Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Comments

top

Supercomputers' Growing Resilience Problems

Bill Barth Re:Hardly A New Problem (112 comments)

Suppose that your job is computing along using about 3/4ths of the memory on node 146 (and every other node in your job) when that node's 4th DIMM dies and the whole node hangs or powers off. Where should the data that was in the memory of node 146 come from in order to be migrated to node 187?

There are a couple of usual options: 1) a checkpoint file written to disk earlier, 2) a hot spare node that held the same data but wasn't fully participating in the process.

Option 2 basically means that you throw away half of your compute resources in case there is a failure. Basically no computations being done today for scientific research are valuable enough to warrant this approach. Some version of Option 1 (periodic checkpointing and restarting) is always more cost effective. These systems, in the US at least, are generally between 2 and 10 times over requested by the science community. Taking half away to seamlessly prevent an occasional job death just isn't worth the lost opportunity to more fully utilize the resources.

Option 1 implies taking some time away from your job to do the checkpointing. The vast majority of the time, some sort of OS-level automated checkpointing would be overkill as well. The author of the code knows better when is a good time to checkpoint and when it's a bad idea. I.e. you might consider checkpointing at a phase of the calculation when the data volume required to restore the state is at a minimum even if that means losing some part of a future calculation. Generally the calculation is cheap to redo since checkpoints of large volumes of data are expensive.

In addition, OS-level checkpointing is a hard problem. E.g. if there are messages in flight on the network, do you try to log them and be able to restart them, or do you only checkpoint when the network is quiet? If the network is never quiet on all the nodes in the job, do you throw in needless synchronization that could ruin the parallel efficiency of the job in order to find a place to do your automated checkpoint? If you decide to log instead, where do you write the log data in order to avoid catastrophic failure of each node, and what's the cost of doing it?

If these were just a bunch of VMs running a LAMP stack, this wouldn't be a hard problem. That's basically solved already. Migrating tasks for HPC jobs is truly a hard problem with tradeoffs to be considered.

about 2 years ago
top

Linux Foundation Offers Solution for UEFI Secure Boot

Bill Barth Re:So (308 comments)

Which is great unless you have 5000 nodes that you need to PXE.

about 2 years ago
top

Neutrino-Powered Financial Trading In Our Future?

Bill Barth Re:I find this depressing (275 comments)

It hasn't really done much for HPC. It has made a dent in low-latency Ethernet, but that doesn't get much use in high-end HPC (that's IB and the Cray and BlueGene networks, FWIW).

more than 2 years ago
top

UK Universities Launch Cloud Supercomputer For Hire

Bill Barth Re:Similar work exists (25 comments)

It's nothing like TG. TG systems basically gave all their cycles away for free through the work of the Resource Allocation Committee--a peer-review body that met quarterly to review proposals and give out allocations of time. This work continues through the XD program under the auspices of XSEDE.

more than 2 years ago
top

Facebook Says Your Email Is @Facebook

Bill Barth Re:There is not even a way to remove it! (346 comments)

You can't get rid of the address, but you can make it so that no one sees it. You can also display to whomever you like whatever address you like. The settings updates you have to make are pretty straightforward.

more than 2 years ago
top

Gamera II Team Smashes Previous Best Human-Powered Helicopter Flight Time

Bill Barth Re:Impressive engineering feat (118 comments)

The contest doesn't require control, and given the power requirements, it's unlikely that even the best cyclists will every fly one of these around the countryside.

more than 2 years ago
top

Linaro Tweaks Speed Up Android, By Up To 100 Percent

Bill Barth Re:battery life (97 comments)

That's not guaranteed at all. The power consumption of a CPU is a function of a huge variety of things. It's possible that while the run time is shorter, the power draw is higher--possibly more than proportionally higher.

more than 2 years ago
top

RMS Robbed of Passport and Other Belongings In Argentina

Bill Barth Re:Clarification, as I live here and study there. (386 comments)

Nobody checks your ID when you go to class in the US either, though there's much less of a culture of people just showing up listening in. It would often, but not always, be easier to detect a stranger in a class here, though there are plenty of 500-person freshman biology lectures, too. Typical classes have ~30 people in them.

more than 2 years ago
top

Go Version 1 Released

Bill Barth Re:Go has some good ideas (186 comments)

You know that Emacs has been able to let you use the tab key to insert the correct number of spaces for the current line for probably 30 years, right? I suspect that Vim can do it, too.

more than 2 years ago
top

10-Petaflops Supercomputer Being Built For Open Science Community

Bill Barth Re:56 gigabit InfiniBand (55 comments)

The 36-port part is the ASIC. The switch boxes have a lot more ports.

about 3 years ago
top

NCSA and IBM Part Ways Over Blue Waters

Bill Barth Re:Typical (76 comments)

It appears to be the latter. The spec is available here. NCSA negotiated a system with IBM, proposed it to NSF under the above linked RFP, went through a peer-reviewed awards process, negotiated an award with NSF, and started working on the delivery and other aspects with IBM and NCSA's other partners. Something went wrong in the last several months, and IBM's pull out was the result. I doubt that there is any more money to be found, and all parties knew what was asked of them in order for the project to be successful.

more than 3 years ago
top

A Closer Look At Immersion Cooling For the Data Center

Bill Barth Re:Rack density (213 comments)

Have you tried fitting a water block in a blade lately? How about 2000 of them? :)

more than 3 years ago
top

A Closer Look At Immersion Cooling For the Data Center

Bill Barth Re:Rack density (213 comments)

Given that you can lay two of these racks back to back and then run them end to end, and that you can remove most of the regular AC equipment from your room, the amount of stuff you can get in your datacenter is the same.

more than 3 years ago
top

Google Scares Aussie Banks

Bill Barth Re:"Responsive and trusted" (150 comments)

It's 3% on my Citibank card here in the US. That's the normal rate for US-based banks for foreign transactions, BTW.

more than 3 years ago
top

Supercomputing, There's an App For That

Bill Barth Re:Seems inefficient (66 comments)

You don't understand how this works. You do the computation ahead of time on the supercomputer to build your reduced order model which you download onto your phone and take out into the field. Once you've downloaded the model, you don't need the supercomputer any more. You can use the phone to do computations using the reduced model as much as you like. If you get into a regime where the predicted error from the reduced order model is too high, you can go back to the supercomputer and update the model. If that happens, then you'll probably have to wait in queue again, but that's not such a big deal.

more than 4 years ago
top

How Do You Organize Your Experimental Data?

Bill Barth Re:Use databases! (235 comments)

You might find this interesting.

more than 4 years ago
top

NSF Gives Supercomputer Time For 3-D Model of Spill

Bill Barth Re:Compute Hours? (102 comments)

Unless we've all got InfiniBand between us, it's not really worth it. These simulations are tightly coupled across the nodes they run on making them very sensitive to the latency and bandwidth between them.

more than 4 years ago
top

NSF Gives Supercomputer Time For 3-D Model of Spill

Bill Barth Re:Compute Hours? (102 comments)

I should have pointed out above that we measure in wall-clock time not CPU time. Most of these codes don't spend much time waiting on I/O, so the two numbers are usually close. We use wall-clock time because that is the time that the user monopolizes the nodes that are assigned to their job.

more than 4 years ago

Submissions

Bill Barth hasn't submitted any stories.

Journals

Bill Barth has no journal entries.

Slashdot Login

Need an Account?

Forgot your password?