23 December 2012

23rd of December

Well the world didn't end as some expected. Just because the Mayan long count calendar ran out doesn't mean the end of the world.


Farm news
As indicated in my last post the UPS died and I ordered another one. Since that post it arrived and has been installed.

I also mentioned the last of the C2Q machines getting replaced. That too has happened so I no longer have any C2Q machines. The entire farm is now i7 based. The farm currently consists of:
  • 2 x i7-970's (12 cores each) with GPU's
  • 3 x i7-3770's (8 cores each) with GPU's
  • 4 x i7-3770's (8 cores each) without GPU

As its summer down here in Sydney its too hot to run most of them without resorting to using the air conditioner so they are mainly off. A few of them (the ones without GPU's) get a burst of activity on the cooler days and then are off when it gets hot again. I am looking at solutions to this so they can be running 24x7.


Raspberry Pi
They had an update dated the 16th of December that seems to have helped a bit. You can now get the BOINC client and manager directly from the repo. I purchased a few 8Gb SD cards (ALDI had SanDisk class 6 ones for $10) so I could re-image them.

There are instructions on the Raspberry Pi download page and links to the latest image.
http://www.raspberrypi.org/downloads

You need a program called Win32DiskImager to put the image file onto the SD card if you run Windows. Once done put the SD card into your Pi and boot. Run through the setup. Once you've done that simply do the following at a command prompt:
  • sudo apt-get boinc-client
  • sudo apt-get boinc-manager

At this point you will have 7.0.27 and can attached to the projects (not that any of mine currently support the Pi directly). Projects that are looking at supporting the Pi that I know of are:
  • QCN (Quake Catcher Network) - You'll need a sensor
  • Radioactive@home - You'll need a sensor
  • Einstein@home
  • Asteroids@home

The first two both need a sensor and already have an app but don't offer native support for the Pi. Einstein has an app but are looking to reduce the memory required and the size of their work units for the Pi. Asteroids are looking at porting their Linux app to the Pi.

After all this I still haven't seen the first Pi that I ordered. According to the forums on the Raspberry Pi web site Allied/RS Components have filled all back orders and have stock. They have been upgrading people to the revision 2 board which has 512Mb memory. The revision 1 that I have only has 256Mb memory. Hopefully it will arrive soon.

The 256Mb Pi that I have keeps losing its internet connection for some reason. Discussions in the forum suggest it can be caused by using a USB hub and/or a Logitec keyboard (I had both). I have tried using other keyboards and even no hub but that didn't fix anything, so it might be a hardware fault.


BOINC news
We got 7.0.42 last week. It has support for intel HD2500 and HD4000 graphics that are built into the 2nd and 3rd generation i5 and i7 CPU's. Its OpenCL capaable so BOINC should be able to make use of them. Of course its waiting on the projects to add support, but given a number already have OpenCL apps that shouldn't take too long.


Parallela
They've almost finalised their board layout and will soon be placing manufacturing orders soon. So its looking good for a May delivery. I won't expect it to be that fast as the Epiphany will only have 16 cores. It is another one of things to experiment with to see if its any good at crunching. At least it should be better than the Raspberry Pi.


Surplus hardware
I still have a pile of old graphics cards I need to dispose of. Nothing wrong with them just that they aren't the current generation so I may as well get some money for them and make some room. Its a bit of nusance using eBay due to the fees they charge. If you are interested drop me a line and we can talk price. Starting at the most recent I have:

2 x Palit GTX570 factory overclocked in original box
13 x GTS450SP these are single-slot cards not sold in Australia. Elsa brand from Japan
2 x Palit GTX295 (rev 2 boards) no packaging
4 x Palit GTS250 no packaging
1 x Radeon HD4550 no packaging
2 x EVGA GT240 no packaging

I also have a couple of intel CPU's
i7-920 (8 cores - ie 4 cores with hyper threading) socket 1366 chip only
i5-3550 (4 cores) socket 1155 in box with intel heatsink/fan

Email for the above surplus items is:
augur tbbs [ @ ] yahoo [ dot ] com
Strip out the spaces and brackets and put a dot in to make it a valid address

09 December 2012

9th of December

This week saw another Core 2 Quad replaced by a 3rd generation i7. There is one left to replace next week and that should be the upgrades for a while. The i7's provide more CPU cores, run faster and use less watts. I am using them as fairly lean (on wattage) CPU-only crunchers.

The UPS on the file server died during the week. The loft stinks of ozone as some component inside has obviously blown up. I awoke on Friday morning to a frantic beeping coming from it and a smelly loft. A new one has been ordered already.


Project news - Asteroids@home
This project has been running for a while but until this weekend only had a Linux app. This weekend the released an application for windows users. They try to map the shape of asteroids using a lightcurve method. Their web address is: http://asteroidsathome.net/boinc/index.php

I had Linux running in a virtual machine on one cruncher for the past couple of months, just for this project. Now I can remove it and use the other machines with a native windows app.


Project news - Seti@home
They had a major crash last week as a result of a local power failure in the Space Sciences Lab at UC Berkeley. It took out one of their older machines. Their science database was also compromised. They didn't manage to shut down their main database machine or the replica database machine before the UPS batteries went flat.

It has taken them a week to restore from the last backup and roll forward. They then had to verify the integrity of the database which took a few days due to its size. You can read about it in this message thread if you want to know the details: http://setiathome.berkeley.edu/forum_thread.php?id=70199

We are back to the usual maxxed-out network connection now. Due to the scheduler time-out happening for the last fortnight they have applied hard-limits of 100 tasks per CPU core at any one time. This should mean more people can get work and those more "hungry" computers won't be able to request thousands of tasks at a time.

29 November 2012

29th of November

Last weekend it was off to the computer shop with another one of the Core 2 quad machines and I brought back a P8Z77 with a 3rd generation i7 and 8Gb of RAM. I still have 2 more to replace.

Basically we're cannibalising the HDD, DVD and power supply to make them. These last three machines still work, its just that they use about 100 watts of power but aren't that fast. The replacements are using about 80 watts and are pretty fast, not to mention having twice the number of cores.

I am deliberately not putting GPU's in these as I want them to be fairly energy efficient although they could be easily upgraded later.


Project news - Seti
Last week their scheduler was having issues and in the end they disabled it. This week they have moved the splitter processes off to another machine in an attempt to improve things.

They are still pushing their Multi-Beam v7 science app in the hope that the extra processing will slow down the work requests but I think all that will do is buy them some time (once they get it rolled out) until the CPU's and GPU's catch up again. It will take a while before the updated science app can be rolled out.


Project news - Einstein
They passed the 800 Tflop processing barrier. This is probably because of people switching from Seti to a more reliable project like Einstein. Of course Einstein has much better funding than Seti.

17 November 2012

17th of November

As I mentioned in my last post I was going replace the old Core 2 Quad machines with i5's. I did one of them (the one that had a failed motherboard). While browsing a suppliers web site last week I find that they have i7-3770's at a cheaper price that I paid for the i5. The only reason why I got the i5 was because they were supposedly cheaper.

Anyway I have ordered a couple of them and they should arrive during the week. I will be back at the computer shop on the weekend to get the i5 swapped over to an i7. No I don't do them myself after killing two KFSN4 motherboards. I am happy to do other bits like install Windows, change drives, etc, just not the CPU's.


Project news - CPDN
ClimatePrediction are not giving out work and won't be for some time. They are planning to upgrade their BOINC software to the current version. Unfortunately theirs is a fairly old version so it will take some planning and testing. They have over 100,000 work units out in the field, but given they have 1 year deadlines on some tasks it could be a fair while before reissues of failed work units happen.


Project news - Einstein
They are coming up to the end of their current Gravity Wave run so there may be shortages or work available. They have also suggested they will be overhauling the website soon.

There is a notice that their test site (albert@home) will be effected by a planned power outage with einstein@home potentially also being effected on Sat 17th of November from 7am to 2pm local time.


Project news - GPUgrid
They are trying to get to a cuda 4.2 app only. Unfortunately they are having problems getting some people to upgrade their drivers. This primarily effects users of the older cards (ie the GTX200 family) as cuda 3.1 seems to work better for them. They may keep one work queue running cuda 3.1, hower their cuda 4.2 app can do some additional science that the older app cannot.


Project news - Seti
After a good run last week they are back to struggling mode again. It seems there is some issue where the scheduler times out when the Astropulse splitter is running at the same time. Requests can sometimes work if you set No new tasks so that it reports current work and then allow it to fetch work. That is you seperate requests to report completed tasks from requesting work.

They still have the issue with the network being maxxed out so nothing can download anyway. It gotten so bad that I don't bother with Seti any more as it takes too much hand-holding just to get work.

If that wasn't bad enough they have cut back the hours the staff can work so they aren't around as much to help out with any issues.

04 November 2012

3rd of November

One of the old Acer Veriton V7900 machines died a while back. I decided it was time to replace it. Unfortunately the Acer case wasn't really suitable as it has next to no airflow, so I purchased the Fractal Designs ARC midi case. You can see it alongside another Acer.
 
 
 
Here you can see them from the top. As you can tell the ARC midi is a bit taller, a bit wider and a bit longer, but not by much.
 
 
 
Here is the inside of the machine. Its another Asus P8Z77 motherboard but this time with an i5-3570 CPU as it was replacing a Core 2 quad. Its got the same Noctua U9B cooler as used on the other P8Z77 builds and the same memory. I would have reduced the memory but it seems the distributors don't want to sell 4Gb Kingston Hyper X kits so it got an 8Gb kit. I reused the old hard disk and Seasonic power supply.
 
 
As you can see its got lots of room and a few 140mm fans. The idea was to have something fairly energy efficent so this one is running using the built-in graphics.

29 October 2012

Parallella

Remember a while back I mentioned Adapteva? They were making a 64 core RISC CPU chip. Well they completed that a couple of months ago. The problem with their development kit was it cost $10,000 and the board was fairly large.

Following in the footsteps of the Raspberry Pi they decided they would make a $100 version of their development board with a dual core ARM processor and the 64 core Epiphany IV processor. It is to be called the Parallella. They started a fund raiser on KickStarter with a goal of $750,000:
http://www.kickstarter.com/projects/adapteva/parallella-a-supercomputer-for-everyone

They exceed their goal by the deadline (Sat 27th of October). Of course yours truly made a pledge. Now we wait for Adapteva to get their development board shrunk and make them available. In return I should be getting a couple of boards.


Farm news
This weekend I spent a bit of time cleaning computers. You know vacumn cleaner and a paint brush type of cleaning. I was going to replace the 140mm fans in the bottom of both of the GPUgrid crunchers, however the Noctua 140mm fans are round so they don't fit into the bracket. Instead I replaced the case fan at the back of the case. They were already Noctua fans but 120mm. I also prefer to use the rubber mounts supplied with the Noctua instead of screwing the fan onto the case as this reduces the vibrations (and noise).

The temporary fans I put into the file server were also removed and put into the spares box. They were swapped out for Noctua fans which has cut the noise down.

I dug out the old file server. Its a Pentium @ 1.8Ghz. It was too slow for number crunching but had a nice case so I put a couple of network cards, a RAID card and a couple of 500Gb hard disks in it about 5 years ago. Its been sitting in the corner for the last 2 years. I powered it up, let it download its 157 windows updates and apply them. It was having a problem with the network card which I tracked down to a 2nd network card not working. Removed the 2nd card and threw it away. Suprisingly the rest of it still works. The CPU heatsink was clogged up with dust so I removed it and gave it a clean and reapplied thermal paste before putting it back. I also had to swap the battery on the motherboard as it had gone flat. It should be right for another couple of years. Why waste the time on it you ask. Well its the backup file server in case the new one dies. So I keep it in a working condition just in case.

I took one of the old Core 2 quad machines which had blown capacitors over to the computer shop. No point in trying to fix it, but I was thinking I might be able to replace the motherboard with another P8Z77. In order to keep the cost down I will try and reuse the case, power supply and so on. I might get an i5-3570 instead of i7-3770 and a lesser amount of memory. Anyway we will see what pricing the shop comes up with. My main concern was if the case was large enough, although they did find the front-panel connector was different to the usual motherboard pins.

24 October 2012

More on the Raspberry Pi

It seems the B model of the Pi has been updated to 512Mb memory. Hopefully the Pi that I have on back-order will turn up with this as the one I have only has 256Mb.

I've been doing a bit of searching to see if anyone has a compiled Seti app. It seems a few people do but none are in the repo. In fact if you do a sudo apt-get install boinc-app-seti it returns an error saying it can't find it but its referenced.

The other issue is there is an armel version and the Pi should use the armhf (ARM hard float) version so the apps need to be compiled for this platform to get the best performance.

The good news is you can download it manually from the Debian pakages page:
http://packages.debian.org/sid/armhf/boinc-app-seti/download

I had to download it to my PC, stick it on a USB memory stick and then copy it across to the Pi. Once its there you'll need to unpack the files into the appropiate directories. The File Manager that comes bundled with the Pi  will allow this.

20 October 2012

20th of October

Again not much to report this week. The weather turned hot so the farm was off for a couple of days. This of course makes a fair dent in the amount of work that can be processed. I give priority to the machines doing GPUgrid as I figure medical advances are more important than astronomy.

Today I finished installing all the GTX660's in the various machines. One each for the 12 core machines and last week I put two into one of the new 8 core machines. I also had to move a couple of machines around to fit the KVM cabling. The 12 core machines have PS/2 keyboard and mouse but the newer machines don't so I need to use USB connectors for them.

The file server was making various noises at the beginning of the week. This turned out to be a case fan that had died and was no longer able to spin up. It was supplied with the case, a Fractal Designs one. I had a couple of old CoolerMaster fans sitting in a box so put them into it temporarily. I have ordered a few Noctua fans. When they arrive I will remove the CoolerMaster fans (they are noisy and also tend to fail after a couple of years).

I managed to find a nasty little bug with the BOINC Manager by accident early in the week. The result is it will lock up the manager and use a whole CPU core. It then needs to be terminated by Task Manager. The developers have been made aware of how to do this and have reproduced it, so they will get a fix out soon.


Project news - Seti
They have commenced their annual fund raiser. The money they raise is used to fund the project rather than specific bits of hardware, which is what the GPU Users Group concentrate on. I will be making a donation, but probably not as big as usual seeing as I gave at the office, I mean the last hard disk fund raiser. As it is I have donated a 5 figure sum since joining Seti@home in 2008.


Adapteva on KickStarter
Adapteva are seeking funding to make their 64 core processor more readily available. This is a nifty way where people can give money (from $10 up) towards their target and as many people can do so as it takes to reach their target. Their target is $750,000. As I type this they are on $405,000. It closes on the 25th of October. The link to their project on KickStarter is here: http://www.kickstarter.com/projects/adapteva/parallella-a-supercomputer-for-everyone

06 October 2012

6th of October

Farm News
Not a great deal to report this week.

Two of the four GTX660's have been installed in one of the P8Z77 machines. The good thing is the P8Z77 has two of the PCIe slots triple spaced, that is there is enough room to put two double-width graphics cards in there and still have a gap between them.

We had a couple of hot days this week, so the farm was powered off for two days. Its spring in Sydney so its probably going to be a long hot summer. Not good for number crunching.


GPUUG fund raising
One of the users has decided that for this weekend only he will match any donations up to a limit so the Green Bank transport fund raiser can be finished off. Well not one to miss an opportunity I made a donation and so have a number of other people. These donations have now completed this fund raiser.


Project news - Asteriods
This is a newish project that is calculating the shape of asteroids from multiple observations. Their science app is written for Linux, so I have setup a Virtual Box machine in order to run some of the work units. They were looking at using Vbox from within BOINC, but have now decided they will port their app to Windows (which represents about 60% of the user base). This is good news as BOINC running in Vbox is not as efficent as running it on the host operating system.


Project news - CPDN
Climate Prediction had one of their servers run out of disk space a fortnight ago. That then meant the upload server also filled up its disks. Its taken two weeks for the project staff to move the files around so we could get the uploads through. The climate models produce large files and with everybody uploading work its a constant battle just to clear space on the upload servers.

The new i7-3770 CPUs are much quicker, so much so that they get through a regional climate model in around 60 hours. The older i7's were taking 100+ hours. I have decided to only run CPDN on the new machines.


Project news - Seti
They have been having a lot of trouble with network bandwidth. This was made worse by all the short work units which were taking longer to download than run. They have moved onto some longer running work units so things have improved a bit. They are trying to get the v7 multi-beam app out, this will do an additional calculation thus taking longer to process and that will reduce the demand for work. Unfortunately its taking longer to get this app ready for deployment.

27 September 2012

GTX660

This is the Palit GTX660 (no not the Ti). I bought these to replace the 560Ti cards. Its a factory over-clocked card.
 
In the cardboard box is a driver disk, power cable, DVI adaptor and the card itself. Its a bit shorter than the 560Ti.
 
 
 
This is a side-on shot. I was trying to show the heatsink in there.
 
 
 
And here we are installed in one of the Asus P8Z77 machines.
 
 
 
Here are the GPU-Z screen shots.
 
 
 
 
You'll need Nvidia driver 306.23 at a minimum as older drivers don't recognise the card.
 
Initial runs have been on Seti using Jason's x41z (CUDA 4.2) app. After that I let it loose on some Einstein work (CUDA 3.2).
 
The Seti work seems to be a bit faster. To be fair they have all been short work units recently so its hard to tell as they only run for 2 or 3 minutes. The Einstein work showed much better speed. I must stress that I have only run a few work units through it at the moment so its a bit early to tell, however Einstein appeared to be around 30% faster. The crunching will continue.

21 September 2012

GPUgrid 100 million

I finally hit 100 million for GPUgrid. Below is my certificate. Crunching continues...
 
 
 
HD 7770
The old HD4850 that I have only supports OpenCL 1.0 and none of the projects use 1.0, so I bought a HD7770 during the week. The idea was it could run OpenCL apps as they are supposedly pretty good. Well they aren't.

I put it into one of the Asus P6T's which already had a GTX560Ti. After a bit of stuffing around to get OpenCL installed its ready. I start up BOINC and it gives a theoretical figure of 1,382 GFlops for the GTX560Ti and 3,520 for the HD7770. That suggests the 7770 is around 2 and a 1/2 times faster than the 560Ti.

I allow Einstein to download some work, the 560Ti gets 2 BRP cuda tasks and the 7770 gets 2 BRP OpenCL tasks. The 560Ti completes them in 2,000 seconds and the 7770 does them in 5,293 seconds. Okay its not a fair comparison as we are comparing a CUDA app with an OpenCL app. So to be fair I get some Milkyway work. They run OpenCL on both. The 560Ti takes 320 seconds and the 7770 takes 373 seconds. While the 7770 is theoretically faster, in real life it isn't.


Upgrades
As you would have noted in the last couple of weeks I have been updating some of the motherboards and prior to that I replaced the GTX570's with GTX670's. Now they have been completed that leaves the remaining GTX560Ti's to be replaced. I have been looking at the GTX660 (not the Ti). That would bring the peak power down from 170 watts to 140 watts and should give a performance boost as well.

15 September 2012

Skeleton Adventure

I finally found a suitable case for the Adventure. This is is it. An Antec Skeleton. In this shot you can see the Adventure 2800 with the power supply underneath.
 
 
 
This is the supporting bar that the back plane of the cards screw into. The black mark I have drawn is where I need to cut it to accomodate the PCIe bridge card. This is in about the 11th slot if one had a normal motherboard.
 
 
 
And here it is again plugged into one of the Asus P6T's. I only have two of these left after recent upgrades to Asus P8Z77 motherboards. The short green card just below the fan is the other PCIe bridge card.
 
 
 
In this picture I am looking down on the Skeleton. You can see the 4 graphics cards starting from the left. You can also see the clear plastic supporting bar across the top.
 
 
 
Another shot, which looks a bit clearer, where you can see the graphics cards.
 
 
 
At the moment I have 4 of the GTS450SP cards installed in it. These may not be its final configuration but they will do for now. Indeed it might not be driven by a P6T even. I am looking at getting another P8Z77 to drive it.
 
I have run some Seti and Einstein work on it since installation. I had some issues with the Einstein GPU tasks erroring out when running 2 per card. I reduced them to 1 per card and they appear to be working fine now although it needs some further testing. Seti tasks seemed to run fine (using the x41z app from Jason) running 2 at a time.
 
 

12 September 2012

Raspberry Pi info

It seems others are also interested in using the Raspberry Pi. I have mine up and running. I've even got BOINC installed on it. Unfortunately there are no projects currently supporting it at the moment.

You have to remember the Pi only has 256Mb of main memory. Mine has 70Mb free when running BOINC, so if the science apps can fit in that then they should run. Its not that fast, so you'd want something that isn't too complex for it to run. I have been looking at the Seti Multi-beam app, which might just squeeze in. There isn't one in the repo, so I have to get the source and compile it myself.

If people ask their favorite projects for science apps they may look into it (Asteriods @ Home has said they will) assuming the hardware isn't too limiting. That means there is no point in asking over at CPDN as their climate models are huge and take a long time to process. Look for an existing message thread or start a new message thread on the project message boards to get the ball rolling. If enough people ask then the projects are more likely to do something. If projects are supportive we can get native BOINC support rather than needing to use an app_info file (also known as the anonymous platform mechanism).

Its possible to run the Pi in command line mode (ie not run X-Windows on it) which should free up a bit of memory. Exactly how much I don't know. Terminal (the program) seems to use 91Mb, which seems rather excessive. Not running it should also free up some memory.


Where do you get it?
The other question I have been asked is where I got mine from. Well I ordered two. The Raspberry Pi website has two suppliers listed. I ordered one from the 2nd link (Allied Electronics and RS Components) and after ordering found out its on back-order for up to 13 weeks. After that I went to the other site (Element 14) who have a local presence here in Sydney, Australia and they had 800 in stock. It arrived about 2 days after ordering, with the case being delayed by a week.

06 September 2012

6th of September

A quick update since the last post.

Raspberry Pi arrived on Monday. Unfortunately I didn't have the right cable to plug it into power (it uses a Micro-USB B socket). Fired it up on Tuesday night, but haven't connected it to the network to see what sort of software is in the repo for it.

I experimented with Linux Mint 13 running in a VirtualBox VM so I could run Asteroids@home work (they only have a Linux app). Got it working only to read an announcement today that they will be creating Vbox work units so there is no need for this image now.

Two new WD 5003 Black drives arrived on Tuesday (SATA III and 64Mb cache). Swapped out the hard disks in both the new machines with these and reinstalled Windows. The windows benchmarks think a WD 5003 is the same speed as a WD Blue (SATA II and 16Mb cache).

Returned the second failed WD Black drive for repair/replacement under warranty.

Dropped off a 3rd Asus P6T for upgrading to an Asus P8Z77 machine. That will bring the farm up to 3 of these P8Z77 machines. I should be picking it up tomorrow.

Sold all 3 Asus P6T motherboards, heatsinks and memory.

02 September 2012

2nd of September

So this is what I get up to on my weekends...
 
Yesterday I took the two Asus P6T's with i7-920 CPU's off to the computer shop. I came back with two Asus P8Z77-V LE Plus's with i7-3770 CPU's. You can see one of them below.
 
Unfortunately they couldn't get a kit for the old Thermaltake 120mm heatsink so it got replaced by a Noctua NH9 (92mm) with 2 fans in a push-pull arrangement. I also replaced the memory as the old motherboard uses triple channel memory and the new one is dual channel.
 
 
 
Out of this I have gained a faster CPU that uses less watts, more memory and the newer graphics cards now have a PCIe 3 slot giving faster transfers.
 
On the down side I had to reinstall Windows due to the chipset drivers being totally different and the hard disk in one of them failed (its still under warranty). I also have a problem with the KVM cabling as the P8Z77 has only a single PS/2 port (it can be keyboard or mouse) but the KVM expects both. I will need to buy some USB KVM cables.
 
As the new motherboard has a couple of SATA III ports I have ordered two new hard disks, same size as before but SATA III and with a bigger cache.

27 August 2012

27th of August

And here is the news for this week:


Farm news
The farm is running fairly well. I am looking at some machine replacements fairly soon to counter the huge electricity price rises we've had in the last 2 years.

I have two machines (Asus P6T) that are mainly GPUgrid crunchers running 24/7. I recently replaced their GPU's and now I am looking at replacing the motherboards. I am looking at the Asus P8Z77, Intel i7-3770 and 8Gb Kingston memory. This should bring the CPU power consumption down a fair bit. I will reuse the existing bits (case, power supply, etc) to keep the price down.


Project news - Einstein
They have discovered 7 new pulsars. These all seem to be from their Binary Pulsar Search work units. The other types of work units (Gamma Ray and Gravity Wave) don't seem to be finding anything.


Project news - Seti
Seti has been releasing their Astropulse for GPU app. They have currently released versions for the ATI and Nvidia cards with OpenCL support and BOINC 7.0.27 or later.

There are a couple of versions yet to be released that will work with older BOINC clients. Personally I wouldn't have bothered with them, the whole idea would be to give people a reason to upgrade to the later BOINC client. Its not as if they need all the extra computing power for Astropulse because they produce them at a ratio of 40:1 (that is 40 multi-beam work units to 1 Astropulse work unit). Astropulse work units are hard to get normally, now they are almost impossible to get.


Rasberry Pi
These are a small (credit card sized) computer that cost about $35 each. The are based around an ARM processor. Originally developed for the education market in the UK with an emphasis on being cheap. You can find details of them here: http://www.raspberrypi.org/faqs

Well it seems both QCN (Quake Catcher Network) and Asteroids@home are interested in them. They run Linux and BOINC has been ported to them. Both projects are looking at creating an app to run on them.

20 August 2012

20th of August

There was a bit of a delay in posting as I went away skiing last weekend, Here is the latest news:


BOINC testing
We're on 7.0.33 at the moment. There was a change to the FLOPS (floating point operations per second) estimate for GPU's which can cause issues if you upgrade while you stilll have GPU work. The estimate is now 10 times the CPU FLOPS as a default. Before it was set to the same value as the CPU. If you have existing work it will usually get an "estimated time limit exceeded" error. Its best to finish all GPU work before you upgrade to this version.


GPU Users Group
Terabit Systems donated some 10 Gbit networking gear to Seti@home. A big thank you to them for this equipment and the guys from the GPUUG for organising this. Comms within the server closet should improve quite a bit once this equipment has been installed.


Project news - CPDN
ClimatePrediction have now recreated most of their work units after their problems a fortnight ago. There are quite a few thousand work units available in the regional (EU or Pacific North West) series. They have had a few issues with filling the upload servers and then have to take it off-line for a couple of days while they make room.


Project news - Einstein
They have been updating their OpenCL app to give it improved speed.

They also had some network changes happening so appear as being off-line at random times during the week. The network changes are all over now so things are back to normal.


Project news - Seti
They had a problem with their version of the server code when they inadvertently removed references to CUDA work. This had the effect of running multiple instances of GPU work units (like 10 or 20) on pre version 7 BOINC clients. Newer BOINC clients worked fine, but those sticking with the old versions had the issue. In the newer clients its referred to as ATI or NVIDIA work. They have since fixed the issue.

In addition to that little problem they also had a 2 day outage due to scheduled electrical work in the Space Sciences Lab. That meant they were off-line for two days. After it came back on-line they weren't giving out GPU work while they worked though the CUDA issue. Its back to its totally maxxed-out 100 Mbit network connection at the moment.

03 August 2012

3rd of August

Today I passed 80 million credits for GPUgrid. Below is my certificate. It seems the last one I took was 30 million back on the 18th of December 2010. On the rankings by RAC (Recent Average Credit) I am now up to position 28 for GPUgrid.
All this can be attributed to two things:
  • Their new CUDA 4.2 science app
  • My two (not so new now) GTX670 cards




 
BOINC testing
We got 7.0.33 yesterday. Unfortunately over at Seti it seems all the existing GPU work units fail with "estimated time limit exceeded" errors. There were some changes to the estimated time calculation in this release so its quite likely related. Until this issue is thought out its probably best to stay away from .32 and .33 for the time being.
 
It you are a Mac user then 7.0.31 became the recommended version. This is because there are a number of installer changes to support OSX 10.8 (Mountain Lion).
 
 
GPUUG news
Things have been fairly quiet because the Seti project staff were rather thin on the ground. However there is possibly a new project to consider. We're awaiting the Seti staff to let us know what they'd like to tackle next.
 
The Transport hard disk fund raiser has a fair way to go yet to finish off and get the additional disk drives the project needs for transporting data from telescopes back to the Space Sciences lab at UCB.
 
 
Project news - CPDN
Climate Prediction had problems during the week. It seems their scientists submitting the thousands of new work units have omitted to send some of the files that are used to make up a work unit. When they get to processing the 2nd month they all fail (each work unit represents 1 year).

After a number of complaints of failed work units the project cancelled all of the work units awaiting to be sent. Presumably they all had this problem. So now we're out of work and hoping they will generate new work units that contain all the necessary files rather quickly..


28 July 2012

28th of July

The farm continues to make steady progress. The GTX670's have been running non-stop on GPUgrid and as of this post have climbed to position 34 on the project rankings by RAC (Recent Average Credit).

CPDN
Climate Prediction has been issuing lots of work units in the last week. The last time I looked they had 22,000 EU and 9,900 PNW (Pacific North West) regional models available. This is in addition to the 137,000 models that are out being processed at the moment.

I wonder if the plumbing will be able to cope. That is BOINC and the server hardware they have, as they have a history of server issues particularly around data storage. Their work units are large and produce 12 trickles of 7Mb each and a final file that is typically 32Mb for the PNW models.

FPGA blog
Last week I mentioned the blog started by Terry trying to get FPGA support into BOINC and some science apps converted. I sent him a note during the week, which he has quoted here: http://www.fpgaathome.org/2012/07/26/message-to-fpga-manufacturers/

He has started on trying to get the Seti@home multibeam application running (or at least parts of it) on an FPGA. If anyone is interested in helping you can contact him via his blog (leave a comment) or send an email to terrystratoudakis (at) gmail (dot) com


GPUgrid
We had a little drama on the project this week. One of the new scientists to join the project issued some rather unusual work units which were all failing. BOINC then starts to reduce the amount of work you can get because your host is returning errors and eventually you get down to being only allowed 1 work unit per day until you return a valid result.

The offending work units were flushed out of the system and corrected ones issued. From reports these new ones take a long time to process. I haven't received any of the corrected ones yet so can't say by how much.

19 July 2012

19th of July

Further to yesterdays post about coprocessors. There is a blog regarding using FPGA's for computing and specifically integrating them into BOINC. Its called FPGA at home and can be found at http://www.fpgaathome.org/

In his post of 18th of July he links to an Xcell article in which they state that they used the XC6VLX240T with a MicroBlaze processor. Looking at the recommended pricing on the Xilinx web site for the board and processor it has them costing approx $2500.

I am not against using FPGA's for computing, in fact I think they should be used. If it can do maths when why wouldn't you use it? Its just that the makers seem to have priced them out of reach. Until they decide to get serious the likes of AMD/Intel will control the CPU market and ATI/Nvidia will have the GPU market to themselves.

18 July 2012

18th of July

Not a great deal of activity on the farm this week. Things have been chugging along without any incidents (touch wood).

The GTX670's have managed to get me up to position 36 based upon RAC (Recent Average Credit) on GPUgrid which is what they spend their time processing. Considering I was down around position 150 or so about a month ago this is pretty good. More importantly they are contributing to some useful science.

BOINC testing
We're still testing 7.0.31 which has a few minor changes since 7.0.28. I have reported one issue to do with project initialisation not honouring the exclude_gpu preferences.

Coprocessors
I have been making some inquires with a number of manufacturers regarding coprocessors. I had initial email discussions with Adapteva who have a multi-core CPU, ClearSpeed who make a maths coprocessor and DRC regarding their FPGA offerings. Unfortunately all seem to have priced themselves around $10,000 which makes them too expensive for volunteers to use.


Future plans
At the moment we're still waiting on Nvidia to release a GTX600 series card to replace the 560Ti. I have three GTX560Ti cards so I am keeping an eye on announcements. The main reason is electricity prices went up by another 18% this month, with the carbon tax getting the blame, and my electricty bill is getting rather large now.

I still need to find a case for the Adventure. Once I have one I will also need a machine to run it and might look for a relatively cheap AMD based motherboard. It doesn't need much CPU power as all it needs to do is drive a PCIe x16 slot to feed the Adventure.

04 July 2012

4th of July

The farm continues to run. 4 of the 5 i7's have been running constantly for the last couple of weeks. The two GTX670's have managed to get their RAC (Recent Average Credit) up to 317,000 for GPUgrid (which is all they are running) with maybe a little higher to go before they plateau.

Unfortunately electricity prices have gone up again, with the introduction of the Carbon tax and having to fund infrastructure. The last quarterly bill came in at $650 which is mostly the computers. Just as well I upgraded to the GTX670's as they should save 78 watts per hour from the cards they replaced. Once an equivilent of the GTX560Ti is available as a Kepler-based card I will look at replacing all of them as well, if only to reduce electricity use.

BOINC testing
This week we got to 7.0.31 which has a few minor tweaks since 7.0.28, but nothing major in it. I have it on one machine and its doing fine so far. Speaking of 7.0.28 it has now become the recommended version, so if you are running an older client I would suggest you upgrade.


New project - Asteriods@home
This is a new project. It is still in start up mode. They are computing the shape of asteroids from multiple sightings. They currently only have a Linux science app. Once the bugs have been ironed out they will port this science app to Windows.


Project news - Climate Prediction
They had a 3 day outage because of network problems connecting to their DB machine. When they came back online yesterday they also had some new work for the regional models, about 1800 work units. I managed to manually get 3 of the machines to pick up work and by the time I got to the 4th machine it had all gone. They have a 1 hour backoff after a scheduler request so it took some time to get them all to make work requests.


Project news - GPUgrid
Ignassi, one of their project staff had his docterate defence yesterday. He is now known as Dr. Ignasi Buch. Unfortunately that also means he is leaving the project, so we wish him well.

Their CUDA 4.2 app is now working well with run times reduced by almost half, so what used to be 8 to 12 hour work units are taking 5 hours. I am not sure how much of this is due to using CUDA 4.2 and how much is due to the newer graphics cards, but either way its impressive.

24 June 2012

24th of June

Farm news
As you would have noted from my previous post I got a pair of GTX670 graphics cards. They have been installed and have been running GPUgrid pretty much non-stop since. I haven't had any issues with them, simply swapped out the GTX570's, reinstall drivers and off they go.

I also did a bit of cleaning. The UPS which looks after the file server was given a good clean out (full of dust due to having a fan that runs flat-out all the time). Once I got it all put back together and turned on I get a loud rattling sound from the shelving unit. I tracked it down to the file server which sits above the UPS.

So I take the file server off the shelving unit and pull one side off and have a look around - nothing to see. I put it back together and power it up while its sitting on the floor, all normal so back on the shelving unit. Power it up again and the rattle returns. I take it down again take the other side off and have a poke around, notice the heatsink is covered in dust, give it a quick brush and blow out and then find the fan at the bottom of the case is loose. I tighten its screws up and off we go again, all quiet this time.


GPUUG news
I made a donation to the UPS fund raiser and finished it off. They have now ordered the parts and shipped them to the Seti@home lab.

The hard disk transport package is still going and looking for donors.


Project news - GPUgrid
They have been beta testing their CUDA 4.2 science app. This has now been rolled out to their "normal" users with up to date drivers. This was good timing for my recent GTX670 installation. I don't get any long work units at the moment so it hasn't been applied to all types of work (they have beta, normal and long work queues) but I expect that will be updated soon.

20 June 2012

GTX670

The farm now has a couple of Nvidia GeForce GTX670 cards. These ones are the Palit Jetstream version. They are factory overclocked with a dual-fan cooling solution and 2Gb of memory.
This is the box they come in, complete with a carry handle on top (why?)

And this is what you get for your money:
  • The GTX670
  • A quick installation guide
  • Driver disc
  • A molex to PCIe power adaptor cable
  • A DVI to VGA adaptor
  • A HDMI to DVI adaptor

Here is an end view. As you can see its a bit wider than 2 slots.

Here is a view from the top. You can see the dual PCIe power connectors towards the middle of the card.

And here we are installed in one of the crunchers. This one is replacing a GTX570 (also a Palit factory overclocked version).

This is what GPU-Z has to say about the card

And here is the Sensors tab while its under load. As you can see its only using 48% GPU load and thats while running a GPUgrid work unit using CUDA 4.2, although to be fair different work unit types give different loads.

16 June 2012

16th of June

As you would have seen from the 2 previous blog posts I spent a bit of time last weekend testing the Adventure. I still can't find a case for it. Its currently packed away again until that gets sorted out. I can't believe nobody makes them over here in Australia, not even a custom built one!


BOINC news
In BOINC land they have been busy switching from using subversion to GIT, so nothing much new development wise.

I have asked for the GPU enumeration to be expanded as BOINC is currently limited to 8 devices but the drivers allow for 64. Dr Anderson has agreed to look into this. We can't stuff that many GPU's into a single machine but some people have managed to put more than 8 in a machine, even I can do that using the Adventure.


Farm news
I have ordered a couple of GTX670 cards which will be used to replace the GTX570's. The 570's will be up on eBay in about a week or so if anybody wants them. Hopefully GPUgrid will be releasing their CUDA 4.2 app to the mainstream fairly soon as it is still in beta testing.

I have been testing the x41x Seti GPU app from Jason G. Its built using CUDA 4.2 and seems to be faster than the older CUDA 3.2 x41g app. It also seems to be more reliable as I haven't seen any -12 errors that pop out from time to time with the older app. This also fits with future GPU upgrades as Kepler-based cards seem to need apps built using CUDA 4.2 to get the expected performance boost from having so many CUDA cores.

Crunching for the other projects continues. I have finished all the Climate Prediction work units that I had on various machines. They now are out of work. In the mean time Einstein is making steady progress, but I am yet to find a Pulsar.


GPU Users Group news
The GPUUG have got 2 current fund raisers going:

  1. UPS - It consists of 6 x UPS and 5 x surge-protected power strips.
  2. Green Bank data transport package. It consists of 3 x hard disk transport cases and 20 x 2Tb hard disk drives. A couple of the cases have already been supplied but they haven't raised enough for the hard disks yet.

Project news - Seti
In the last week or so Seti had a number of power issues to their building that resulted in one server "Thinman" and a couple of new work stations getting fried. This is why the GPUUG have started a UPS fund raiser.

11 June 2012

Adventure testing (part 2)

So part 2 of this saga...

Yesterday I used one of the other P6T's to test with. Points to note:
  1. It helps if you put the cards into the correct slot (start from the opposite end to the bridge card)
  2. It likes to have a monitor plugged into the 1st card (the one nearest to the bridge card).

While opening and closing the case many times yesterday the hard disk in the machine died. Just as well I had recently put an SSD in there and backed up the hard disk before starting. Fortunately I have a couple of spare WD Caviar Blue drives I was going to upgrade machines with, but this one had a Caviar Black, well it did until it died yesterday.A quick instal of a new drive and copy the files back and it was up and running again with no work units lost.

Today its Maul's turn. You can see Maul being used with the Adventure and a single GTS450 SP at the bottom of the screen. The power supply is on the bottom right


Another shot of the Adventure and power supply sitting beside the P6T.


Here I have loaded up 4 of the GTS450 SP's into the Adventure


And just to see how it handles it another 2 GTS450 SP cards in the P6T as well


A screen shot of BOINC running a few work units on the GPU's
So this leads to an interesting thought, what if I were to get 2 Adventure's and plug both into one of the P6T's. Say 4 cards in each like I have done in this case, that would mean an 8 GPU machine is within possibility.
The GTS450 SP isn't a particularly powerful card. They are probably better running only a single task per GPU, but they only use one slot which means it can have some room either side for better air flow. They have also been superceeded by the newer Kepler-based cards.

10 June 2012

Adventure testing (part 1)

This weekend is a long weekend, Monday is the Queens birthday public holiday. So I decided time to try getting the Adventure going, even though I don't have a case for it.

We start with Obi-Wan, the retired Jedi Knight I mean Acer Veriton V7900. Its an Intel Core2 Quad Q6600 machine with a single PCIe slot. Seeing as its retired I figured why not use it.


Next we have the old Corsair HX1000 power supply (1000 watts) which was also sitting around doing nothing. The bag is full of varous power cables as its a modular power supply.


And lastly we have the Adventure 2800 itself. The small board at the back is one of the bridge cards which the cable coming from the PC plugs into. 


After putting it together the 1st step was to see if it would power up without any add-in cards. This it did fine. There are a row of diagnostic LEDs on the front of the board that indicate it powered up.



The next step was to connect it to the PC. This bit also worked. However as soon as I added a graphics card to the Adventure the PC wouldn't boot. Probably something to do with the on-board Intel graphics and having an expansion board with a graphics card. Before anyone asks, I have run a couple of different graphics cards using the single PCIe expansion slot in it for a couple of years without any problems. Okay so it won't work with the Acer Veriton.

03 June 2012

3rd of June

Not much to report this week. Crunching continues on the farm...


Project news - Einstein
They've discovered 9 new pulsars. 7 are in the Parkes data and 2 in the Arecibo data.

They are also sending out framed certificates to the users who find these new pulsars. Unfortunately I don't seem to have come across one.


Project news - GPUgrid
As mention last week they are beta testing new apps. It seems CUDA 4.0 will be the minimum version and CUDA 4.2 for the Fermi and Kepler based cards. All of the latest batch of CUDA 4.0 ones failed for me and a number of other users.


Project news - Seti
They were off-line for a few days while they repaired the faulty power cable on campus at UC Berkeley.

They also have a badly thought-out feature where they have limited reporting of work to 64 work units at a time. They introduced this limit as users get an "HTTP Internal Server Error" at some point, believed to be above 1000 work units. This has lead to much frustration as many machines have well over 64 work units to report, mine included. It also created another bug where it tries to resend some of the tasks as "lost tasks" even though they were just reported.

26 May 2012

26th of May

Farm news
The farm has been chugging away all week with no dramas. During the week I tested the Nvidia 301.42 drivers on one machine. This weekend I took the opportunity to upgrade all the machines to it. I also upgraded the BOINC client to 7.0.28 so they are all the same.


Nvidia releases 301.42 drivers
Nvidia has released updated drivers. Nothing new in that, except this version addresses the "sleep bug" that the 295.x and 296.x drivers introduced for Windows. That means it should be safe to upgrade to this version so you can run CUDA 4.2 apps.


Project news - Einstein
I forgot to mention last week that Einstein have an OpenCL app for the ATI graphics cards. Its been in testing for a while on Albert@Home (the Einstein beta test project). It has now been released on the main project.

They also discovered 3 new pulsars in the data from Arecibo.


Project news - GPUgrid
They have been beta testing a CUDA 4.2 science app for a few weeks. They will be rolling it out publicly next week. The normal sized work units will still use CUDA 3.1 for the time being with the CUDA 4.2 app being used for the "long" work units.
Linux users will need 295.x drivers (or later) and Windows users will need 301.x drivers (or later) to run the new science app.

22 May 2012

22nd of May

Not much to report this week. The farm had the weekend off as I went down to the snowy mountains.


BOINCstats
There has been a number of changes to their web site. A few people have complained that they have to use multiple tabs to see things now, where as before they were on a single page. I haven't looked at my stats recently so can't comment on the changes.


Farm news
Apart from the weekend off its been crunching pretty flat-out on 3 of the machines.
The GTX570 in Sulu is starting to build up its RAC (recent average credit) as its been running constantly for about 2 weeks now. Its currently up to 90,000 which is pretty good for a single GPU.

The older EVGA GTX570 was sold on eBay, so its off to a new home. As I mentioned before I am looking at GTX670 cards to replace the remaining Palit GTX570's with. There is about 60 watts of power saving between the two generations of cards.


GPUUG fundraisers
The GPU Users Group have been doing a couple of fund raisers.
1. A hard disk transport case so they can ship drives between telescopes and the university.
2. Another hard disk transport case and 20 x 2Tb hard disk drives, again for transporting data.

I have donated a bit to both. Actually one was meant to finish off the 1st fund raiser, but I picked the 2nd by mistake. To fix it up I made another donation.


Project news - Seti
They spent most of last week off line due to a power failure. They are now back on-line and everything is up and running.

A 15kV underground power line heading from the campus power substation to the hill where the Space Sciences Building is located, shorted out around 2030 on Tuesday night. PG&E, the local power company, spent two days crawling the line from manhole to manhole looking for the problem. Apparently while looking for the first short, they managed to cause another short in the line.

13 May 2012

12th of May

Another SSD
I installed the Intel SSD in Spock on Monday. I copied all the stuff over from the hard disk, repartitioned it, formatted it and then copied a few things back. Adjust the defrag settings and run the SSD toolbox software and its done. The 3 main crunchers now have one each.


Heart transplant
This weekend I went off to the computer shop with Spock and the i7-970 CPU I mentioned on the 5th of May. It had its CPU transplant and a Corsair H80 cooler installed. Its been having a bit of a work out this afternoon when I got back home.

I have also had a look at the Corsair-Link solution to see if it might help with cooling. It allows monitoring various sensors in the machine and adjusting the fans and/or H80 cooler. Its both a hardware device that you plug fans and the H80 into and software for controlling them. They sell in the US for $100 each. I am not sure about Australian pricing.


GTX570 for sale
The GTX570 that was removed from Spock is now up on eBay. It turns out I had the original box with all the bits in it, so its packed up again and ready to go to the highest bidder.


GTX670's are out
Nvidia have now released the GTX670's based upon the Keplar chips. I have been having a bit of a look online as possible replacements for the 2 remaining (factory overclocked) GTX570 cards that I have. The Palit ones look quite good, but I need to see if there are any real world examples of people using them for crunching.

05 May 2012

5th of May

I bought this on eBay seeing as the computer shop couldn't get one. Its an i7-970 cpu. There is a headsink in the box too which is the main reason why the box is so big. I don't need the heatsink.



This one is destined to upgrade Spock from an i7-920 (8 cores @ 2.67Ghz). I have already upgraded its memory and ordered an SSD for it. They tried to deliver the SSD on Friday except nobody was home at the time. The computer shop don't have a Corsair H80 in stock, so I will have to wait another week before it can get done. This will make it the same setup as the other 12 core machine:

Intel i7-970 CPU (12 cores @ 3.2Ghz)
Corsair H80 water cooling
12Gb DDR3 memory
120Gb SSD
500Gb HDD
GTX560Ti graphics card


BOINC testing
We got 7.0.27 this week, which has a number of changes. Currently I have it running on my usual test machine without any issues.


CPDN project news
A couple of their upload servers crashed with hard disk errors last week. They managed to get one back up fairly quickly but the other took most of the week. It was back online and uploads going through but is now reporting a problem with one of the log files.

I have made a couple of donations to them so they might be able to replace some of the hard disks. I have asked how much they cost as I don't know how much a server grade 2Tb hard disk is in the UK, for that matter I don't know how many drives they would need.


GPU User Group news
They did a couple of quick fund raisers to supply a new RAID controller card with battery backup and 4 x 24" monitors for the project staff at Seti. Both have been delivered. The staff were quite impressed with having new large screen monitors. Matt (one of the project staff) posted a picture of his desk looking nice and neat using some books to stand the monitor on.

Quake Catcher project news
They also went off line for a few days. Now that they are back online it looks like they have updated their web site.

25 April 2012

ANZAC day 2012

Today is ANZAC day. the term ANZAC stands for Australian and New Zealand Army Corp. Its a day where we remember the sacrifices made by defence personel since the forming of the ANZAC force in World War I. Its like Remembrance day in the UK.


Other SSD installed
As I mentioned in my previous post I got a couple of Intel SSD's. One was installed in the i7-950 machine and the other I installed this week in the i7-970 machine. Installation was painless this time and it just worked out of the box.


CPU upgrade on the way
I have purchased (via eBay) another i7-970 chip. This will be used to replace an i7-920 in one of the other machines. Once I get it I will get a second Corsair H80 cooler installed by the shop at the same time.

Today I spent a bit of time getting it ready. I already have spare memory so the machine was updated to 12Gb of RAM. While I was at it I flashed the BIOS to the latest one. It had a GTX570 in it but I have swapped it out in favor of a GTX560Ti (which I also had spare). The GTX570 will be listed on eBay soon, it hasn't had much use so no point keeping it.

CPDN woes
They have had 2 of their upload servers fail with hard disk errors. I have a number of completed work units which cannot upload, along with most other users.

I have made a donation which I hope they will use to replace some hard disks. I would encourage anyone who also does CPDN work to make a donation if they can. The link to their donation page is hard to find (I had to ask where it was in one of their message boards). It is: http://climateprediction.net/content/donations

Coupled with this, or maybe as a result of, they have released 65,000 hapam3p_pnw (Pacific North West regional) work units.

21 April 2012

21st of April

This week I did a couple of upgrades. The most notable one was this thing. This is the Corsair H80 liquid cooling solution. The i7-970 was running rather warm and even with a 120mm CoolerMaster heatsink it wasn't coping very well.

In the kit you get a couple of Corsair 120mm fans, the radiator, tubing and cold plate. Its a closed-loop system so the radiator, tubing and cold plate are already connected and filled with coolant. All you need to do is install them and the fans.









This is looking at the back of the machine. We've used Noctua fans instead of the Corsair ones. They move more air, last longer and are quieter.








Here is another picture showing the radiator at 38mm thickness, plus the 2 fans at 25mm each.








And here is another photo looking towards the back of the case.








And lastly some stats from SIV showing it peaking at about 58 degrees C under load. Previously it was peaking around 70-72 degrees.





The other upgrade was replacing the OCZ SSD (which was only 64Gb) with an Intel 520 series one thats is 120Gb. Its in the i7-950 machine. After a bit of stuffing around with a Sata cable that wasn't quite plugged in and the DVD reader disappearing as a result I went to do a firmware update, but it turns out its already up to date. They give you a "data migration tool" which is a version of Acronis to copy your existing data across. I ran that and then removed the old SSD. There is also a "tuning kit" that disables a number of things such as the indexing service and Defrag. You have to download the software from the Intel web site.

I did some fiddling with the memory settings on the i7-970 during the week as well. With the help of Red-Ray on the Seti forums (author of SIV) I managed to get it running at 1600Mhz. Unfortunately it seems the 12 core Gulftown memory controllers are somewhat slower than the Bloomfield ones. The Gulftown i7-970 was getting around 10,000 MB/Sec at idle. The older Bloomfield i7-920's are able to get around 15,000 MB/Sec.

I have another Intel SSD to install in the i7-970 machine, but seeing as it got the H80 installed this weekend I figured it can wait a few days before I touch it again.

14 April 2012

14th of April

Milestones
I managed to get to 53 million for GPUgrid and both Einstein and Seti are coming up on 12 million. Below are my BOINCstats numbers.

 



12 cores finally working
On a whim I decided to fire up the 12 core machine and update it. That took a few hours installing windows updates, latest drivers and then BOINC. Its been powered off for the last 6 months. It still didn't work. It passes memory tests but as soon as you try and use it for number crunching it complains of memory errors. So I powered it off and went to bed. Around midnight I had an idea so started it up and went into the BIOS and set the memory speed to 1066 and let it download a single Einstein work unit. It started running so I let it go for a couple of minutes. Still going, so that looks like its working (they usually fail straight away). I allowed it to get a full 12 work units and went back to bed. In the morning they are all done and without error.

Comparing some run times it seems its slightly slower than the other machines. This is an i7-970 so its running at 3.2Ghz and the others I was comparing with are i7-920's which run at 2.67Ghz. The only difference is the memory speed. The i7-920's are all running their memory at 1600Mhz, but this one can only run at 1066Mhz. Anyway at least its going and I may fiddle with the settings later. I might just get an unlocked i7 (980x or 990x) for it so I can run the memory at its rated speed.


BOINC testing
7.0.25 became the latest "recommended" version earlier in the week. The main differences are:
  • Support virtual machines using VirtualBox
  • Support for OpenCL scheduling (ATI and Nvidia)
  • Support for 'distributed storage' projects (there are none at the moment)
The scheduler has been rewritten so that it will better share resources across projects if you run multiple projects.


GPU Users Group news
The 10 hot-swap hard drive fund raiser was finished off and the drives have been delivered to Seti at home. They also asked for a JBOD card so they could try and connect one of the existing servers to the JBOD. That too has been delivered.

We are waiting on news of another fund raiser, which should be targetting bandwidth improvements, but still need the approval of the project staff before this can proceed.