Most server boards i've looked at have onboard video. However this is not shared memory video in the northbridge like you see on desktop boards, it is connected to the southbridge via either 32-bit PCI or PCIe x1 and usually tied in with the remote management stuff. I don't know for sure how well it performs but my suspicion is worse than intel integrated graphics.
The block diagram in TFA shows 4 QPI interfaces, so theoretically yes. Um no, 4 interfaces would only allow five devices to be directly connected.
I'm betting they use no more than 3 QPI links for CPU-CPU interconnect per core, making a 2-hop system (which is still a challenge). The intel diagram I saw had 3 QPI links on each processor for CPU-CPU (in what was afaict a 2 hop system). Then there were four IO hubs each connected to two processors using the remaining QPI links.
They were leaning on that "oh, our bad, we got some 'demo units' by mistake" excuse pretty hard. My guess is they were just passing on what thier supplier had told them until they could prove otherwise.
It's kind of a difficult position to be in. Newegg probablly didn't want to lie but they probablly didn't want to be publically seen as doubting thier suppliers either until they could prove that the supplier was indeed shipping fakes.
Xeon is brand and nahalem-ex a codename. These chips will be branded as xeon.
The xeon brand has been used for many architectures over the years and I don't see any reason it wouldn't continue to do so just as the names pentium and celeron have been recycled.
What this probably will mean though is that the writing is on the wall for core 2 based xeons now there is a nehalem based xeon that supports more than 2 sockets and more than four cores per socket. I'm sure they will stick arround for a while for legacy support reasons (just as you can still buy core 2 chips for the desktop) but eventually they will be phased out.
I'm(perhaps optimistically) assuming that that means all the RAM in an up to 8 socket system wouldn't be more than one hop away from any core. These only have 4 QPI links so in a system with more than 4 sockets (remember you have to hook up the IO hubs somewhere) 8-socket system some processors will be two hops away from each other.
Does it have the memory I/O bandwidth to keep up with the CPUs? IIRC it has quad channel ram so i'd expect the answer to this to be similar to running a quad-core on dual channel ram.
At least according to toms hardware going from dual to triple channel on a quad core doesn't make much difference. so it seems one channel per two cores is about right with current tech.
When will I be able to actually buy a mother board with 8 of these 8 core CPUs Dunno, I expect you will see the first boards and servers built arround this CPU arround release time (i'm sure the server and motherboard vendors already have engineering samples). I dunno if any of them will be 8-socket though.
and what kind of a frame rate would Crysis get on that rig? Probablly not that much better than what it gets on current hardware.
I doubt there will be many single socket systems based on this CPU. Remember this isn't on the same socket as any existing CPU and it really isn't designed for such low end systems.
The real news is that this means intel will have a current gen product for machines with more than 8 cores and/or more than 6 ram channels (6 ram channels with 2 sockets per channel* means 48GB with reasonablly priced modules or 96GB with very expensive 8GB modules,).
* in theory you can put three modules on one channel but it limits you to DDR3-800 speeds and none of the boards i've looked at supported it.
For better or worse the linux kernel doesn't have a properly designed stable ABI (linus says he does this deliberately to avoid ever getting locked into a driver API/ABI because too much becomes dependant on it) so it's very hard to fix the kernel without introducing ABI changes that require third party modules to be recompiled (debian do try but don't always succeed, not sure about ubuntu).
The problem it it's only very recently that a system for managing this automatically has been introduced and there is no standardisation of such a system accross distros (i'm not even sure if distros outside the debian family have a system at all)
So if you get the drivers directly from NVIDIA they won't operate within this system and will have to be manually fixed each time.
It sucks but there isn't a whole lot the distros can do to change the behaviour of stuff that is installed ignoring the package manager and module management system they use.
Sounds like a lot of hassle to me, especially as different plugs require the wires cut to different lengths to wire them properly and safely.
I also dislike travel adaptors, they usually seem poorly made and often try to fit multiple socket types and fit none of them well.
Personally I think the best option would be to put an IEC on your extention then have short IEC leads for the different plug types. Afaict i's usually pretty easy to scrounge up a local IEC lead from somewhere if you've missed a plug type from your set.
Actually I don't think Mozilla actually requested debian to change the name, debian changed it themselves because firefox was trademarked and debian want to be free of intellectual property issue. Then you didn't follow the issue.
Debian seem to think trademarks are ok but copyrighted logos under a non-free license aren't. I can understand where this position comes from though i'm not sure I agree with it.
Debian claim they were given permission by mozilla a long time ago to use the firefox name on thier unofficial builds of firefox despite the fact they didn't use the branded logos (the firefox source has always included a set of "unbranded" logos, you see them on nightlies and alpha releases among other places).
Later someone from mozilla came along (seems they got new people in charge of such things) and basically said that if debian wanted to use the firefox name they had to both accept the non-free logo and give mozilla veto power over any patches debian applied. Neither of these conditions was acceptable to debian.
They could have done what the nightlies and alphas do and used the release codename as the name of the browser but a browser whose name changed with every release would be pretty confusing for users. So they switched to using thier own name. Soon afterwards they also switched from using the mozilla supplied unbranded logos to using thier own logos (though I don't think this was caused by a licensing issue but just because they looked better than the unbranded mozilla logo and also fit with the new name).
If linus had written all of linux that would be true but he didn't not by a long way.
So unless linus had a policy of getting either permission to relicense the code or a copyright assignment from each and every contributor (afaict some opensource projects do this but linux isn't one of them) then he can't relicense it without a heck of a lot of work to
1: work out who wrote each line of code 2: contact all those authors to get permission 3: rewrite any code which the authors refuse to give permission to relicense.
Apache on linux (at least in all the setups i've seen) starts as root so it can bind port 80 but then switches down to a lower privilage user to do the actual serving. Some damage could still be done of course but hopefully it's limited compared to the damage root can do.
Apache on windows defaults to running as "localsystem" (roughly the windows equivilent of root)
You can run it as another user but apparently ( http://httpd.apache.org/docs/2.0/platform/windows.html ) that user has to have "Act as part of the operating system" privilages. MS describes said privilages as "This user right allows a process to impersonate any user without authentication. The process can therefore gain access to the same local resources as that user.".
So it seems either way to run Apache on windows you have to give it what ammounts to root privilages.
All Sky1 shows usually was re-re-re-runs of Stargate anyway. I was under the impression that sky1 showed a lot of big american shows before anyone else in the UK (or at least they used to).
Yeah they fill the rest of the time with repeats but frankly most channels do that.
Unfortunately I think the likely long term outcome of this is that even more gamers move to consoles where it is easier to control piracy without resorting to such draconian measures.
Pirating on consoles requires getting your console physically modified which costs money, voids the warranty and carries a high risk of getting your console permanently banned from online services.
Yes you will need a case tall enough to take cards without risers (which means 3U afaict) but I would guess getting the same IOPS any other way would take up way more than 3U of rackspace.
WOW NICE motherboard there, TWO io hubs to give seven of x8 electrical/x16 mechanical slots along with a x8 link for the onboard SAS and an x4 link for the onboard dual port gigabit.
One problem I see is that afaict trim can't just be passed through. The controller needs to take careful steps to handle trim in a way that keeps the raid data consistent.
My (possibly incorrect?) understanding was that 3-4x 7200rpm drives arent a drop-in replacement for a 15k in all situations Not all but I would think most.
If you are comparing a big array of 15K drives with an even bigger array of 7.2K drives I would think it likely that the application in question is one that is capable of generating large numbers of requests in parallel (most likely some kind of database server).
Frankly, I think all parties benefit with a software solution. IMO the best solution is a hybrid. Let software be in charge of finding the normal balance between cool, quiet and fast but have safety limits in the hardware that kick in when thing go too far outside normal.
BTW does anyone know if anyone has retested the affect of heatsink removal under load with more modern CPUs since toms hardware did it back in 2001. I'd like to know if AMD have got thier act together...
The only thing that could have been done, is adding a safe guard which fires a software alarm and either shuts down or massively underclocks the 3D core in case a temperature threshold is crossed. (That's how it's done to protect CPU in case of faulty fan). Back in the P4/athon XP era toms hardware demonstrated ( http://www.youtube.com/watch?v=BSGcnRanYMM ) removing the heatsink from various chips while gaming and the P4 kept going albiet at slideshow framerates and reconvered fine when the heatsink was reattatched . Unfortunately i'm now aware of any similar tests since
Nvidia need to realise (like AMD did after that video) that thier overheat protection systems need failsafes implemented at as low a level as possible so that even if the fan system fails the chip can't cook itself.
One thing to remember about PCs is that there is a wide range of voltages present. At the high end you have 12V for analog stuff and bulk power distribution* while at the low end you have signal voltages of less than 2V (afaict) and core voltages of under a volt.
I could easily see a A badly fried graphics card putting 12V onto the data lines. That would almost certainly damage whatever was on the other end (northbridge for LGA775 stuff, IOH for lga1366 stuff, CPU for LG1156 stuff) and maybe stuff beyond that too.
*There has been a shift in PC architecture design in recent years from distributing power at 5V and 3.3V to distributing it at 12V. 12V has lower cable losses and the power has to be routed through switchers near the load anyway to power modern high current low voltage devices like CPUs.
It's an option, the thing is most people doing this are in it to make money. That means they have to have some kind of financial transactions with the rich world whether to pay for access directly or (more likely) to pay for advertisements.
Just like with online gambling I would expect the pressure to be applied to the companies who help the profits flow to the offshore companies performing actions that are legal (or at least impractical to prosecute) in the offshore companies jurisdiction but illegal in the customers jurisdiction.
Remember the big advert networks are what enable these sites to make money by placing customers adverts on sites that the customer would be unlikely to associate with by choice.
Slashdot Mirror
Most server boards i've looked at have onboard video. However this is not shared memory video in the northbridge like you see on desktop boards, it is connected to the southbridge via either 32-bit PCI or PCIe x1 and usually tied in with the remote management stuff. I don't know for sure how well it performs but my suspicion is worse than intel integrated graphics.
The block diagram in TFA shows 4 QPI interfaces, so theoretically yes.
Um no, 4 interfaces would only allow five devices to be directly connected.
I'm betting they use no more than 3 QPI links for CPU-CPU interconnect per core, making a 2-hop system (which is still a challenge).
The intel diagram I saw had 3 QPI links on each processor for CPU-CPU (in what was afaict a 2 hop system). Then there were four IO hubs each connected to two processors using the remaining QPI links.
They were leaning on that "oh, our bad, we got some 'demo units' by mistake" excuse pretty hard.
My guess is they were just passing on what thier supplier had told them until they could prove otherwise.
It's kind of a difficult position to be in. Newegg probablly didn't want to lie but they probablly didn't want to be publically seen as doubting thier suppliers either until they could prove that the supplier was indeed shipping fakes.
Xeon is brand and nahalem-ex a codename. These chips will be branded as xeon.
The xeon brand has been used for many architectures over the years and I don't see any reason it wouldn't continue to do so just as the names pentium and celeron have been recycled.
What this probably will mean though is that the writing is on the wall for core 2 based xeons now there is a nehalem based xeon that supports more than 2 sockets and more than four cores per socket. I'm sure they will stick arround for a while for legacy support reasons (just as you can still buy core 2 chips for the desktop) but eventually they will be phased out.
I'm(perhaps optimistically) assuming that that means all the RAM in an up to 8 socket system wouldn't be more than one hop away from any core.
These only have 4 QPI links so in a system with more than 4 sockets (remember you have to hook up the IO hubs somewhere) 8-socket system some processors will be two hops away from each other.
there is a diagram of an example 8-core setup at http://www.hardwarecanucks.com/wp-content/uploads/intel_nehalem-ex-8-core.jpg
Does it have the memory I/O bandwidth to keep up with the CPUs?
IIRC it has quad channel ram so i'd expect the answer to this to be similar to running a quad-core on dual channel ram.
At least according to toms hardware going from dual to triple channel on a quad core doesn't make much difference. so it seems one channel per two cores is about right with current tech.
When will I be able to actually buy a mother board with 8 of these 8 core CPUs
Dunno, I expect you will see the first boards and servers built arround this CPU arround release time (i'm sure the server and motherboard vendors already have engineering samples). I dunno if any of them will be 8-socket though.
and what kind of a frame rate would Crysis get on that rig?
Probablly not that much better than what it gets on current hardware.
I doubt there will be many single socket systems based on this CPU. Remember this isn't on the same socket as any existing CPU and it really isn't designed for such low end systems.
The real news is that this means intel will have a current gen product for machines with more than 8 cores and/or more than 6 ram channels (6 ram channels with 2 sockets per channel* means 48GB with reasonablly priced modules or 96GB with very expensive 8GB modules,).
* in theory you can put three modules on one channel but it limits you to DDR3-800 speeds and none of the boards i've looked at supported it.
For better or worse the linux kernel doesn't have a properly designed stable ABI (linus says he does this deliberately to avoid ever getting locked into a driver API/ABI because too much becomes dependant on it) so it's very hard to fix the kernel without introducing ABI changes that require third party modules to be recompiled (debian do try but don't always succeed, not sure about ubuntu).
The problem it it's only very recently that a system for managing this automatically has been introduced and there is no standardisation of such a system accross distros (i'm not even sure if distros outside the debian family have a system at all)
So if you get the drivers directly from NVIDIA they won't operate within this system and will have to be manually fixed each time.
It sucks but there isn't a whole lot the distros can do to change the behaviour of stuff that is installed ignoring the package manager and module management system they use.
Sounds like a lot of hassle to me, especially as different plugs require the wires cut to different lengths to wire them properly and safely.
I also dislike travel adaptors, they usually seem poorly made and often try to fit multiple socket types and fit none of them well.
Personally I think the best option would be to put an IEC on your extention then have short IEC leads for the different plug types. Afaict i's usually pretty easy to scrounge up a local IEC lead from somewhere if you've missed a plug type from your set.
If you play high-definition video on YouTube, you exercise a CPU.
Especially if you try and do it under linux.
Actually I don't think Mozilla actually requested debian to change the name, debian changed it themselves because firefox was trademarked and debian want to be free of intellectual property issue.
Then you didn't follow the issue.
Debian seem to think trademarks are ok but copyrighted logos under a non-free license aren't. I can understand where this position comes from though i'm not sure I agree with it.
Debian claim they were given permission by mozilla a long time ago to use the firefox name on thier unofficial builds of firefox despite the fact they didn't use the branded logos (the firefox source has always included a set of "unbranded" logos, you see them on nightlies and alpha releases among other places).
Later someone from mozilla came along (seems they got new people in charge of such things) and basically said that if debian wanted to use the firefox name they had to both accept the non-free logo and give mozilla veto power over any patches debian applied. Neither of these conditions was acceptable to debian.
They could have done what the nightlies and alphas do and used the release codename as the name of the browser but a browser whose name changed with every release would be pretty confusing for users. So they switched to using thier own name. Soon afterwards they also switched from using the mozilla supplied unbranded logos to using thier own logos (though I don't think this was caused by a licensing issue but just because they looked better than the unbranded mozilla logo and also fit with the new name).
If linus had written all of linux that would be true but he didn't not by a long way.
So unless linus had a policy of getting either permission to relicense the code or a copyright assignment from each and every contributor (afaict some opensource projects do this but linux isn't one of them) then he can't relicense it without a heck of a lot of work to
1: work out who wrote each line of code
2: contact all those authors to get permission
3: rewrite any code which the authors refuse to give permission to relicense.
Apache on linux (at least in all the setups i've seen) starts as root so it can bind port 80 but then switches down to a lower privilage user to do the actual serving. Some damage could still be done of course but hopefully it's limited compared to the damage root can do.
Apache on windows defaults to running as "localsystem" (roughly the windows equivilent of root)
You can run it as another user but apparently ( http://httpd.apache.org/docs/2.0/platform/windows.html ) that user has to have "Act as part of the operating system" privilages. MS describes said privilages as "This user right allows a process to impersonate any user without authentication. The process can therefore gain access to the same local resources as that user.".
So it seems either way to run Apache on windows you have to give it what ammounts to root privilages.
All Sky1 shows usually was re-re-re-runs of Stargate anyway.
I was under the impression that sky1 showed a lot of big american shows before anyone else in the UK (or at least they used to).
Yeah they fill the rest of the time with repeats but frankly most channels do that.
Unfortunately I think the likely long term outcome of this is that even more gamers move to consoles where it is easier to control piracy without resorting to such draconian measures.
Pirating on consoles requires getting your console physically modified which costs money, voids the warranty and carries a high risk of getting your console permanently banned from online services.
There are server boards out there with pelnty of PCIe slots e.g. http://www.supermicro.com/products/motherboard/QPI/5500/X8DTH-i.cfm (shamelessly grabbed from a board in a picture linked from another post here)
Yes you will need a case tall enough to take cards without risers (which means 3U afaict) but I would guess getting the same IOPS any other way would take up way more than 3U of rackspace.
WOW NICE motherboard there, TWO io hubs to give seven of x8 electrical/x16 mechanical slots along with a x8 link for the onboard SAS and an x4 link for the onboard dual port gigabit.
http://www.supermicro.com/products/motherboard/QPI/5500/X8DTH-i.cfm
One problem I see is that afaict trim can't just be passed through. The controller needs to take careful steps to handle trim in a way that keeps the raid data consistent.
My (possibly incorrect?) understanding was that 3-4x 7200rpm drives arent a drop-in replacement for a 15k in all situations
Not all but I would think most.
If you are comparing a big array of 15K drives with an even bigger array of 7.2K drives I would think it likely that the application in question is one that is capable of generating large numbers of requests in parallel (most likely some kind of database server).
Frankly, I think all parties benefit with a software solution.
IMO the best solution is a hybrid. Let software be in charge of finding the normal balance between cool, quiet and fast but have safety limits in the hardware that kick in when thing go too far outside normal.
BTW does anyone know if anyone has retested the affect of heatsink removal under load with more modern CPUs since toms hardware did it back in 2001. I'd like to know if AMD have got thier act together...
The only thing that could have been done, is adding a safe guard which fires a software alarm and either shuts down or massively underclocks the 3D core in case a temperature threshold is crossed. (That's how it's done to protect CPU in case of faulty fan).
Back in the P4/athon XP era toms hardware demonstrated ( http://www.youtube.com/watch?v=BSGcnRanYMM ) removing the heatsink from various chips while gaming and the P4 kept going albiet at slideshow framerates and reconvered fine when the heatsink was reattatched . Unfortunately i'm now aware of any similar tests since
Nvidia need to realise (like AMD did after that video) that thier overheat protection systems need failsafes implemented at as low a level as possible so that even if the fan system fails the chip can't cook itself.
One thing to remember about PCs is that there is a wide range of voltages present. At the high end you have 12V for analog stuff and bulk power distribution* while at the low end you have signal voltages of less than 2V (afaict) and core voltages of under a volt.
I could easily see a A badly fried graphics card putting 12V onto the data lines. That would almost certainly damage whatever was on the other end (northbridge for LGA775 stuff, IOH for lga1366 stuff, CPU for LG1156 stuff) and maybe stuff beyond that too.
*There has been a shift in PC architecture design in recent years from distributing power at 5V and 3.3V to distributing it at 12V. 12V has lower cable losses and the power has to be routed through switchers near the load anyway to power modern high current low voltage devices like CPUs.
IMO if you are going to count episodes then you should also count expansion packs
half life: late 1998
opposing force: late 1999 (1 year)
blue shift: mid 2001 (1.5 years)
decay: late 2001 (0.5 years)
hl2: late 2004 (3 years)
ep1: mid 2006 (1.5 years)
ep2: late 2007 (1.5 years)
It's an option, the thing is most people doing this are in it to make money. That means they have to have some kind of financial transactions with the rich world whether to pay for access directly or (more likely) to pay for advertisements.
Just like with online gambling I would expect the pressure to be applied to the companies who help the profits flow to the offshore companies performing actions that are legal (or at least impractical to prosecute) in the offshore companies jurisdiction but illegal in the customers jurisdiction.
Remember the big advert networks are what enable these sites to make money by placing customers adverts on sites that the customer would be unlikely to associate with by choice.
What if they only allow known crawlers from major search engines?