paid on savings includes expected taxes. For example on my best MMA (Money Market Account) I am making 4.00% APR (www.dollarsavingsdirect.com) which because I am right in the middle of the 25% bracket means I am making after taxes 3.00% which after historical expected inflation is -0-. That is right, I find the best I can make to come out even. So I would expect, if taxes were eliminated, over the next few years in the same inflation environment 3.00% on my cash. What this really means is I expect inflation at best.

Point is, because tweaking some law looks like a good idea, for those like me who save because they should and is the right thing to do right now. Not all things will be equal. To assume nothing else will change. “4% is great if I didn’t have to send 25% of it to congress to waste”, won’t work out that way. The same will apply to bond rates. Now maybe you would want to effeminate taxes because these rates are baked into everything that borrows money. Car loans have higher rates because the person buying the bond (loaning you the money) has to pay taxes and will only take that bond if they can make their taxes. Same goes for school loans, house loans etc. So maybe again eliminating taxes on interest isn’t a bad idea. Oh wait, but if you do that, then why take the risk in stocks? If your just going to pay some of that in taxes, the stocks minus bonds rate will be smaller, net of taxes. So now people are all buying bonds and not providing companies with capital they need to make jobs.

The next time someone says they have all the reasons why changing taxes will be great, punch them (including me, because I still hate our tax system), because they have no idea what all the other 300+ Million Americans are going to do to compensate for that.

Note I mention a car loan above, I in no way encourage anyone to ever take a loan or lease a car. I think mathematically it is a good way to keep your self from getting ahead. When an auto company is making more on leases than the product, be worried.

That is right Multi-Core CPU’s while great for use desktops and laptops to support many background processes running is bad for scientific simulations. I will present data for this using Namd 2.6.

In the old days CPU builders focused on increasing single core speed. Most systems had only 1 core and only high end server had two cores, which lived in two sockets. Many research applications (Like Namd, or the better known Gromacs which is used in Folding@Home. Already were built to run on multiple CPU’s using MPI. MPI allowed researchers to tie together the power of many smaller systems to reach supper computer performance better than that available on purpose built supercomputers. Now even such purpose built machines are using commodity processors and MPI to reach new performance numbers.

To MPI a multi core CPU looks like N CPU’s where N is the number of cores. Thus now with modern quad core CPU’s users can run ‘mpirun -np 8 namd2′ (run namd on 8 cpus/cores). So what is bad about having many cores? CPU builders are rapidly increasing the total performance available in the same socket 1U (1.75 Inch) systems by adding cores. While it is great that a 8 total core box has a lot of performance, in the past that extra performance came from singe core improvements. Thus serial codes (those who can not use multiple cores) benefited and MPI codes benefited also.

With multi core, CPU builders have been lowering the performance of individual cores. That is serial applications, or applications with serial portions will now run slower. Look at the plot of namd running on a cluster of AMD cpus.

The CPU types are dual core Opteron 2218’s at 2.6 GHz and quad core Opteron 2356 “Barcelona’s” at 2.4 GHz. The Barcelona is AMD’s current (as of 12/2008) CPU and from the data it is shown that on 1 core the 2218 is faster. So if NAMD was a serial code the older 2218 would be a better choice. Now in the case of NAMD which scales fine to 8 cores we see that having quad core (if that is the only way to get more performance in a box) is ok. Remember that 4 cores total of 2218’s costs the same about as 8 cores of 2356’s because of the dual vs quad issue. This does not include the cost of power, rack, network etc. Thus for the same cost the 2356 is better because on 8 core namd reaches .45 Days/NanoSecond. While the 4 cores of the 2218 reaches only .70 Days/NanoSecond. Thus in a cost/performance at these small numbers the 2356 is a great deal for small labs running parallel codes up to a few tens of cores.

That is fine for many labs, and they will benefit greatly. The problem, and why Multi-Core is bad for science. Is at the margin. Some place some researcher is trying to run Namd not at 32 cpus as in the plot, but at 2048 and 4096. Namd will have a hard time reaching this limit. Many codes will have no speed improvements from 32 cores and up. Scaling many times has to do with network performance and memory bandwidth. Some applications can not be made parallel! Thus the the above user reached a given performance on 2218’s at 2048 cores, he will need many more 2356 (the newer and better CPU) cores to reach the performance he had before. That may not be possible.

As CPU builders add more and more cores and individual core speed drops, many researchers will find them selves needing to find new ways to make their applications which worked great at smaller number to scale further to just maintain performance on new hardware.

There are limitations to scaling. For example the simulation above was ~29,000 atoms. Atoms in the simulation are spread across cores, thus my upper limit is 29,000 cores. Not good, we need individual core speed to increase and be easily accessible to the programmer. More on that latter.

I recorded how I removed vocals from a pop song to be sung in wedding. Yes I know the quality of the recording is poor, it was done really fast with no rehearsal.

First I must say Happy Thanksgiving to everyone. This is one of my favorite holidays. Why? I love food and good traditional home cooked food, and Thanksgiving is no better way to get it. It means even more now that I have been living on my own for the last 6 years.

Podcast

I have been kicking around the idea of trying to start a pod cast on High Performance Computing (HPC). This is a topic I am intimately familiar with, but I have a lot to learn. My goals would be for sysadmins like my self to learn from other sites how they run their systems, funding models, and in use management software. This would be balanced with user applications. These talks would be for the graduate student, research scientist, and faculty, who would like to hear about tools for developing HPC and Terascale applications as well as tools for those who wish to do some form of science.

In short I am looking for a co-host and software projects to talk to. For a co-host I am looking for someone who is familiar with HPC but less from the admin perspective (Like my self) and more for the science factor. Remember these applicaions will from from large FEA and CFD applications to the biological sciences. Curiosity, a good speaking voice, and a large science background would be helpful, to ask good questions so that those listening who are in those fields will get answers they are looking for.

The show would be modeled after FLOSS Weekly. My own company will have a site up in the future. Shows will be between 30 minutes and 1 hour long, and will start at once every two weeks. I don’t know when I will get this started, mostly hinges on finding a co-host.

And trust me I speak much better than I write. Co-hosts do not need to be close to Ann Arbor, all interviews will be done over Skype. People interested in being co-hosts should email me at: brockp@mlds-networks.com and people who have projects to nominate for being features can ether comment or email. Feel free to mention your own project.

Right now I am in my hotel for the last night before I go home from Austin Texas from this years Super Computing

This year was not that bad. I was able to touch base with PGI about there 8.0 release with support for GPGPUS from Fortran and CCFF (Common Compiler Feedback Format). The GPGPU support looks simple, but it always does. I am sure it is not as easy as it looks, but it is going to be much simpler to work with GPU hardware from Fortran, and if the regular -Minfo -Mneginfo flags work could be very useful. The CCFF stuff looks really good. Lots of data about what the compiler did and what the code did when it ran. Turns out this information in held in XML and they have published this (they claim, I can’t find it) thus third party tools should support it. I for one harassed Allinea about adding support for this profile information into OPT and maybe DDT.

DFT’s on GPU’s. Two papers and one application. Both papers reached the same performance, in both cases much better than the CU-FFT support shipped with CUDA. I hope Nvidia (and ATI) look at these papers and add their support. In many cases, optimizing for bandwidth on the card provided the best performance. This says to me that GPU’s are already in the case of CPU’s which is memory is the bottle neck and there is only marginal improvements that need to be made in the compute engine.

In one of the papers the author looked at total cost for a application they were running. This is great because in my own testing I found copying data to and from real memory to device memory was slow and just awful. Thus moving small parts of your app to GPU’s does not help in most cases. Ether your entire app needs to live on the card (well compute should, pre and post processing not so much) or don’t even bother. In his case GPU’s were good only if using his wonderful optimized FFT. If using the stock CU-FFT, forget it, no point. Copy to and from the card crushed performance. Still promising. I am happy though, that someone pointed out the real downfall of GPU hardware.

The dev’s of NAMD have added CUDA support for GPU’s to their code. This is beta and not GA yet, expect to see it soon. He pointed out something that I didn’t think of. That was, the best performance to and from RAM and device memory is using pinned memory. This is memory that can not be paged out. This allows RDMA style applications to make sure that where the data is placed in memory is where it should be and that the address has not moved (yes the OS moves pages). Well RDMA fabrics like IB also want pinned memory. Thus MPI stacks want to use pinned memory also, but also de-pinns memory that is nolonger needed. What happens when the card and MPI both work with these buffers? It gets un-pinned, data goes into memory in wrong place.

Now you can code around this, but regular research grad students should not have to worry about this. I talked to Jeff Squires of OpenMPI and he was aware of this and Nvidia came to him already that day before I did and have asked about this.

Here is the last bit of insight from the NAMD dev. Why can we not do RDMA between GPU memory in systems with multiple cards? Why can we not do RDMA from GPU memory across and RDMA fabric (think IB, iWARP) to another GPU or RAM in a remote node? This is a great idea! Talking to Jeff should be doable and should help out performance very much on these types of MPI+GPU applications we _will_ see in the future.

Attention Ovid-Elsie Alums.

During the last O-E playoff game WOES ran out of space on their streaming server for people out of town to listen. I (via MLDS) would like to provide WOES with the extra server, software and bandwidth to support an unlimited number of listeners.

What do you think? Please comment, let me know:

• Did you try to listen but couldn’t connect?
• Do you want a stream other than Real?
• Do you know anyone who wanted to listen but didn’t know you could online?
• If you would listen from my stream!

If you are a O-E staff member, student or parent. Please ask O-E Schools and WOES to accept my offer. Or if you have any ideas please place them in the comments. If you wish to email me directly my email is brockp@mlds-networks.com

Brock Palen, O-E Alum Class 2002

Building ParMetis

This will work on 10.4.x and 10.5.x for PPC and Intel I expect. Though I have not tried it. This was done using OpenMPI

tar -xzvf ParMetis-3.1.tar.gz
cd ParMetis-3.1

Edit Makefile.in make sure CC and LD are set to mpicc. Set set INCDIR and LIBDIR to nothing.

make

This make will fail with ‘cant find malloc.h’, malloc.h is in /usr/include/sys/. Re-open Makefile.in and set INCDIR = -I/usr/include/sys/

make

For those with impatience

tar -xzvf ParMetis-3.1.tar.gz
cd ParMetis-3.1
CC = mpicc
LD = mpicc
INCDIR =
LIBDIR =
make
INCDIR = -I /usr/include/sys
make