Re: Future ARM laptops
The latency will kill you when moving around a 3D wire mesh, the more complex the object the more the latency will kill you. I doubt the average browser will be able to cache the size of a typical CAD file
231 posts • joined 19 Feb 2008
Cisco do a C880 (which is actually a OEM Fujitsu), HPE (who now own SGI) and also Atos Bull all do 8 Socket boxes as well as IBM Power
But if you need any sort of high availability you will always need at least two.
If your workload can be split up much better going for smaller nodes with one spare.
These days with such high core counts it's generally the memory costs that determine the server cost.
Do you run a single server with high core counts and high memory requiring 64GB or 128GB Dimms or lost of small ARM servers with 16/32GB Dimms. All of a sudden real estate becomes an issue.
Having a common kernel is a good idea, what Apple should do is think about different personae depending on form factor which will have to be adhered to my software vendors, I.e. a cut down interface with different functionality in mobile mode and a full desktop experience when linked to a monitor and keyboard/mouse.
I actually quite like some of the touch features in windows 10 in desktop mode
Now Hyperconverged is something entirely different.
But converged/Hyperconverged also means, get all your support and lifecycle management from one place. Rather than constant finger pointing and time wasted working out which patches/firmware updates work together across different vendors.
Think about all of the non paid for support in most departments.
e.g. Barry in Accounts has a small issue with windows (does not know how to do something), So shouts out, "Does anybody know how to do XYZ".
Shiela in also in accounts says yes and helps Barry.
Now most people in an office know windows/office to some degree as they have used it either at home or in another job.
Now if you have a linux desktop with Libre Office, the probability of somebody knowing how to do something goes down massivily.
Even worse is something does not work in Linux, you do a quick google and the answer is to run some horrid (to the end user) bash command to fix something.
When I was there it was £20 for an evening meal. (you made sure you booked a hotel with breakfast included) and that was it.
So if you ended up working late which was not unusual and got back to the hotel at say 9/10pm the last thing you wanted to do was go in search of a cheap place to eat.
Even the cheapest burger in the hotel was often £15. Add in a couple of drinks (even non alcoholic) and a side and your £20 budget is blown
The last benchmark Oracle released for SAP was 2016, An 8 Socket Machine with 256 Cores.
Giving a SAPs per core performance of 2787. A total SAPs score of 713480.
In the same year a Dell Poweredge R730 had a SAPs per core of 2677. Yes this was only a 2 Socket box with a SAPs rating of 117780.
The same server with Skylake has a per core SAPs rating of 3129.
So yes SPARC won on overall numbers. However I am yet to meet the customer who has a DB server requirement of 713480 SAPS.
Even a requirement of 117780 SAPs for a DB server is very rare. Yes the whole landscape might require lots of SAPs but a bunch of 2 Socket x86 servers will a) be cheaper, b) won't have all your eggs in one basket.
Yes there are workloads that have been tuned for SPARC, likewise for intel. At the end of the day it comes down to cost. x86 is a good enough workhorse at a far cheaper price.
Ex CSC myself
I find it very sad what has become of the one big outsourcing companies.
Companies with hundreds of technical/functional consultants and analysts and
Programmers. They could have come up with innovative SaaS solutions new ways
of doing business, but they forgot their DNA. Instead the good left sometimes ending
up in start ups.
Alas the accountants and restructuring managers are in charge.
Very odd. I live in London and already have fibre to the home, currently the max offer is 300Mb but surely just needs a switch upgrade at the exchange.
Alas my internal home network is nowhere near that speed so anything over 200Mb would be wasted (damn victoran terrace houses with no hollow walls, so have to rely on powerline and wifi)
With Oracle purchasing Sun, they effectivly killed off any software/hardware partnerships that where out there.
If Sun where still independent would they have maintained a good partnership with SAP such that a HANA port would have been on the cards (like it is with IBM power).
Would they have had a greater relationship with developers who would have embraced SPARC/Solaris as a development platform.
The second Oracle came on board the software providers went, urgh not a friendly competitor/partner, I'll stick to x86 with Power as our only RISC based system.
Rather than design a large aircraft to carry the payload up to a certain height. Why not design a smaller aircraft that attaches to the payload (with it's own fuel), that uses the engines on the payload.
The Payload then drops the aircraft bit, which flys itself home.
No need for large fusalages, automate the whole thing with no need for people on board.
Even if the hardware is locally maintained, it's the software that goes tits up when the hardware fails. And it's the software management that has been oursourced.
The software is old and does not have automated resyncing (taking referencial integrety into account),
there are likely to be hundreds of little scripts that have been created to get around functionality gaps, many of which will not have been documented or architected properly (likely to have been created to fix a P1 issue).
I've been part of many an outsourcing deal, with perspecitive from both sides.
The recipient doesn't know what they don't know, the people giving up the system don't know what they have forgotten and are generally pissed off they may be redundent soon.
The best moves have been the ones where a ground up review/re-write/rearchitect/retest have been part of the move adding new functionality, plugging gaps properly and properly re-testing.
Well they already have an enterprise class cloud with virtustream.
Also why build a cloud when you can hook into existing clouds, without the need to build them.
AWS can now take on VMware workloads.
DellEMC storage can use existing cloud storage providers as an additional storage tier within
and array. You can even backup to existing cloud storage providers out of the box.
My personal opinion. Why go out to the cloud, why not invite the cloud into your own data center.
I.e. offer you spare capacity out to the cloud providers who can then distribute/create HA nodes
on a smaller granular level, thus taking the risk away from large sites.
Why the hell would you not run DBs on Virtual machines?
30% overhead? Not seen those kinds of overheads for years.
90% of DBs I see out in the marketplace, do not require anywhere near the amount of
compute capacity available on modern platforms with 18+Cores per Socket in a two Socket server.
If you have the kinds of requirements for 8+ Sockets for a DB then you probably want to run the DB on Prem anyway or at least in a private/sourcing deal as no cloud can provide the kind on SLAs needed with such a DB which is likely to be business critical. (Or the company is crap at archiving data)
We can get a good indication of the performance differences when we look at the Ryzen benchmarks against the desktop skylake processors that include the new AVX instructions.
As an example in Cinebench
Ryzen 1800X achieves a single threaded score of 162 and a multithreaded score of 1637 (16 threads)
10:1 MT/ST ratio
Intel 6900K Also achieves 162 and a multithreaded score of 1490
9.12:1 MT/ST ratio
Intel i7-7770k (skylake) achieves 201 and 985 (with 8 threads only)
Lets assume a 16 thread skylake will run at the same 4.2-4.5Ghz and has double the multithreaded performance of the i7-7770k the multithreaded result would be 1970
9.8 MT/ST ratio
But are we likely to see a 16 thread skylake run at that frequency range?
From what I have read, the are creating a common socket for E5/E7 yes they may get a different name, but the E5 equivalent will have the AVX accelerator and the E7 the XML acceerator. Personally I'd love them to merge the lines as currently the software I deal with in only certified on E7 and the associated servers carry a large price premium.
I assume you mean the Skylake server release?
Not a massive jump in IPC from Broadwell to Skylake and Skylake to Kaby Lake even less so.
The E5 2699v5 is rumored to only have 28 cores (although a 32core rumor has also been spotted).
But the clock speed rumor is only 1.8GHz, So I think this will be neck and neck, Skylake with a little higher IPC but fewer cores, is may just come down to Multi-threading efficiency and frequency)
I'm really getting confused about the market Apple is trying to target these days.
Creatives? So if you do photographic/Film work you need lots of memory! But you can't go beyond 16GB.
If you want to be creative on the move and use all of that processing power, your battery won't last with such a thin device.
Any creative I know is very particular about desk appearance/tidyness, yet no Apple docking station, so you have endless adapters clogging up desk space that can easily fall out when working. And you probably have a number of external devices. Cameras/Screen Calibration devices/Tablets/Scanners, etc
Remember when Apple released the Dustbin MacPro with two Graphics cards? Where was the 5K monitor to go with it. They released the iMac with a 5K Monitor at the same time.
Business users? Well at this price point, strained corporate budgets will not stretch to these prices, and do you want to risk the chance of loosing that VGA/HDMI adapter before you turn up to a customer site?
Apple need to refresh MacOS to make it tough friendly and catch up with convertibles, consider who actually uses a MacBook Pro.
Alas they have become the B&O/Bose of the IT world when it comes to computers. Form over function!
MS have really managed to out Apple Apple with the new Surface range. (regardless what you thing of Windows 10)
Whilst AMD chips have not been able to compete with Intel at the high end, APUs have been great performers for the price.
However third party Laptop makers have only every paired these chips with crap hardware, preferring to put equivalent Intel chips into "Premium" laptops.
I for one am hoping the Zen performs as promised especially in the server market.
1) Kids keep taking it and whenever I'd like to use it it is out of charge
2) Still able to run all my apps without much slowdown after OS upgrades
3) Not a SIM enabled model so pretty useless when out and about unless I tether with phone
4) I always have my phone on me, which is usually charged and can do everything on that
5) Tablet screen is still too small to watch a movie and sound is crap (I don't always want to wear headphones)
So yes it's a nice to have but no need to upgrade every couple of years.
Every hour you spend building the infrastructure and calling vendor a) because the product from vendor b) or c) does not play ball you are not adding value to the business.
I love driving fast cars around a track. I have also in the past built myself a good track car (because I enjoyed it), but all the time I spend building the car I can't drive it.
In terms of compute and storage scaling there are now rack scale products out there where you don't need to scale compute and storage together, you can buy compute dense nodes and storage only nodes.
Yes build you own can be cheaper from a capex point of view but not from an opex point of view.
If (an unlikely scenario I know), Apple were to shrink wrap OSX (even with a tight hardware compatibility list) it would really hurt Microsoft. Given that phone/tablet integration with Microsoft is so poor compared to iOS/MacOS.
Set MacOS free and Apple would see even more adoption of iphone/ipad at the expense of Android as well (Android integration with desktop/laptop is pretty much non existent unless you use a google apps).
The biggest upset Apple could cause Microsoft is to sell OSX so it can be installed on any PC, They could even get away with a tightly controlled hardware list.
Undercut Windows 10 price, and build in greater integration with iOS.
As much as I hate itunes, the integration with the iphone/itouch/ipad for transferring content is so much easier compared to the crap you get on Android (so far no music transfer app has managed to copy 128GB of music to my Android phone, so I have to insert the SD card into my PC and transfer manually)
Windows phone music transfer to Groove Music has to be via Onedrive, sorry I'm not paying a subscription just to get music onto my phone.
Linux is not the answer to Windows 10. The UI does not appeal to the average user and you still have to use the CLI to get stuff working, which turns most people off.
Sorry if I want to buy a (hyper)converged platform, I want one number to call and support/test the entire stack. If something goes wrong I want that vendor to triage between the different component companies.
Nutanix are going to end up with far too many options and supporting the systems will be a nightmare.
They are spreading themselves far to thin.
In Both OSs, standard apps (photos, music, film, etc) have been replaced with new ones with less functionality.
Otherwise win 10 anniversary addition has not caused any issues. But what does piss me off is Cortana. When issuing voice commands, eg play jimi Hendrix it always looks for groove music subscription service and complains I'm not a subscriber, but if I type the same phrase it launches groove music and plays local content.
Biting the hand that feeds IT © 1998–2019