Feeds

back to article NEC tag teams with HP on high-end x86 servers

Japanese IT supplier NEC has expanded its technology alliance with HP to cover its future Project Odyssey x86 systems, and also to kick in a little help with the development and testing efforts. That's no surprise. Like HP, NEC still has a business peddling Itanium-based servers running the HP-UX variant of Unix, but it needs to …

COMMENTS

This topic is closed for new posts.
Anonymous Coward

"Obviously, with Linux being an open source platform, it is easier to create add-on high availability software for it as well as make contributions to the Linux kernel that make it more ruggedized."

As opposed to not having to do anything with Windows Server which already has standardised native high availability clustering, and supports hot CPU and memory add, NUMA control, etc, etc...

2
3
Silver badge

"The companies are aligning their engineering, technical, and testing teams to harden Linux and Windows as part of the re-upped deal."

You seem to have missed this bit !

1
0
Anonymous Coward

"NUMA control, etc, etc..."

Suggest you Google ( or Bing in your case ) SGI UV servers

This latest SGI UV product family can scale a single system image (SSI) to a maximum of 2,048 cores (4,096 threads) because of its innovative NUMAlink® interconnect.. SGI UV supports the latest Intel® Xeon® processor E5-4600 product family, and can operate unmodified versions of Linux - SUSE® Linux Enterprise Server and Red Hat® Enterprise Linux®. SGI UV is the only solution that currently leverages the power of Intel's latest CPU beyond 4 sockets and 1.5TB memory per system.

Single System Image

1
1
Silver badge
Facepalm

Re: AC

".....SGI UV is the only solution that currently leverages the power of Intel's latest CPU beyond 4 sockets and 1.5TB memory per system....." Erm... No. Hp Proliant DL900 series are eight socket, using Intel E7 CPUs with up to ten cores each, and lots more memory than 1.5TB. The old DL980G7s we have scale to 4TB of RAM if you can afford the really large DIMMs.

1
0
Bronze badge

Re: AC

Dammit Matt -- ya beat me to it.

And HP has brought BACK MCSG for linux, so my support on the 12 of those I have in 3 clusters will be extended too.

Ripping hardware workhorses those.

1
0

@AC "innovative Numalink technology", do you work for SGI by any chance?

1
0

Re: AC

They probably mean Xeon E5, as Xeon E7 is by no means "latest" since it is waiting to be upgraded to Ivy Bridge EX in Q1/2014. Currently it is based on now ancient Westmere microarchitecture.

Xeon E5 is based on Sandy Bridge uarch, and the upgrade to Ivy Bridge is imminent (in couple of weeks) - however, E5 is limited to 4 sockets unless you use 3rd-party connectivity solution such as NUMAlink.

0
0

SGI UV is a cluster.

You got it wrong. Sgi is using a hypervisor to trick the linux kernel into believing it is running a single smp server, when it is in fact a cluster. ScaleMP also use the same approach, they also have a 2048 cpu linux server, but they are also running a cluster. These clusters have awful latency to nodes far away, they may look like single image, but it is not. They are only used for parallell HPC workloads. Google on ScaleMP too, and see it also uses a software hypervisor tricking linux kernel into running a single server. These servers are clusters, and unusable as smp single servers.

0
1

You got it wrong. Sgi is using a hypervisor to trick the linux kernel into believing it is running a single smp server, when it is in fact a cluster. ScaleMP also use the same approach, they also have a 2048 cpu linux server, but they are also running a cluster. These clusters have awful latency to nodes far away, they may look like single image, but it is not. They are only used for parallell HPC workloads. Google on ScaleMP too, and see it also uses a software hypervisor tricking linux kernel into running a single server. These servers are clusters, and unusable as smp single servers.

UPDATE: here is a link on the similar clustered server ScaleMP sells (2048 cores, 64TB RAM):

http://www.theregister.co.uk/2011/09/20/scalemp_supports_amd_opterons/

"....Since its founding in 2003, ScaleMP has tried a different approach. Instead of using special ASICs and interconnection protocols to lash together multiple server modes together into a shared memory system, ScaleMP cooked up a special hypervisor layer, called vSMP, that rides atop the x64 processors, memory controllers, and I/O controllers in multiple server nodes. Rather than carve up a single system image into multiple virtual machines, vSMP takes multiple physical servers and – using InfiniBand as a backplane interconnect – makes them look like a giant virtual SMP server with a shared memory space...."

0
1

NEC processors

Will they have a Z80 compatible mode still?

0
0
Silver badge

hmm

>Kittson will be socket-compatible with the existing Itaniums and be made with the same 32nm processes used to make the Poulson Itaniums.

>The "Kittson" Itanium processors, which everyone expected to come out in 2014 or so, are almost certainly the last of the Itanium chip line even though HP and Intel will not admit it.

Wow RIP Itanium. We hardly knew ya.

0
1
Silver badge

Re: hmm

Still this link is always worth a good laugh even if they quit updating it to show the even sadder true trajectory. Some epic fails end with a whimper instead of a bang.

https://en.wikipedia.org/wiki/File:Itanium_Sales_Forecasts_edit.png

0
1
Anonymous Coward

If they (Intel and HP) were smart Kittson would never see the light of day. If Kittson is the last, I expect to see a lawsuit from Oracle against Intel and HP claiming they perjured themselves in court and then Oracle will go on a fishing expedition looking for the proof that is almost certainly there. HP and Intel both know that Kittson will be last. At most they might juice the chips a bit after several years once the sales have declined even further.

0
1
Bronze badge

Hmm

Half of the comments blocked by Adblock. Weird or what.

0
0
This topic is closed for new posts.