* Posts by scorched_cpu

11 posts • joined 25 May 2012

Heaps of Windows 10 internal builds, private source code leak online

scorched_cpu

Tron replay?

Sounds like the Tron Legacy movie. This year we put a 10 on the box! And oh yeah by the way flynn OS I mean EncomOS is available for download on the web. And yes I know, comparing win10 to flynnOS is a horrible insult.

Reg man howls over HPE Moonshot IoT box

scorched_cpu

Does it help that it meets NEBS level 3, and will operate at up to 55c and accept -48VDC? conditions. Come on that has to mean something for datacenter type of processing at the edge. The original design concept was for it to be installed in retail chain stores, to sit in a dirty closet or on some managers desk.

What I can say is that I just left HPE ES, but my work there led to many a raised eyebrow in the DoD for use in tactical situations.The HPE Federal team has a sub contractor ready to offer it in Mil-Spec 810G and others for use in extreme harsh environments. I concepted having 3 or 4 of these in a transit case giving me 16 cartridges in 4U. That provided me upto 256 cores of compute. Add that the m510 holds 3 and m710x holds 4 of the newer format M.2 storage. So on the m510 2 of M2. could be 1TB meaning 2TB per cartridge, the intent was to use some form of san virtualization to create a hypervisor cluster with upto 32TB of storage. Might not sound like much, but in a truly portable/tactical situation it was. And the variant idea was to use 12 m510's to provide 192 cores of hypervisor cluster, and 4 m710x with the GPU to provide a tactical capable of providing field ready VDI. Overarching goal was to provide a software defined datacenter in a box.

I was personally asking the Edgeline team to enable support the m800, a quad TI based 66AK2H processor. So in this config it offer 4 cores of ARM and 8 cores of C66x DSP per SoC, giving 16 cores of ARM and 32 Cores of DSP per cartridge. My goal here was due to the PXI options for running the already supported National Instruments FlexRIO cards, it would be able to support the concept of the above mentioned SDDC with Software Defined Radio. With the HPE capability to run NSA suiteB functions to provide field crypto without the need for an over priced mil-spec'd harris setup. Again many in the DoD and intel agencies did like the idea.

Not sure if that helps anyone see the vision for future use of the EL4000. But it was doing well for my job.

scorched_cpu

Re: 4 computers in one 1U chassis?

Did you even read from the link? The chassis is only 17"x23". What is not stated is that each cartridge is only approx 6.5"x7.24" and only 0.78" thick. The diagram is 100% incorrect, all 4 cartridges are up front stacked 2 per side. The rear is 2 PSU's and the choice of PCIe or PXI frame.

HPE sharpens knife for next salami-slicing staff redundo round

scorched_cpu

Re: Pathetic company

I agree. I am in the role of a distinguished technologist at this company, have been here just under 2 years and can't wait to get out. Management is atrocious and for any fellow HPE'ers out there, in case you have not heard, the company just canceled your college tuition assistance benefits to save money for the new-co merger.

HP slaps dress code on R&D geeks: Bin that T-shirt, put on this tie

scorched_cpu

Internal Video

The new head of HPE released a funny internal video countering all the rumors of a dress code. HPE does not have an official policy, and he goes on to say we are all adults and to dress appropriately.

EMC, HP blockbuster Borg: Big potential or total CATASTROPHE?

scorched_cpu

HP SL4540

The author makes a comment of the HP 4500 as being ideal for EMC SDDC software.

Funny he should say that, because the HP SL4540 is the ONLY platform supported on the newest VIPR 2.0 for commodity servers.

http://www.emc.com/techpubs/vipr/new_feature_in_vipr_2-1.htm

Crazy Texans dunk servers in DEEP FRYERS

scorched_cpu

Re: Not new and not green

I totally agree, and I didn't fully run all the calculations and understand that fluid is more efficient at removing heat.

But the thing I was trying to imply, is that the grcooling isn't really all that efficient. It uses a 3.5KW pumping system that requires a water inlet of upto 120gpm to exchange heat. So the cost and efficiency of this makes no longer makes sense as the infrastructure to get water in a datacenter and power to run the pumps negates alot of the benefit.

scorched_cpu

Re: one word:

these liquids are non-conductive and while not entirely non-flammable, the flashpoint for off the shelf mineral oil is at typically 120-140C, and commercial versions are up to 200C.

scorched_cpu
Thumb Down

Not new and not green

These ideas are old just with twists, just look at toms hw back in 2006!

http://www.tomshardware.com/reviews/strip-fans,1203.html

Plus, mineral oil comes from petroleum distillates while making gasoline and while the overall cost to acquire it is low, it still requires crude oil to make and hence not green at all. Then in one video they claim it saves 20% power by replacing fans, this may be true for calculating the entire datacenter, but i just don't see it for small environments, a server with a high end cpu fan only consumes 7-8W max, which is typically less than 10% of the overall wattage a server consumes. Add in mineral oil cost of manufacturing, cost of cooling infrastructure and your breaking even at best. I have seen datacenter shipping container manufacturers with this type of cooling, so again, there is nothing new about what they are doing.

NetApp streaks ahead, avoids trip-up by bum flash leg

scorched_cpu

Chris Mellor has forgotten

Funny but the author has forgotten his own article about netapp's off array flash and integrated links to ONTAP, way better than EMC will have.

http://www.theregister.co.uk/2011/06/17/netapp_project_mercury/

So while NetApp has yet to acquire a company similar to xtremeIO, you can rest assured they are looking at lots of technology. And the OEM E-series deals, while limited had one of the biggest wins, the 55PB lustre FS for the Sequoia Supercomputer at LLNL.

scorched_cpu

Re: Why do you keep bringing up flash drives outside of the array?

PAM is inside the array as part of the read cache mechanism.

Biting the hand that feeds IT © 1998–2019