Amazon Makes It Almost Impossible To Calculate Their "Virtual CPU" Equivalent (informationweek.com)

← Back to Stories (view on slashdot.org)

Amazon Makes It Almost Impossible To Calculate Their "Virtual CPU" Equivalent (informationweek.com)

Posted by timothy on Thursday December 24, 2015 @11:24PM from the n-divided-by-q dept.

dkatana writes: AWS started out defining its virtual CPUs as being composed of EC2 compute units, or ECUs, which it defined as an equivalent to a physical Xeon processor. However, a virtual CPU now looks suspiciously variable... A virtual CPU is whatever Amazon wants to offer in an instance series. The user has no firm measure to go by. From the article: [B]y doing a little math, you could actually compare what you were getting in virtual CPUs in EC2 versus Azure. Also by doing a little math, you knew how to compare one Amazon instance to another based on the ECU count in each virtual CPU. Microsoft didn't look too bad in the comparison. That is one of the casualties of the nomenclature change. I have searched for updated information on how a virtual CPU is measured and found nothing comparable to the definition of the 2012 ECU measure. I have questioned Amazon representatives three times between Oct. 27 and Dec. 21, and don't have much of an answer."

114 comments

Min score:

Reason:

Sort:

Totally true by Anonymous Coward · 2015-12-24 23:35 · Score: 1

In fact Amazon doesn't know themselves... I have asked them to compare to the i7 series and they simply don't have a clue.
1. Re: Totally true by mSparks43 · 2015-12-25 03:24 · Score: 2
  
  when I did some testing a while back. $20 a month would get you processing of a crappie mobile you could buy for $50. Amazon has never had good cpu.
  so it's probably more that they don't want to say.
  anyone who's benchmarked them against alternatives has nothing good to say.
  http://openmymind.net/Why-I-Di...
2. Re: Totally true by lgw · 2015-12-25 04:56 · Score: 2
  
  That's a good link. My real complaint about AWS is what's mentioned at the end:
  
  It also feels like a lot of services are stuck at version 1.0, lacking that polish and continual improvement
  This is what annoys me. SQS is a good 1.0 version of message queues, but the features are the just above the minimum you could possibly call a message queue. DynamoDB is a good 1.0 version of a NoSQL DB, just above the minimum you could possibly call a NoSQL DB.
  These services are years old, but look like what most software does at version 1.1 or so: minimal features, no glaring bugs, but nothing great either.
  I can't say anything about EC2 CPU performance, as I've never benchmarked that myself, but aren't they just (mostly older) Xeons? The way the T2 instance is described makes it sound like the cores are oversubscribed, but I haven't heard that about the other instances.
  In any case, unless you're CPU-bound, it doesn't matter. I'm more concerned about price for the memory I need. For compute-intensive jobs, EC2 Spot is cheap if you're really fault tolerant.
  
  --
  Socialism: a lie told by totalitarians and believed by fools.
3. Re: Totally true by mSparks43 · 2015-12-25 08:17 · Score: 1
  
  each insrance is one of who knows how many virtual machines running on old xeons.
4. Re: Totally true by lgw · 2015-12-25 13:26 · Score: 1
  
  Right, but they clearly promise you a number of cores. The easy assumption is that the cores aren't oversubscribed, and none of the docs suggest they're oversubscribed, except the docs for T2 (seems like it must be). Can't tell for sure, of course, unless someone can find something definitive from Amazon.
  
  --
  Socialism: a lie told by totalitarians and believed by fools.
5. Re: Totally true by mSparks43 · 2015-12-26 11:29 · Score: 1
  
  what they promise is by "virtual core".
  that has no L1/2/3 cache.
  and will be balanced on the machine based on actual cpu cycles (my "real" 3.5ghz cpus can "turbo mode" to nearly 7ghz on heavy loads - Linux cpu-governor)
  it's an extreme version of the old hdd spec lies.
  http://tiemensfamily.com/TimOn...
  300...
  compared with 10s of thousands for a "real" core.
6. Re: Totally true by lgw · 2015-12-26 14:03 · Score: 1
  
  I can't make much sense of your post, and the linked article is a bit confusing as well. But if the claim is that the 2 cores from an m1.large are about what you'd expect from 2009, I can totally believe that - the m1 instance is pretty old. I'd expect them to be sub-2-GHz Xeon cores. Xeon is usually damn slow compared to the consumer cores from the same year, and it's not sensible to compare Xeon performance to consumer performance (or, at least, it's Intel's brain damage, not Amazon's).
  From the docs, m4 is 2.4 GHz Haswell, and m3 is Ivy Bridge - they don't say the frequency so probably pretty low.
  As far as cache, I don't know what hypervisor Amazon uses. I know VMware (which I'm sure they don't use) protects the CPU cache as long as you don't oversubscribe cores. "World switching" a core, as they call it, is the most expensive operation you can do in a hypervisor, so any hypervisor worth a damn will try to avoid that.
  
  --
  Socialism: a lie told by totalitarians and believed by fools.
7. Re: Totally true by godefroi · 2015-12-28 03:46 · Score: 1
  
  I believe they use Xen.
  
  --
  Karma: Poor (Mostly affected by lame karma-joke sigs)
Quit whining by Anonymous Coward · 2015-12-24 23:35 · Score: 0, Funny

If you don't like it use something else
1. Re:Quit whining by pijokela · 2015-12-24 23:42 · Score: 2
  
  That is not really possible. Amazon is getting to a monopoly position if you want to use all the modern Cloud stuff. Sure, you can get boxes from many different providers, but AWS has a ton of other services that you cannot buy from others and even more importantly, all the 3rd party Cloud services are running on AWS so they are faster if you are on AWS too.
  I would not be surprised if in 10 years Amazon would be a verb for running server software like google is now for search.
2. Re: Quit whining by t1oracle · 2015-12-24 23:53 · Score: 1
  
  Google and Microsoft will always have their own clouds. Those are two real competitors.
3. Re:Quit whining by vikingpower · 2015-12-25 00:05 · Score: 4, Interesting
  
  Alternative: build your own private cloud out of the smallest servers you can find that still suit your need. I did so: a private cloud on HP Microserver (gen 8). The things consume almost no power when idle. Taken together, they provide quite the computing power ( 64 cores, Xeon E3 ) and quite the storage (32 TB). Cost me around € 450 in electric power per year. Am not dependent on Azure or Amazon. Use no bandwidth when doing cloudy things.
  
  --
  Religous speak to God. Insane are spoken to by God. When all shut up, one can finally hear Shostakovich in peace
4. Re: Quit whining by DThorne · 2015-12-25 00:36 · Score: 3, Insightful
  
  I'm not sure I'd call it whining - it's a significant investment for many to set up cloud computing the way they need it so I think it's a fair demand that if your gas station insists on measuring it's product in frackles that it give you a fair conversion rate for litres to frackles. My needs would be something predictable like rendering 3d images, so the first thing I'd be doing is measuring render times for the same image there and here. That's easy and relatively cheap to do, and the numbers scale directly, but I can see this being a serious issue for other uses.
5. Re: Quit whining by Anonymous Coward · 2015-12-25 00:48 · Score: 0
  
  All clouds are data mining scams, Amazon's isn't even good.
6. Re:Quit whining by Anonymous Coward · 2015-12-25 00:49 · Score: 0
  
  There are several services that compete, they just aren't as well known or are avoided. Of the boy boys, Google are shit at this stuff, so much is lacking, so any sane designers skip them; and we all know what Microsoft are like for vendor lock-in and moving targets, so they're fucked off at the start except the MS-only shops.
  The future is CDNs and caching networks, everything is specific or small time. The reality is almost no one needs "cloud" for power or scale, and if they do, they'll have their own bespoke DCs scattered around the world.
  You're a moron if you think the biggest online store on the planet will become a verb like "google" for a service 99.9999% of the people aren't even aware exists.
7. Re: Quit whining by Anonymous Coward · 2015-12-25 00:52 · Score: 4, Interesting
  
  Microsoft is a joke and since Satya insist on moving everything to India, treating people like shit, and firing everyone after a few years for "IP" reasons, it's a trainwreck of a company that only still survives because of its initial brand-name and market penetration.
8. Re:Quit whining by Anonymous Coward · 2015-12-25 00:54 · Score: 0
  
  Yeah! Stop asking for better products and information. Bow down to your corporate overloads and hand over all your money for whatever they deign to allow you to purchase!
9. Re:Quit whining by ShanghaiBill · 2015-12-25 01:48 · Score: 2, Informative
  
  Alternative: build your own private cloud
  The whole point of doing computing in the cloud is to handle variable demand. I spend 98% of my time writing and debugging code, and only 2% running the final model. But when I run it, I want to do it at scale. On AWS I can rent a dozen K80 GPUs for a few hours a week. There is no way it would make sense to own them, even if I could afford to do that.
10. Re:Quit whining by Anonymous Coward · 2015-12-25 02:32 · Score: 0
  
  What the fuck don't you understand about 'use something else'?
  Corporations don't give a shit about people asking for things - all they care about is the bottom line. If enough people use someone elseâ(TM)s product and then they will change.
11. Re: Quit whining by NormalVisual · 2015-12-25 02:39 · Score: 2
  
  And with all the crap they're pulling with the telemetry and whatnot in Windows 10, they seem hell-bent on losing that as well.
  
  --
  Please stand clear of the doors, por favor mantenganse alejado de las puertas
12. Re: Quit whining by Anonymous Coward · 2015-12-25 02:44 · Score: 0
  
  Lucky we don't want any fancy cloud crap because it's expensive and insecure.
13. Re:Quit whining by Anonymous Coward · 2015-12-25 04:55 · Score: 0
  
  If you don't like it use something else
  "Love it or Leave It" is a common sentiment, but often not one's optimal choice, as per Albert O. Hirschman's "Exit, Voice and Loyalty."
14. Re: Quit whining by Anonymous Coward · 2015-12-25 05:01 · Score: 0
  
  Microsoft has about half as much of the market by all reports as Amazon. They are definitely no joke and giving Amazon a run for it's money. They have a comparable set of services and are not tied to Windows. Do your homework before talking out of your ass.
15. Re: Quit whining by lgw · 2015-12-25 05:06 · Score: 1
  
  It's odd how much Azure seems to struggle to keep up with AWS. MS has no shortage of cash, and at the very least they could offer "MS SQL instances in the cloud" as cheaply as they needed to go get people locked in, It's not like NoSQL and message queues are rocket science these days, and the AWS versions have fairly minimal feature sets.
  In the old days MS was all about lock-in, and while you never wanted to touch v1.0 of anything from MS, by the time 2.0 came out they were usually ahead of the pack in terms of feature checklist (quality not so much). I guess they've lost that spirit.
  Google spent so long just offering a limited set a quirky services in the cloud. It seems like they're smart enough to stomp the competition: it always baffled my why they didn't. (They certainly know how to operate at scale as good as anyone.)
  
  --
  Socialism: a lie told by totalitarians and believed by fools.
16. Re: Quit whining by Lumpy · 2015-12-25 05:49 · Score: 1
  
  But they not only measure in frackles but they also break your fuel gauge so you dont know how much they actually put in your car.
  
  --
  Do not look at laser with remaining good eye.
17. Re:Quit whining by Anonymous Coward · 2015-12-25 06:13 · Score: 0
  
  Amazon is getting to a monopoly position if you want to use all the modern Cloud stuff.
  They are? I thought they were losing immense amounts of business to all of the other sizable players in the space (GAE, Azure, IBM, Oracle), had a terrible reputation for performance, price, stability and features. But that's just the word on the street.
18. Re: Quit whining by Anonymous Coward · 2015-12-25 06:21 · Score: 0
  
  What? Have you even seen MS's offerings? They can't build DCs fast enough because so many people are signing up. Their stuff is at least in par with Amazon if not better.
19. Re: Quit whining by Anonymous Coward · 2015-12-25 06:39 · Score: 0
  
  They include in their numbers Office365 and exchange subscriptions.
20. Re:Quit whining by Anonymous Coward · 2015-12-25 07:12 · Score: 0
  
  Why not go back to the old way of doing things and have a cloud platform for all the developers in the whole organization? Reserve a slot and go nuts. Sure beats giving someone else extra profit to do the same thing for you.
  If you are independent, why not collaborate with people you are familiar with on github to do the same thing? The cost savings would be worth it.
21. Re: Quit whining by Cederic · 2015-12-25 07:17 · Score: 1
  
  (They certainly know how to operate at scale as good as anyone.)
  Operating at scale when you're bespoke from the chip up is easy.
  Well, relatively easy.
  Operating at scale when any twat out there can do whatever the fuck they like to your commodity infrastructure (which must be relatively commodity or you can't sell it) is a very different proposition.
  I'm confident Google could step into that market if they chose, but it's a deviation from their standard operating model.
22. Re:Quit whining by Cederic · 2015-12-25 07:32 · Score: 1
  
  We have multiple tier 5 data centres. Amazon/Azure is still useful for dynamic scaling for some of our workloads.
  Some of the more optimised and specialised services can also really exploit cloud pricing. When part of our business can service millions of consumers for a few hundred dollars a year with Amazon we just can't get close to that price point in house.
  But fuck it, The Good, The Bad and The Ugly just started on TV so I'm going to kick back, drink some vodka and enjoy cinematic perfection.
23. Re: Quit whining by Anonymous Coward · 2015-12-25 07:36 · Score: 0
  
  I'm not so sure about that, but who cares? They still have 15 + percent of the market and growing. Amazon has stagnated at 25 to 30 percent.
24. Re:Quit whining by Anonymous Coward · 2015-12-25 08:01 · Score: 0
  
  Disagree. The whole PROPER point of doing computing in the cloud is to handle variable demand. That's a good reason to rent anything, including capacity.
  The whole reason companies actually do so boils down to stuff executives read in airplane magazines, buying into huckster fueled hype, and listening to Gartner and suddenly believing they're IT experts. So the jackasses end up spending per month what they could spend once every three to five years on non-scaled workloads just so some genius can write in the company newsletter or a press release that they "moved to the cloud".
25. Re: Quit whining by mlts · 2015-12-25 08:21 · Score: 1
  
  Microsoft survives for one reason: They are so well entrenched in the enterprise. Other than small companies that LDAP can work with, AD will be found as the core authentication and management mechanism of most companies out there.
  Because of this, if MS sees losses on other fronts, they can just ratchet up Windows Server license fees, and still come out ahead, as they have a captive audience.
26. Re: Quit whining by soulhuntre · 2015-12-25 08:36 · Score: 1
  
  What? Have you even seen MS's offerings? They can't build DCs fast enough because so many people are signing up. Their stuff is at least in par with Amazon if not better.
  This. All of this. More and more clients who traditionally went AWS (advertising campaign back ends, social media startups and so on) are picking Azure. The AWS tools are just crude in comparison and the Azure offerings are typically more complete, more robust and much better documented.
  3-4 years ago suggest Azure hosting was suicide for a potential contract, these days it is an advantage and makes the guys pushing AWS look like they aren't keeping up. The perception is shifting fast.
  
  --
  --> Fight tyranny and repression.... read /. at -1!
27. Re:Quit whining by Anonymous Coward · 2015-12-25 08:49 · Score: 0
  
  With HP would update their MicroServer line to the G9s... but other than that, I have seen those in use as well as at a local Linux convention. Those are nice little servers for when rackmounted boxes isn't needed. They don't break the bank either.
  HP has a lot of cool technologies, the Moonshot rack that can stuff 45 blades, each the size of a VCR cartridge, in ~5 rack enclosure.
  Public clouds are great for variable/peak loads, but for base computer tasks, having those servers in house is better overall, especially with the fact that one has physical control of security. Since the servers will be paid for, either at the cloud provider's location or at the business's location, might as well keep all but peak load in-house.
28. Re:Quit whining by vikingpower · 2015-12-25 18:07 · Score: 0
  
  Disagree. For a developer, the whole point of computing in the cloud is to be familiar with it. To that end, a private cloud makes eminent sense. And "at scale" then means: 8 servers (HP MicroServer Gen8).
  
  --
  Religous speak to God. Insane are spoken to by God. When all shut up, one can finally hear Shostakovich in peace
29. Re: Quit whining by Anonymous Coward · 2015-12-26 11:54 · Score: 0
  
  Google and Microsoft will always have their own clouds. Those are two real competitors.
  With any luck, Microsoft will be dead and buried by then.
Uhhh by Anonymous Coward · 2015-12-24 23:37 · Score: 0

Who cares? Your stuff is gonna get hacked anyways and the Internet (in the US) is slower than dirt due to crappy ISP throttling.
Mostly irrelevant to most people by AmiMoJo · 2015-12-24 23:42 · Score: 3, Interesting

Most people don't care about the exact performance, so it's not with spending the money and effort to precisely define or guarantee it.
Amazon is generic and cheap. Microsoft has really good integration with visual studio and .NET. Those are the factors people choose by.

--
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
1. Re:Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 00:33 · Score: 0, Insightful
  
  Microsoft is shit. PERIOD!
2. Re:Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 00:33 · Score: 1
  
  Considering that most modern CPUs have performance counters and such, there is no excuse not to be able to measure it, particularly when you are charging for it. How big is this house you want to sell me? Well, somewhere between 100 and 300 sq. m., nobody cares, so why are you whining about it, am I right?
3. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 00:38 · Score: 1
  
  CPU built-in performance counters are as reliable as the size of the fish catch described by your alcoholic uncle.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
4. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 00:43 · Score: 4, Funny
  
  After rereading my post, I realized I needed to post a clarification for the easily offended: "your alcoholic uncle" is used as a generic example and not referring to the AC's uncle.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
5. Re:Mostly irrelevant to most people by StormReaver · 2015-12-25 00:44 · Score: 1
  
  This behavior is what I (and others) predicted when this whole "cloud" stupidity began: as providers become more entrenched, the prices and complexity slowly increase. Those who become dependent on this particular brand of idiocy will be taken to the cleaners.
6. Re:Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 00:45 · Score: 0
  
  Consistent performance does matter.
  Try logging in to an ec2.micro instance someday, that's a half-hour on Windows.
  What happens is you get 2 full cores for 10 seconds, then 98% of the CPU is not available.
7. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 00:48 · Score: 0
  
  For one, yes, I don't have an uncle.
  Also, can you please explain why not? For sure I can imagine OS-related things going on in the background, but are you really saying that a counter on the CPU of FLOPs is completely random? Certainly the business of optimizing algorithms depends on these things.
8. Re:Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 00:52 · Score: 0
  
  Cheap? Amazon is wildly overpriced, have you seen how much they'd have you pay for bandwith? You better be serving static 1990's HTML and nothing else.
  The benefit of Amazon is that they have a very friendly UI and an easy solution for almost everything. But the price per month is far from competitive.
9. Re: Mostly irrelevant to most people by Dunbal · 2015-12-25 00:55 · Score: 1
  
  Fuck the easily offended politically correct cunts. They are useless anyway.
  
  --
  Seven puppies were harmed during the making of this post.
10. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 00:57 · Score: 0
  
  I'll admit that the memory bandwidth benchmarks have a fair amount of scatter to the results, but I have found FLOP counts and cache misses to be fairly reliable measures. If you don't measure the performance of your code, how in the world can you know if it is any good? I assume of course that you have a job big enough that you would want to use Amazon's servers instead of your cell phone or your Beowolf cluster from 20 years ago.
11. Re:Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 00:59 · Score: 0
  
  I wouldn't say generic. Amazon's cloud ecosystem dwarfs anything else out there. They've got some amazing tools.
12. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 01:01 · Score: 0
  
  Says the person with such a useful post. Can you explain why modern CPU performance counters are not sufficient? Because as someone that works on optimizing CPU performance, I would be glad to learn more. Performance measurements and algorithm benchmarks have changed a lot in the last few years.
13. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 01:39 · Score: 2
  
  The uncle clarification was to make sure that you and others knew I wasn't just trying to be a troll. When I get marked as troll, I want to earn it. As far as the counters, it's not that their random, it's that they are optimized to give inflated tallies for benchmarks. So, in some areas, they're going to vastly overstate their performance (and thus, the "fish story" analogy) but, when you actually call on that level of performance, it isn't there. Which areas that do this will vary by manufacturer and model. The companies that try to offer serious and fair benchmarks have had a nightmare time trying to resolve this. They often have to reverse-engineer where each processor is lying, figure out an estimate of percentage it is lying on that particular test and modify the final results by that percentage. Of course, the company that made the chip will say the adjustment is unfair, then they rebuttal, etc. If that alone wasn't bad enough, chips (though this is more common on graphics chips) that are labeled as mid-range are, in fact, higher-end chips that have been stepped down by firmware, that remove the firmware imposed speed limiter for benchmarks only. These are just the examples I know of; I have no doubt there are more.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
14. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 01:58 · Score: 0
  
  By what do you mean "optimized to give inflated tallies"? It's true different operations take different amounts of time on a CPU clock, but I don't have a clue what that sentence of yours means. Are you saying if you set the correct counters for making roofline plots with PAPI or whatnot, the CPU is going to lie to you?
  And graphics chips have no performance counters. If you're comparing CPU and GPU performance counters, your knowledge is worth the water that fell through the net when my uncle got the fish.
15. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 02:08 · Score: 0
  
  For reference, see e.g., http://www.spiral.net/software/roofline.html or http://crd.lbl.gov/departments/computer-science/PAR/research/roofline/ . This is something that you can't do on any GPU in 2015. (I admit that I don't do much on GPUs myself, but colleagues at a CUDA research center confirm this.)
16. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 02:15 · Score: 0
  
  As an alcoholic uncle I resent that statement
  At least people know when I am lying!
17. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 02:45 · Score: 1
  
  When it comes down to names of CPU calls, I'm in over my head. My former father in law is the engineer, not me. He said the specifics, I just couldn't retain them. But, the short was, when it came to internal performance marks, CPUs lie. Whether that specific function call does (or even can) lie and which CPUs do it, I'm unable to say due to being out of my area of skill. Sorry about the graphics kludge in my statement. I was mixing apples and oranges. I was meaning that both types lie, in their own way, to fool performance benchmarks, not that they had the same type of internal metric. I've followed this nightmare more from the side of the benchmark suites, not from the engineering side.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
18. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 02:47 · Score: 2
  
  Unrelated side comment: Where the hell are all my line breaks going?
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
19. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 03:11 · Score: 1
  
  OK, you win the internet for me today.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
20. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 03:25 · Score: 1
  
  First link didn't work but, the second one did. Thank you, that was informative. Sadly, I don't have the background to follow some of it. It's helpful to have this now assuming the manufacturers don't try to game it too much. Of I hadn't already posted, I would mod you up.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
21. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 04:00 · Score: 1
  
  I found the fix. ...and it sucks. Apparently, Slashdot now carries over your desktop posting preferences to the mobile site. This didn't used to happen. My desktop was set to HTML codes posts. There is no way I'm imputing proper HTML code using a smartphone touchscreen keyboard!
  So, to the fix. You can only do this on the desktop version of the site. There is no mobile accessible setting for this. Login to the desktop site, open any story, then reply to any post. Select options from the post window (not the main Slashdot title bar) then go to the Formatting drop down box in the pop-up window and select "Plain Text" then click the save button at the bottom of the window (you may have to scroll to see it). You will still be able to put in HTML tags but, you don't need to post proper code anymore.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
22. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 04:35 · Score: 1
  
  Hey, I've been staying analytical and being honest when I'm outside my skill set. The personal attack is uncalled for.
  You don't need to take my word on it for benchmark cheating. A Google search will do. Here's an older one but, it went all the way to a class-action suit that Intel had to settle: Intel lawsuit. The most recent article about CPU cheating was 2009 though GPU cheating is still rampant in the news. On the CPU side they have either cleaned up their act enough to stay out of the news or just haven't been caught. I can't say for sure which.
  Though none of this will be detailed enough for you as your knowledge is more specialized and most news is not going to go to that level. I can't fix that.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
23. Re: Mostly irrelevant to most people by Dunbal · 2015-12-25 05:37 · Score: 0
  
  There are different types of usefulness. Your post, for example, which attempts to impress the reader with your apparent technical knowledge is only public masturbation rather than adding anything further to the discussion. It has nothing to do with political correctness which was brought up by GP and addressed by myself. However if you wish to find the knowledge you seek I suggest other threads or failing that, perhaps another source than slashdot.
  We're all specialists son. In fact I am willing to bet I have higher doctoral and post doctoral qualifications than you. Go masturbate somewhere else.
  
  --
  Seven puppies were harmed during the making of this post.
24. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 06:14 · Score: 0
  
  There is nothing left to game for pure number crunching. In the beginning, we could just count FLOPs and all was fine. Then computers changed and we all knew that it was meaningless in the 90s and 00s, and models of performance benchmarks became so complicated nobody could use them. About 5-10 years ago some people figured out you could get an estimate of theoretical performance for a given algorithm within about 5-10% error based on max FLOP/s and max memory bandwidth on a given system.
  It is just applying scientific principles to computation.
25. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 06:23 · Score: 1
  
  Today's complexity creates a lot of the trouble in getting reliable performance data. That makes perfect sense to me.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
26. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 06:37 · Score: 0
  
  And the CISA got passed so now the government doesn't even need a warrant to take your stuff. Good luck with your cloud goals of 2016. Fuck that, I'm happy we built our own private cloud that we own everything on.
27. Re: Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 07:08 · Score: 0
  
  (Not the same AC.)
28. Re: Mostly irrelevant to most people by TheReaperD · 2015-12-25 07:13 · Score: 1
  
  Sorry. That's the hard part of having a long conversation with an AC.
  
  --
  "Be particularly skeptical when presented with evidence confirming what you already believe." -
29. Re: Mostly irrelevant to most people by Cederic · 2015-12-25 07:38 · Score: 1
  
  I'm not a specialist. I took a career choice to become a generalist, on the grounds it lets me get involved in solving problems with a broader scope and greater scale. I call this fun.
  I have no post-graduate qualifications either. I got on with learning pragmatic useful shit instead.
30. Re: Mostly irrelevant to most people by BronsCon · 2015-12-25 07:51 · Score: 1
  
  Of course, your post wasn't at all masturbatory.
  
  --
  APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
31. Re: Mostly irrelevant to most people by Dunbal · 2015-12-25 09:44 · Score: 0
  
  More akin to sodomy I think.
  
  --
  Seven puppies were harmed during the making of this post.
32. Re:Mostly irrelevant to most people by Anonymous Coward · 2015-12-25 23:54 · Score: 0
  
  Azure is expensive, especially if you want to scale. The lowest cost (and useful because it's the first tier to include HTTPS) is $75/mo. That's very expensive compared to Amazon, where the least expensive node with HTTPS is only $9.50/mo or so. So on Amazon, you can have a load-balanced, auto-scaling solution for $30-$40 per month for your application, while in Azure you have to spend at least $150 per month just to get started.
  Their Azure Database performance is also crap. Unless you pay a few hundred per month, you can't get the IOPS that you need.
  Azure is expensive enough that we're having to put multiple web applications on the *same* instance. And for most of our applications, it's too expensive to even think about load-balance / redundancy / auto-scaling.
  So for the small-medium business category, Azure is far less of a value then Amazon.
33. Re: Mostly irrelevant to most people by phorm · 2015-12-29 06:54 · Score: 1
  
  I try to use the phrase "one's alcoholic uncle" which is less likely to be misconstrued as personal. :-)
Variable Units by TheReaperD · 2015-12-24 23:43 · Score: 1

I read this a they're worth whatever they say it's worth (at the moment) and you're going to whine and complain about it but, not do anything about it. Until they start loosing a large dollar value of customers over it, they're not going to fix it.

--
"Be particularly skeptical when presented with evidence confirming what you already believe." -
what a surprise by climb_no_fear · 2015-12-25 00:19 · Score: 2

that the "Cloud" is nebulous....
So just takes the worst measurement you can find? by WoLpH · 2015-12-25 00:34 · Score: 1

Imho there is only 1 valid measurement for cloud solutions, how much will certain performance cost you. And you can only really find out by testing it for _your_ specific goal. Some apps require more cpu, others more disk, others more ram... there's no single number to indicate your price/performance index.
If you need well defined performance... by drolli · 2015-12-25 00:45 · Score: 3, Insightful

then buy dedicated instances.
I like micro instances/instances which do not occupy full physical processors at Amazon because of availability and price for low-impact/bandwidth applications. For all other use lambda or dedicated instances.
Virtual CPUs are anyway difficult to asses - to me it may be very relevant to have the 1st level cache of the core which i run on undisturbed by other applications (since changing the cache hits is a big deal for specific numerical problems), and for you 20% more share of the CPU may be important.
1% of computation time not spend in my task on a physical processor can do as much damage as 50% change in speed.
A small side remark: the price for the different VCPUs also varies.
1. Re:If you need well defined performance... by Sarten-X · 2015-12-25 03:09 · Score: 1
  
  This is exactly the problem, and solution.
  From a computer science standpoint, modern algorithms that rely on single-thread speeds are obsolete and should be avoided for new development. From a sysadmin perspective, CPU speed is the least important metric compared to the rest of the system components. For project management, the system management toolset and feature support should be the prime concern.
  Amazon's never really tried to hide the fact that their hardware is dissimilar behind the scenes. That will impact your computation speed, naturally.
  
  --
  You do not have a moral or legal right to do absolutely anything you want.
welcome back to the mainframe by known_coward_69 · 2015-12-25 00:50 · Score: 0

beg for CPU time and be happy when the admins throw you a little bone just like back in the 70's
1. Re:welcome back to the mainframe by Anonymous Coward · 2015-12-25 01:35 · Score: 1
  
  This is a scam by ISPs. They won't give you a high-bandwidth uploading network connection at your home or office. So you have to pay rent to run some slow computer at some remote facility.
  It's not 1990 anymore... why can't you host a server at your own place.
2. Re:welcome back to the mainframe by known_coward_69 · 2015-12-25 01:45 · Score: 1, Insightful
  
  where can't you get high speed connections? NYC and LA are up to 300mbps and higher for time warner. other markets as well.
3. Re:welcome back to the mainframe by Anonymous Coward · 2015-12-25 02:56 · Score: 0
  
  We share a pair of ISDN lines between just over twenty developers in downtown Seattle. That's the fast connection available in our building since Comcast doesn't provide service to the block, despite having a government granted monopoly, and CenturyLink is at capacity and can't get permission from the city to dig up the street. At my previous job, we shared a T1 between about sixty people because of nearly the same reasons. Not everywhere can get high speed connections, but your millibit example is just ridiculous. You can transmit faster than that with smoke signals.
4. Re:welcome back to the mainframe by Anonymous Coward · 2015-12-25 03:01 · Score: 1
  
  where can't you get high speed connections? NYC and LA are up to 300mbps and higher for time warner. other markets as well.
  The first commonly used modems were a *thousand* times faster than that. Connections here in the US are very slow, like the typical 1.5 Mbps T1 or DSL connections in the Seattle area for businesses, but the claim that it takes more than three seconds to transmit a single bit, is a lie.
5. Re: welcome back to the mainframe by Anonymous Coward · 2015-12-25 03:32 · Score: 0
  
  > 300mbps
  At that speed, it would take 53,333 seconds to send a standard 80x25 page of text. Sounds like you irrationally hate the US and are lying to try to promote your position.
6. Re: welcome back to the mainframe by Anonymous Coward · 2015-12-25 03:45 · Score: 0
  
  Actually you're off by a factor of 2.73. The GP didn't lie as badly as you claimed. The first common modems were 110 baud, which is not a thousand times faster than his claim.
7. Re: welcome back to the mainframe by Anonymous Coward · 2015-12-25 03:48 · Score: 0
  
  Screw you. So he lied by a factor of a thousand or by more than 300. The point is still the same. He is a damn liar. Connections in the US are many magnitudes faster than what he claimed.
8. Re:welcome back to the mainframe by tepples · 2015-12-25 05:26 · Score: 1
  
  Seattle's zoning law is anti-utility. For acceptable Internet performance, leave Seattle.
9. Re:welcome back to the mainframe by Anonymous Coward · 2015-12-25 08:12 · Score: 0
  
  You can get high speed internet anywhere, it's just a matter of how stupidly expensive it is. You can find an ISP to run fiber anywhere for the right price.
10. Re: welcome back to the mainframe by Anonymous Coward · 2015-12-25 10:17 · Score: 0
  
  Sounds like you've never been to Seattle. Even the phone monopoly loses when fighting the city to try to get permission to run cable. At the startup where I've worked for nine years, our T3 order is nearly nine years late. The city simply will not let Seattle upgrade phone lines.
11. Re: welcome back to the mainframe by Anonymous Coward · 2015-12-25 10:33 · Score: 0
  
  We've been waiting for nearly sixteen years for an upgrade to our T1.
12. Re:welcome back to the mainframe by Anonymous Coward · 2015-12-25 13:37 · Score: 0
  
  Thanks for the advice, Captain Douchebag.
13. Re: welcome back to the mainframe by Anonymous Coward · 2015-12-26 03:56 · Score: 0
  
  I've got a business dedicated 100/100 fiber to my house for $90/month, with worldwide jitter in the low single digit milliseconds, 24/7, and mainland USA jitter in the fractional milliseconds.
Which specs? by Anonymous Coward · 2015-12-25 00:50 · Score: 0

The 'bursting' nature of the new AWS specs are straightforward enough: if your application is 'bursty' in nature - that is, it wants bursts of CPU power between long idle intervals - then the shared vCPU options are viable and while you're running your spec might very well exceed that of an ages-old 1GHz Xeon processor.
But specs are a difficult thing to nail down: are you focused on matching the math processing specs of the old processor? The I/O specs? Do you expect a more modern CPU to be slowed down so that your application takes the same amount of time to execute as it would on a Xeon?
Personally, I'd rather my application complete serving a web page five times faster, then sit idle for four times as long (while the hypervisor has the CPU off serving other vCPUs) than to match the spec of a Xeon processor.
How much can you pay? by Anonymous Coward · 2015-12-25 00:54 · Score: 0

But that's not how it works. He makes full use of EC2, there is a pain associated with moving. They estimate the pain and charge $1 dollar less.
Extracting the maximum that any company will pay for that service.
It's the old IBM mainframe model, do you think it makes sense for any company to use an IBM mainframe with less power than a normal server at a bucketload more cost? No, its the pain of switching that's expensive, so IBM milk as much as possible.
1. Re:How much can you pay? by Anonymous Coward · 2015-12-25 01:21 · Score: 1
  
  Except there's still a lot of potential market to capture. Microsoft and Google can charge less and get those people before they get locked in to Amazon.
Newsflash! by Anonymous Coward · 2015-12-25 01:29 · Score: 0

Newsflash! People applying non-cloud mindsets and methodology to cloud computing having trouble making sense of Amazon.
Seriously the CPU units are not a mystery, and plenty of people (myself included) have managed to do very successful capacity planning in AWS.
Why not just benchmark it? by Anonymous Coward · 2015-12-25 01:33 · Score: 0

Make a cloud benchmark tool and run the same code on different instances....
MIPS! by Anonymous Coward · 2015-12-25 02:05 · Score: 0

MOPS and such.. Perhaps there is a need to implement a negligible cost measurement infrastructure providing some industry standard and normalized metering in those cloud instances.
AWS by fluffernutter · 2015-12-25 02:38 · Score: 1

I was considering using AWS... but it seemed to complicated with their calculator to get any kind of estimate of what it would cost me I went with dedicated servers instead.

--
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
1. Re:AWS by Anonymous Coward · 2015-12-25 03:24 · Score: 0
  
  I was considering using AWS... but it seemed to complicated with their calculator to get any kind of estimate of what it would cost me I went with dedicated servers instead.
  Calculators ARE pretty hard. That is why I use AWS.
  I like AWS.
  AWS is like Jaws (we need a bigger bot), although normally there's less blood involved and the running time tends to be a tad longer.
Microsoft makes it almost impossible to be license by mrmaster · 2015-12-25 03:21 · Score: 2

Big deal. Try asking azure support about licensing and they will flat out say I don't know. Assuming you bugged them at least a half a dozen times about the same questions beforehand.
Article is wrong on two things by pgn674 · 2015-12-25 03:35 · Score: 1

> "A virtual CPU is whatever Amazon wants to offer in an instance series." No. The vCPU (Virtual CPU) aspect of an AWS EC2 Instance is the county of virtual cores that are exposed to an OS. In desktop computers, a quad core Intel CPU will appear to have four courses when looked at from inside the OS (my go-to way to count them in Linux is to run top and press 1). A quad core hyperthreaded Intel CPU will appear to have 8 cores. The vCPU metric simply tells you what the OS will show you, and tells you how many processor threads can run concurrently. > "If you deal with server sizing and instance price comparison, then the measure -- previously expressed as an EC2 Compute Unit or ECU -- is kaput." ... "It's the closest thing you'll find to an acknowledgement that ECUs are still in use behind the scenes, but Amazon no longer wishes to define them due to the changing nature of its underlying hardware." Yes, ECU (Elastic Compute Unit) metrics are still used behind the scenes, but Amazon does publish them. Even for new Instances. Check out these URL's: http://a0.awsstatic.com/pricin... http://a0.awsstatic.com/pricin... Of course, this isn't very parsable by human eyes. So someone started an open source project to display this data, and its available at http://www.ec2instances.info/ So yeah, TFA is wrong.
1. Re:Article is wrong on two things by pgn674 · 2015-12-25 03:47 · Score: 1
  
  Ugh, sorry about formatting. Here, I think this is right:
  
  "A virtual CPU is whatever Amazon wants to offer in an instance series."
  No. The vCPU (Virtual CPU) aspect of an AWS EC2 Instance is the county of virtual cores that are exposed to an OS. In desktop computers, a quad core Intel CPU will appear to have four courses when looked at from inside the OS (my go-to way to count them in Linux is to run top and press 1). A quad core hyperthreaded Intel CPU will appear to have 8 cores. The vCPU metric simply tells you what the OS will show you, and tells you how many processor threads can run concurrently.
  
  "If you deal with server sizing and instance price comparison, then the measure -- previously expressed as an EC2 Compute Unit or ECU -- is kaput." ... "It's the closest thing you'll find to an acknowledgement that ECUs are still in use behind the scenes, but Amazon no longer wishes to define them due to the changing nature of its underlying hardware."
  Yes, ECU (Elastic Compute Unit) metrics are still used behind the scenes, but Amazon does publish them. Even for new Instances. Check out these URL's:
  http://a0.awsstatic.com/pricin...
  http://a0.awsstatic.com/pricin...
  Of course, this isn't very parsable by human eyes. So someone started an open source project to display this data, and its available at http://www.ec2instances.info/
  
  So yeah, TFA is wrong.
This is not a big deal by Anonymous Coward · 2015-12-25 03:46 · Score: 0

Amazon used to publish things in terms of ECUs and they switched to just telling you what the hardware actually is. I have no idea why that would not be considered an improvement. The article mostly complains about the t series being vague but that is no surprise - the t series has no fixed amount of cpu time. They have burstable cpu time that allows extra performance for short periods but a tiny amount of continuous cpu.
Mbps vs. mbps spelling lame by tepples · 2015-12-25 05:27 · Score: 1

Back when Usenet was popular, members of certain groups in alt. used to call that kind of post a "spelling flame" or later a "spelling lame". If you want to complain about mbps vs. Mbps, please also have something helpful to say about the rest of the post.
How about OpenStack? by Anonymous Coward · 2015-12-25 08:15 · Score: 0

How about OpenStack, which can be considered a bug-free cloud for a bug-ridden world?
1: Their code base is proven for production.
2: Their support for redundant VMs and moving them across clusters is top notch.
3: OpenStack works flawlessly out the box.
4: KeyStone authentication is bulletproof, beats IAM on every front.
5: OpenStack offers two ways to store data -- Swift for objects, Cinder for block-level I/O.
If OpenStack wasn't going to crush AWS in the next year or two, IBM, HP, and RackSpace wouldn't be betting the farm on this.
1. Re:How about OpenStack? by Anonymous Coward · 2015-12-25 10:06 · Score: 0
  
  IMHO, the losers in the cloud market are joining forces around OpenStack. This rarely ends well. In fact, I don't think it's ever ended well. There's usually a reason why the winners won and losers lost, and "best technology" is rarely the deciding factor. Gathering together a bunch of losing technologies under one umbrella rarely makes them winners.
  I'm not passing judgement on the objective merits of OpenStack. Maybe it's the best thing since sliced bread. Maybe it's better than sex. Doesn't matter. The fact remains that AWS and Azure continue to dominate the market, and OpenStack is going nowhere.
history repeating again? by Kishin · 2015-12-25 09:31 · Score: 1

The old VAXen that DEC sold came with a measure called VAX Units of Performance (VUP's). You bought stuff with a certain amount of VUP's rather than raw performance specs. Now Amazon is selling some Amazon-specific measure of performance that benefits them more than us. More signs the cloud is just re-inventing the mainframes and minicomputers of old. Cheaper, faster, more flexible, and so on for sure. Same concepts, though, plus what it seems is the same bullshit. :)
Deliberately obscure? by spud1955 · 2015-12-25 14:36 · Score: 1

Maybe its all meant to be like mobile phone plans. So complicated that no one can workout which better or cheaper