40 posts • joined 1 Jun 2007
Well, Tony? We're waiting for the correction article now that you've been properly embarrassed by the facts.
I expect Tony Smith will print a retraction when he is proven wrong next week.
Criteria for invalidation of his assertions:
1) 32-bit apps run faster on the A7 than they do on A6
2) 64-bit apps run faster than 32-bit apps on A7
3) 64-bit apps on A7 can reach 2x performance of 32-bit apps on A6
The article's understanding of CPU architectures is shallow and its conclusions are faulty. Many others have pointed out some of the reasons why. I imagine AnandTech will shortly embarrass the author by performing a real analysis when they have hardware in hand. Once that is accomplished, surely The Register will print a full article, not an update comment tacked on to this one, detailing the extent of their error. How about it, Tony?
In a shocking turn of events,
French government board recommends taxing products made by non-French companies - quelle surprise! Protectionism is alive and well in Socialist-paradise, where Asian and American companies subsidize the continued existence of French culture. A million young people voted for Hollande. Their punishment will be to pay more for what little luxuries they can afford with their lack of jobs.
time to go 128bit
If those drives average 1TB each, that means the number of bytes that Seagate ships is too large to be addressable by a 64-bit pointer. Of course disks are addressed by blocks, which are now mostly 4KB, but mmap() is out of the question if you were to virtualize all of this year's disks.
Re: Nice article
The reading here is to be done between the lines, unless you have access to internal Nvidia documents. The facts we have are thus:
1) Nvidia touts new architecture with a fast quad core and an additional low power core for live migration (effectively the same big.LITTLE architecture ARM is proposing). It will require changes to the Linux (and presumably Windows) kernels to support asymmetric scheduling and core transition for their android customers.
2) No Linux scheduler changes ever arrive from Nvidia or anyone else to support this.
3) Nvidia's repurposes the extra core as a touch detection device, a job for which it is far too powerful.
The conclusions here are either that the chip is actually broken and cannot be made to work as intended (very likely given that it is Nvidia), or that they failed to create software that does this to any benefit. My money is on both, but you'll notice that they are not trying this again for their future chips, which means that it's not just a hardware failure.
"Both the A57 and A53 employ the 64-bit extensions to the ARMv7 architecture that were announced when the 64-bit ARMv8 architecture was unveiled last October at that year's ARM TechCon."
ARMv8 is not an extension to ARMv7 by any means. It is in fact an entirely different instruction set architecture (ISA) that is not at all compatible with the previous generation. This is very different from, say, x86_64 or PPC64 that are extensions to their predecessor ISAs.
"The A57 is essentially a Cortex-A15 with the addition of the ability to handle 64-bit processing,"
Yes, except not really. Some parts are similar, but looking at the new ISA, ARMv8 is actually quite a bit more RISC-y than ARMv7-A, doesn't use predicated instructions, omits complicated multiple load/stores, and is in general better designed for parallelization and reordering by a modern CPU. Its pipeline would be quite a bit different than an A15 with its relatively complex ARM and variable length Thumb instructions. It's far more likely that the A57 is a 64-bit processor with backwards support for V7-A.
"Expect A7/A15 mashups using big.LITTLE to appear in products next year, they say."
It's nice to see that hardware designers still don't understand how software actually works. While there may in fact be chips that sport fully coherent A15+A7 complexes, the number of products that actually use them as ARM imagines will be countable on zero hands. Ask Nvidia about the Tegra 3 and its companion core for details on why this doesn't work in the real world.
Open source always triumphs
Wait, I thought Google were the good guys, do no evil and all that. My understanding was that the definition of open is "mkdir android ; cd android ; repo init -u git://android.git.kernel.org/platform/manifest.git ; repo sync ; make". There must be some aspect I'm missing here.
The US spends almost no money on Canadian border security. Sure, they like to inconvenience travelers between the two countries, but there are two thousand miles where there is nary a fence separating the two, and a person can simply walk right through. The border is even larger and more porous if you consider the Alaskan border.
The reason for this is simple: the standard of living in Canada is about equal to that of the United States. There is no great rush of Canadians attempting to flee poverty and oppression by entering into the US, just as there is no rush of Frenchmen flooding into Germany due to porous borders. Mexico, on the other hand, is a problem for the US because of the standard of living discrepancy just as North Africa is a problem for the EU.
Europe will seal its borders, whether external or internal, or its standard of living will sink so low that the osmotic pressure for immigration disappears. I'm fairly sure that you will not be pleased with the latter outcome.
math is hard
Apparently you don't understand the difference between selling software and selling hardware. See, with software, you make one GM disc and then replicate it a billion times. The only cost to you after non-recurring engineering is the price of a DVD or a download. When you sell hardware, you have to actually make every piece and pay for components on each one beyond the non-recurring engineering work you did developing the hardware and software that runs on it. By your numbers, Apple is making 24.3% selling devices of which a significant amount of money needs to be paid out for each sale, whereas Microsoft is making 31.8% selling plastic discs that cost nothing.
If you need it spelled out: Microsoft spends FAR FAR FAR more on R&D than Apple does for significantly smaller returns.
the users get hurt
Google and Microsoft can have a slap fight in private or public, but releasing this exploit before a patch is available is putting users at risk. The guy, likely a typical Google aspie, got his panties in a bunch because Microsoft wasn't taking him seriously, so he decided that he would show them by putting this out there and proving how right he was. I doubt he had the support of his superiors on this one.
I'm not one to defend MS's lousy security record, but the point of the disclosure protocol is to protect the millions of people out there who use this stuff. Even those that are proactive about security may be bitten by a zero-day exploit with no patch available.
Also, why do we not have an icon for Eric Schmidt with horns? Surely he deserves that much.
Shaft the UK buyer?
"Sounds like the usual shaft the UK buyer business model"
In California, the sales tax is 9%. At that rate, a $499 iPad costs $543 dollars. If you're paying £429 (VAT included), you're getting it cheaper than we are ($536 according to el reg).
the crux of the issue
But the article avoids the main point of contention in the SQL world (no, not Mysql versus Postgres or Oracle vs. everything else); whether it is pronounced "sequel" or "S Q L".
virus company scams
"Intego acknowledges there is no iPad malware to defend against as yet"
Doesn't Intego VirusBarrier X6 then qualify as the first piece of iPad malware?
Adobe can't even get a 64-bit version of the software out, their Linux port is barely functional, their Mac port is woefully behind the Windows version in performance, and security holes abound. This has been true for years and is not getting any better.
My best guess is that the Flash codebase is such an utter mess that Adobe can't do a thing with it. They're just scrambling to patch the leaks and keep it running on current platforms. They can't admit this, of course, but given that they're unable to make headway on any of the issues plaguing their most popular product, they don't have to. If this is the situation, you can expect it never to improve until a complete rewrite of the software takes place. If Adobe is currently doing this, they should be up front about it and give us a deadline for the fix. If they are not then the future of Flash is behind it.
32 bit code strikes again
Looks like our friends in the storage industry are running up against the limits of their 32 bit controllers (2^32 = 4.3 billion). Unfortunately it seems like we're in for a rough ride since they will need to bump up the sector size again soon. This is not unlike the way FAT file systems handled this problem when they exceeded 2 GB disks, and again we will have large inefficiencies in storing small files (unless your FS packs them like Reiser 4). That 10% you gain will be chewed up rather quickly, especially when they step up to 32 or 64KB sectors.
if you have to ask, you can't afford it
ARM's RVCT compiler produces far better code than GCC. It has much better dead code stripping and inlining, it can handle long jumps better (no trampoline sections), is much smarter about Thumb interworking, it can actually use the VFP, and generally generates code that is both faster and smaller. Its only real flaw is the license manager that slows down your compiles (do you hear me ARM? Per-process license is a stupid, stupid idea).
It is very expensive.
Their RVDS debugger, on the other hand, is a steaming pile of shit.
"in a world dependent on innovation, it means the ability to design technology that is as broad and creative as the people it serves." Were I writing about women in technology, I would refrain from using the word "broad" to describe "the people"
buy it now
Will her platform include mention of her contributions to Proposition 8, ending gay marriage in California (and screwing many of her gay employees), or will she focus on her nonsensical 2.6 billion dollar purchase of Skype? eBay is run like a forced labor camp and its success is largely due to it being first to market. Frankly, I'd rather have another unqualified actor.
The most likely reason for blank user agents: the Googlers have decided that they want to encourage websites to be standards compliant instead of detecting the browser type and building a page for that one. This sounds pretty consistent with a company that has just released a minority browser platform.
Does Russia REALLY need a non-ascii TLD, and right away? Isn't this something that can be handled in the application layer? How about setting up DNS resolvers in Russia to resolve *.PO (the Cyrillic first letters of Russia) to .po.ru? Seems a lot simpler to me. This will turn into a phishing minefield as ascii-like Cyrillic letters will be used. Don't get me started on the issues of filtering Chinese characters. I have no idea how to input one, or even how to look one up.
For what it's worth, my country doesn't use letters that even look like Latin, and we do okay with our ascii TLD.
"focused on customer care and explaining electronics to the customer rather than just supplying".
This is the funniest thing I've heard all week. I've never been to a CPW, but I've been in plenty of Best Buy locations. Their employees are warm sacks of skin in a blue shirt. They couldn't tell you the difference between DRAM and a hole in the ground, much less explain why you should buy one TV over another. In the fine tradition of used car salesmen, they will invent anything and everything in response to your questions, and then will try to push their worthless extended warranty on you. They are best handled with a cattle prod, keeping them at distance while you shop.
Cult of the Non-Dead Cow
Where is the X-prize to develop non-animal derived insulin so that the head of PETA doesn't have to harm animals in order to have the insulin she requires to continue living?
I'll get my coat, we're going for steaks.
Halfassed, as always
Adobe didn't have time for Linux until they saw that Microsoft was out to eat their lunch. Suddenly, the love spreads.
And by supporting Linux, of course they mean i386 Linux only. Have no dreams of running anything they release on your ARM Linux handheld. Hell, they won't even recompile for AMD64 (still waiting for flash support there...).
Simply corrupts Windows. Not ideal, sure, but the hardware is intact and you can reinstall. Hardly a brick.
"Unfortunately there's no word yet on the read/write speeds on the disks, which is rather important."
Unfortunately there's no price info, either.
Lightweight, so what?
While this embedded hypervisor is nothing to be scoffed at, why are we so happy about VMWare having stripped features off of their flagship product? I can buy 2GB of flash (SD) at retail for 25 USD (so figure out how much Dell pays). Why didn't Dell go all the way and include the full version of ESX server plus management console and so on instead giving customers the emaciated 32MB version? Surely if my navigation device and my ipod and my phone can all have 2 or more gigs of flash memory, my Dell servers can afford it.
Again, I'm not knocking the product. It's great. But let's not forget that in this context, "embedded" does not have to mean "reduced".
"WGA's goal is not to punish the people who purchase these programs; they, of all people are the most victimized,"
What? How am I, a paid-for Windows user, victimized when someone else pirates Windows? In what bizarro world does that make sense? I am victimized when THE SOFTWARE I PAID FOR intentionally disables functionality THAT I PAID FOR because Microsoft's unwanted agent on my machine, WGA, defaults to the assumption that I am a criminal.
"The goal is to give these people a tool to let them know they have been victimized and can do something about it."
I know I've been victimized when I have to drive over to my parents' house and reactivate their computer. And I can do something about it: their next computer will be a Mac*. Thanks for informing me, Microsoft.
*and I hate Macs, so don't paint me as a fan-boy.
No Architecture Type
I notice that their website doesn't specify the instruction set. Probably an unlicensed MIPS derivative, given the number of cores they were able to squeeze on one die. In particular, they neglect to mention if it has a floating-point processor (I'm assuming it does not). Otherwise, it's very interesting since it already runs Linux and has on board PCI-e as well as GbE controllers.
Does not go far enough
The burden of creating unbreakable machines should NOT be on the State of California. These vendors should be competing with each other by making sure each of their products is the most secure and actively encouraging hacking attacks and comparisons by the state's auditors. The fact that they are generously willing to do what the state demands of them to meet the minimum security requirements, instead of competing to meet the maximum security requirements, leaves me (a California voter) fearing the worst. Let's decertify them all, and not let them come back until they return with a machine that DOES NOT get hacked. They get to submit one machine every two years. If that machine is hacked, well, better luck next time.
While I'm here, who in the hell thought putting wifi on a voting machine was a good idea? These guys should be laughed out of competition. I think I'm going to start voting by mail.
Ashlee Vance on the sauce?
Seriously, what in the hell was that detour about animal abuse? You admittedly have nothing bad to say about a person and you feel compelled to make something up. This seems like a personal problem to me. I encourage you to air these issues out with a trained professional, not on these pages.
Amusingly enough, the stock photos on the MS open source page linked above are the same generic developer photos that appear in the Visual Studio 2005 installer. Who know the intended VS users are also open source enthusiasts?
How about removing the numerous fatal bugs out of Visual Studio 2005 before we move on to VS2008? Every release since VS6 gets worse. I'm not even asking for it to work well (like detecting changes to header files when rebuilding), I'd just like it to crash not nearly as often.
$9500 is not that big of a deal. Sure, it'll set you back some. But given the cost, plus power and cooling and rack space you'd need for equivalent performance out of general purpose processors (for a very specialized application), it may well turn out to be a bargain. Obviously, this throws these out of contention to be accelerators for your web server. Odds are that you aren't doing something that mathematically intensive to need one of these anyway. If you're simulating nuclear explosions, however (or, y'know, listening in on citizen's encrypted transmissions), the performance you can get out of the Xylinx part with good custom configuration can blow away a rack of GPPs.
I'm no foaming-at-the-mouth rabid Apple fanboy, and in fact usually put function over form. But why, oh why, could they not have made it look like any other plain, boring, normal phone? That thing is just hideous. And let me tell you, in mixed company, "it runs Linux" will hardly work as an excuse to justify that.
"The 1TB drive comes in two flavors; business and pleasure — identical cousins, yet different as night and day."
Subtle. A little Patty Duke reference for those of us who watched too much TV in our younger days. Funny.
"Indeed, if Seagate's new ES.2 disk drive for enterprise were to be played by popular 1960s sitcom actor Patty Duke, one would notice it adores a minuet, The Ballet Russes, and crepe suzette. Meanwhile the 7200.11 for PCs would apt to love to rock and roll. A hot dog makes it lose control."
Not so subtle. Now you're beating the reader over the head with the joke. Not funny.
Brevity is the soul of wit.
The quote in question (which everyone seems to love) is: "The primary duty of an exception handler is to get the error out of the lap of the programmer and into the surprised face of the user."
I am guilty of copying this into our internal development wiki and our Bugzilla quips, so I'm as unoriginal as the next guy. But, Miss Stob, here's the reasoning. First, it is a standalone comment that is simple, yet funny. True, you've uttered many a brilliant observation, but most of them require context to appreciate. This one is a standalone (context-free, for your XSL fans). Second, it is both simple and contains an implied vulgarity, perfect for simple and vulgar folk like myself. Third, and this one may have escaped you, it does not require that I engage my English to Americanish translation engine (an enterprising post-doc may want to compare and contrast this with hardware CISC to RISC translation). I find your writing, with all of its charming colloquialisms and contortions of the language not fathomed by us continental types both challenging and fascinating to read. However, it seems that most of my coworkers find your tracts, well, intractable. This unfortunately rules out many of your more clever statements for mass market purposes.
Thus, like a singer that has released one catchy pop song after a repertoire of deep and meaningful music, you've captured the hearts and minds of the hoi polloi with what in your mind may at best be mediocre. There is no fighting it. I suggest opening up a Cafe Press store with your new signature phrase, perhaps selling bumper stickers.
I may be missing an important aspect here, and that is key management. That is, the encryption takes place on the drive, and the drive holds the key (I'm guessing) until told to forget it. So where does theft protection come in? If someone steals my laptop, can they read the drive unless they purposefully wipe it out? That's absurd.
Either this press release, er, hardware review is missing some important information (such as a password required at startup for the drive to correctly read its contents or other key management scheme), or this feature is, well, useless.
It's About Time
As an outsider to the Solaris community that's been interested in trying it out, let me say that it's about damn time this happens. If you go to the OpenSolaris web site, all bright eyed and eager to download a new operating system, you will walk away in bitter disappointment. Sure, it says the word "open" in two dozen languages on the web page, but when you go hunting for an installer disk to download, suddenly you are cast into a maze. Nevada builds? What the hell is Nevada? Oh, it's what they're calling the OpenSolaris code base. You'll need to download these components and build them. Well, how do I install it? Oh, you can't do that, you need to have a Solaris machine up already to build on. But you can get started if you go to Sun's site and download their Solaris Express Enterprise Pro Champion Edition (after dutifully registering), and then enjoy that pleasant install experience. And when that's done, you still have the work ahead of you of getting ON (what the hell is that? Oh, OS and Network. Sorry, I don't work at Sun) built and updated. Did I miss anything? We haven't gotten to packages to make the system usable yet.
I just want an ISO that says OpenSolaris and installs THE OpenSolaris system. In a usable state. Then I'll be able to test my apps against it and claim they work. "Well, it runs on Belenix" doesn't quite feel the same, does it?
Ashlee Vance, always the senseless Sun critic (sometimes rightfully so, but more often not) misses the point, as usual, in eagerness to bash Sun. I look forward to downloading this disk when it is available.
- Fee fie Firefox: Mozilla's lawyers probe Dell over browser install charge
- 20 Freescale staff on vanished Malaysia Airlines flight MH370
- Neil Young touts MP3 player that's no Piece of Crap
- Review Distro diaspora: Four flavours of Ubuntu unpacked
- Sysadmins and devs: Do these job descriptions make any sense?