Slashdot Mirror
Notebook Makers Moving to 4 GB Memory As Standard
akintayo writes "Digitimes reports that first-tier notebook manufacturers are increasing the standard installed memory from the current 1 GB to 4GB. They claim the move is an attempt to shore up the costs of DRAM chips, which are currently depressed because of a glut in market. The glut is supposedly due to increased manufacturing capacity and the slow adoption of Microsoft's Vista operating system. The proposed move is especially interesting, given that 32-bit Vista and XP cannot access 4 GB of memory. They have a practical 3.1 — 3.3 GB limit. With Vista SP1 it seems that Microsoft has decided to fix the problem by reporting the installed memory rather than the available memory."
Because OS X Leopard CAN access 4GB of RAM. Let's see if Apple also joins the fray. Then again, 4GB is way too much, because after all 640KB should be enough for everyone.
for obvious reasons :D
:D
Nice "fix" though, then people can keep adding memory and think it helps
What an ingenious solution! No wonder Microsoft is the leading innovator in IT! That's SO innovative it makes me sick. It's hard enough trying to explain to customers why they shouldn't waste their money on 4GB of memory and a 1GB video card only to lose a quarter of it in real life, now the OS is trying actively trying to make me look like a liar too.
'I want one of those to run pine and vi!'
OEMs releasing computers with 4GB of memory standard? Finally, we have machines that'll be able to run emacs...
Stop with the kludges and force the developers along. 32 bit came with the 386 era and lasted a good while - a very good run indeed. 64 bit would last beyond our lifetimes anyway, I doubt we will even come close to the limits of addressable memory there (hopefully this isn't the new 640k comment) -- so there is no point in stalling it indefinitely.
It's now official, new machines have more RAM than my first computers had hard drive. I'm not sure how to use such power, though you can bet it will be with an open source OS.
This means that my next machine will likely be able to run the entire operating system from RAM if I want to use it that way, even with a GUI and applications I would have qualified as memory intensive just a couple years ago.
Back in my day when we chiseled our bits into stone and sent them by mule train from village to village...
Since 4GB is Vista's supposed "sweet spot", this seems to make sense. But the addition of RAM (which I could not find anywhere in the article that said it would be standard) is partly due to the price of 2GB modules "likely" dropping. If I was a memory manufacturer, and knew the big names in notebooks were going to quadruple memory, you better believe I'm raising prices. If 4GB ends up being standard, it's more like "consumer pays for" standard, than "steering wheel in a car" standard. This seems more "likely" to me.
Its pathetic how poorly Microsoft has moved towards 64 Bit vs. Other OS's out there...
Solaris Way Back in the 90's with Solaris 7 I think... Had 64 bit support with perfect 32 bit
support of backwards compatability.
OS X goes a step further one OS Package and support for PowerPC, 32 Bit and 64 Bit and Intel 32 Bit
and 64 Bit. And appliactions seem to work for all of theme for the most parts (with the ovious
exceptions of apps that require the advanced features of the newer Chips.
Reporting that you have 4 Gigs installed is not a real feature it just makes it easier for the
hardware companies to scam people saying here buy this with 4 Gigs of Ram and the OS says there is 4
GIgss of Ram while it only supports 3. I would be Pissed If I knew I couldn't access all my RAM.
Say I had VMWare on my Laptop and I allocated a VM with 2 Gigs and an other with 1.5 Gig and ran both
figuring that I had 4 Gigs of Ram available. I would be annoyed that I couldn't run both of my VMs
and Not knowing seeing that it supports 4 gigs of RAM I would want to know who is taking up 512 Megs
of Active Ram. I could blame Windows for being more of a memory hog. I could blame VMWare for sucking
up all the extra memory to run. But the fault is the Hardware Manufacutre put more ram then the
PC with the preinstalled Software can handle to make a few bucks and Microsoft just plays in their
hand making everything look hunky dory.
If it says you have 4 Gigs install It should also say there are 3.3 Gigs that can be access
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
to fix the problem by reporting the installed memory rather than the available memory
:P
I wouldn't have expected any other `solution' from MS
I am putting myself to the fullest possible use, which is all I can think that any conscious entity can ever hope to do.
In another few years, I might be able to run Eclipse.
From the summary: "They claim the move is an attempt to shore up the costs of DRAM chips, which are currently depressed because of a glut in market."
The article says: "While first-tier notebook vendors such as Dell, Hewlett-Packard (HP) and Toshiba are planning to roll out 4GB notebooks starting from the first quarter of 2008, the move is expected to give a boost to the DRAM market, according to memory module makers."
The article does not say that this is a deliberate attempt to increase DRAM price. And if it was, wouldn't it be illegal?
It sounds like they should be making wiis and not memory. Solve a few problems at once
Check out my sysadmin blog!
My Laptop uses the ram as video memory, if they use 256/512 MB as video ram it is not that much more than the maximum allowed by 32bit operating systems.
32-bit Linux has similar limitations, too. OS needs to reserve *some* addresses to access other hardware such as CPU registers, PCI cards, etc. Since 32-bit CPUs can only count up to 2^32, it cannot address all the locations in RAM. This is definitely not just Microsoft.... :-)
Or at least, supervision by people who know how computers work? 4GB is perfectly sensible for a 32-bit x86; the virtual address space is only 4GB, but the physical address spaces is larger (at least 36 bits on all popular processors). Yes, that means it's awkward to use more than 4GB in a single application, but so what? Using more than 4GB across the system is perfectly transparent.
Also, what's with slamming Microsoft over the "slow" transition to 64-bit? 64-bit XP has been out for, like, three years now. It runs 32-bit applications, because the x64 architecture makes it so ridiculously easy you'd have to intentionally break it. 64-bit Linux does the same, because it takes, like, a line of code to do so. If software makers aren't producing 32-bit apps, it's probably because their customers haven't demanded they do so yet; and the customers probably haven't demanded it because it's unusual for a single application to need 4GB of RAM. Finally, those applications that can frequently use gigondo amounts of RAM in a single virtual address space (e.g., Oracle) for the most part had 64-bit binaries available right out of the gate.
Laptops, which have been a nice fat cash cow to OEMs, are steadily becoming cheaper and cheaper. Nowadays, we have retailers selling laptops for about 500 euros with specs that a couple of year ago were selling for a lot more than 1500 euros. A while back the average laptop price tag was about 1200 euros and nowadays it went down to 800 euros. That's a heck of a revenue cut.
So the OEMs are forced to add another fancy selling point, like upgrading the specs once again, in order to keep making the big bucks. They don't give a damn if it brings any added value to the product or if it even functions properly. What matters is some fancy little side remark on the laptop's brochure that makes their fancy little product be picked by the vast hordes of consuming sheep. Who cares if it makes sense or if it's even usable. What's important is that them flock falls for that "OMG! IT'S N+1!!!" and promptly spend their cash, specially for the "it's bigger than my neighbour's" bragging rights.
This sort of thing isn't exactly new. In fact, it's the repeat of another similar marketing push, which was the "32-to-64bit" campaign. The fancy stickers advertising the new and improved 64bit 'puters for the "OMG IT'S TWICE THE BITS!!" effect were all over the place, which earned quite a few hardware sales. Yet, the fact is that the brand new 64-bit 'puter could only run on the 32-bit legacy mode, as they were shipped with a 32-bit operating system and the OEMs shipped hardware without ever thinking on releasing 64-bit drivers or even releasing the hardware specs.
So those OEMs will, once again, sell hardware that will not be usable by the user, at least as advertised. It doesn't matter to them. The only thing that matters is the sales revenue, specially in this day and age where we are starting to see sub-300 euro hardware. And screw the consumer.
Slashdot, fix your code or at least hire someone who is competent at it to do it for you.
If you also use a resource-laconic OS, it will mean more ram for your software.
Bloatware will always exist and lean programming will also continue. A word processor can be made to run slow on future super computers or really fast on 90s era computers and its all depended on the programming philosophy and skills. People who care will always complain or work with every little byte of ram, even if you have tenths of GBs to spare.
I've got a MacBook (not the x3100 model), 2.16Ghz BlackBook, and the max in theory it can hold is 4gigs but according to apple the max it can support is 3gb, and according to the forums, it can access 3.3.
:(
What happened to 4gb? a chipset limitation or Apple castrating their hardware like usual?
ps. this is an honest question, I can't seem to get a straight answer from anyone
In all fairness 512meg is a reasonable about for XP, 1gig however is better. I don't see your average joe 6-pack user benefiting from more than 2gigs.
:)
After all the auto-updating software for their printer, mouse, keyboard, webcam, etc.; all the spyware, adware, trojans; and all the extra applications like AIM, anti-virus, anti-malware, non-driver device software (syncing, calling home, etc.), and media software playing music in the background, I can see joe 6-pack user making use of more than 2 gigs if he actually wants to do something with his computer.
Nostalgia for XP will fade. As existing PCs fail, they will be replaced by Vista boxes.
Vista is an interim solution to a problem that existed only in the minds of Microsoft's stockholders, which has caused and will continue to cause problems for actual customers. I have the feeling that by the time Vista is out in any significant quantity, Microsoft will have obsolesced Vista and moved us onto something else.
The higher the technology, the sharper that two-edged sword.
How many Microsoft engineers does it take to change a light bulb?
None. Microsoft just redefines darkness as the standard.
Well, there's spam egg sausage and spam, that's not got much spam in it.
Windows 32-bit operating systems really do not play well with more than 3Gb of memory. Check out what Microsoft says about it.
The address space you'd normally use to address memory beyond 3.5gigs (or thereabouts) is reassigned for talking to other stuff. It's a simple 32bit limitation - can't only be talking to the RAM after all. As others have pointed out throughout this thread, the problem does not exist in 64bit architectures.
5 years ago, nobody would have thought that we'd run into this problem at all. Remember those times? Everybody and their mum was just about getting ready to jump onto the 64Bit bandwagon with AMD charging in front. And then, while nobody (especially not AMD) was paying attention, we kinda veered off-course into a multi-core world instead and all of a sudden, people stopped caring about 64bit. After all, you had a larger net performance gain from upgrading to 2 32Bit cores than to one 64Bit one. And now, we're finally running out of address space.
The trouble is that in contemporary chipsets in 32-bit mode the upper 1G or so of physical memory overlaps with the address space for the PCI bus.
Error: password can't contain reverse spelling of ancient Chinese emperor
it even took years to fill up
Our first PC had a 30MB HD capacity. And that's the first PC, not the first computer.
I'm only 29!
I'm guessing this is Moore's law in action, in that the memory chips have scaled. Notebook motherboard space hasn't increased, so they probably aren't increasing the number of chips, but going to the next gen of memory chips.
Remember that the 3.2-3.3 GB limit is not imposed solely on the RAM itself. The computer as a whole only has a maximum number of addressable "points." I believe that 32 bit Windows or Linux allows for a maximum number of addresses around 4 GB, so you'd think that you could use 4 GB of RAM.
Wait a moment and think it out.
Estimate that components such as your processor caches, motherboard I/O destinations, Network cards, CD/DVD drive will take up about 1/2 GB of the theoretical 4 GB. These MUST have addresses or they cannot function.
Now add in all of your vRAM (the amount of ram on your video card), that ram will also need a set of addresses. We'll estimate 256 MB of vRAM.
So now you've taken your theoretical 4 GB of ram space, subtracted 512 MB for essential system components needing addresses, subtracted 256 MB of vRAM on your video card needing addresses. So, total, you've just taken away 768 MB of your theoretical RAM limit. 4 GB (Theoretical limit) - 768 MB (used addresses by components and video card memory) = 3.25 GB of RAM. Systems with 512MBs of vRam have a 3GB limit for memory.
Now consider the slap in the face SLI 8800 GTX's would be to system addressing. They take up 768 MB of vRAM each. So that is 1536 MB of vRAM total. Now you are probably down to something like 2GB of RAM addresses available for the system.
Heh. So the point is, the world NEEDS to get it's butt over to 64-bit sometime soon. Gamers are going to start to feel the burn soon when suddenly they have no more RAM to play with while SLIing.
This applies to both Windows and Linux. 64-Bit doesn't have this limitation. The only ones implying it is a Windows problem is those like Twitter the Troll and Communist Zonk.
I think next year I'll finally upgrade my system. I have a TI-99/4A. I bought it a few years back when they went on sale at Sears for $100. It has 256 bytes of RAM, but will store programs in the 16 KB of video RAM.
Fifteen colors, 256 x 192 resolution, 3 MHz processor, plenty for coding and reading Slashdot. But Firefox takes forever to load from the cassette tape.
So for ten times the cost (ouch) I could get 250,000 times the RAM, 1,000 times faster clock speed, and at least 10,000,000 times the storage.
I hope my games still run: Parsec, Congo Bongo, and my favorite Alpiner.
Well, I guess now I have to find exactly which is the job I do that can be done more efficiently using either the Mayan Calendar or the French Revolutionary Calendar...
Yes it is.
The Linux kernel devs solved this back in 2004.
"I've got more toys than Teruhisa Kitahara."
Why would notebook makers want to keep memory prices propped up, when those prices increase their costs, and the price of their notebooks they have to sell?
The much more obvious reason is that Vista needs more RAM, as does XP and everything else that inevitably bloats (including Linux).
And even though 32b Vista can't use all the 4GB of RAM, people will want it anyway, because they won't know/understand that the extra RAM isn't helping. And with all the other problems Vista has, the old solution to just throw more RAM at a buggy platform will still be popular.
--
make install -not war
How many Linux engineers does it take to change a light bulb?
None. Linux lightbulbs last forever. They don't need to be changed.
Since RAM is so cheap, why aren't there more "external RAM" systems like Gigabyte's I-RAM? That seems to be the only product offering RAM structured as a drive.
And it's got limitations. What about more products, that draw power from USB instead of PCI (that's all the I-RAM uses the PCI for anyway)? Maybe a 6Gbps SATA connection, to compete with PC-2700 RAM's 21Gbps access rate. If the memory manager could access several PCI-e SATA cards, how about 8 SATA connections to a single bank of RAM, for 24-48Gbps rates, usable as actual RAM (not a RAM drive)? And how about a few 10Gb ethernet ports, for Network Attached RAM, shared memory among distributed apps? Sure, the latency and coherence are problems for parallelizing these serial buses, but with devices to work on, the SW to control and harness them could get underway.
There's enough cheap RAM now. How about some 400GB RAMdrives, or just 400GB RAM for preloading at OS boot every app you'll launch in a desktop session? How about launching separate fresh images in RAM as backup for apps with memory leaks, the OS switching among them to keep them fresh (and killing/relaunching to keep the pool fresh)? There are all kinds of RAM-limited scenarios that lots of RAM could solve. We can afford it now. Let's have it. The increased demand will keep the prices higher, but for a good reason instead of these inaccessible notebooks.
--
make install -not war
And what about drivers? IIRC 64 bit linux supports far more hardware than windows does.
The summary is a bit misleading. Notebook manufacturers are in a gruesome race for ODM/OEM contracts. The only reason they have for spec'ing more RAM is because the customer (DELL, HP, etc) are paying for it and the price for the larger RAM spec is lower.
http://www.maxineudall.com/2010/02/should-economists-be-sued-for-malpractice.html
It appears you failed to notice that the architectures of AMD's and Intel's multi-core processors are both x86-64. That means that we are upgrading to two 64-bit cores.
Slashdot, fix your code or at least hire someone who is competent at it to do it for you.
"It does not do to leave a live dragon out of your calculations, if you live near him." - Tolkien
Booting Solaris Systems to Either the 64-Bit Kernel or the 32-Bit Kernel
http://developers.sun.com/solaris/articles/64_bit_booting.html
This document provides answers to frequently asked questions about booting the Solaris Operating Environment (OE) to either the 64-bit kernel or the 32-bit kernel. Booting the 64-bit kernel has been available on Ultra workstations with UltraSPARC processors (Sun4U systems or higher) since the first release of the Solaris 7 OE.
Standard operating practice for Redmond is to rollout each new OS with a strict requirement that bumps up against the current hardware maximum generally available. Just because Vista can't address 4GB doesn't mean they can't check for it and require it anyway. Redmond and hardware companies have always had an incestuous relationship - a more than gentleman's agreement that hardware and software will bloat up hand in hand. Everyone knows that.
I'm not a market analyst but I have to wonder, I don't see why it would be illegal to increase the market value of your product by taking steps to encourage the demand? That's the goal of advertising and many other common market practices.
I work for the Department of Redundancy Department.
Surely you should say "There's very little in this world...". Otherwise all /.ers would have girlfriends.
Actually, the memory limits are due to buggy 32-bit PCI device drivers that can't have their address spaces reallocated to above the 4 GB barrier - so MS decided to limit the amount of RAM available to XP rather than deal with POS drivers.
Server 2003 and XP 64-bit don't have this limitation, though.
P.S. - I'll reccomend 64-bit OSes for the masses when Flash and Java Web Start work. Not until then. (I've been running 64-bit desktops for several years now - Fedora, CentOS, Debian, and some technical workstations running XP 64-bit and DB servers running Server 2003 64-bit.) You'd have to forklift upgrade a large portion of your stuff to work with 64-bit - you have to rewrite a fair bit of your driver core to handle the larger memory address size, something that hasn't had to have been done since Windows 95 came out!
P.P.S. - The 4 GB per-process virtual address space is not completely true in Windows - you can only access 2 GB of it, with the other 2 GB of address space reserved by the operating system. IIRC, you can change it to have 3 GB available to the process, but MS does not guarantee that all third-party DLLs will work properly.
Why can't I mod "-1 Idiot"?
Nothing new here. Ubuntu 7.10 32-bit -- the latest shipping version -- only supports 3.3 Gb or so. They haven't bothered to turn on BigMem in the kernel. I upgraded two desktops and one laptop last week, all to 4 Gb of RAM, and was seriously disappointed when the BIOS showed 4 Gb but Ubuntu did NOT.
Learning HOW to think is more important than learning WHAT to think.
Well, if your field of study was French Revolution or Mayan history, I bet you'd find it pretty useful.
Your ad could be here!
"it's bigger than my neighbour's"
It is bigger than your neighbour is? I wouldn't consider that a viable selling point, unless you live next door to some seriously small people.
Dealing with lawyers would be a lot less tedious if they all looked like Casey Novak.
Deleted
Yes, but they also released 32-bit Vista which was just stupid. Given that 95% of computers capable of running Vista in the first place are also capable of running it in 64-bit mode, they should have just made the leap and been done with it. The reason that there are still problems with 64-bit drivers is that writing 32-bit-only drivers is still a viable option.
Dewey, what part of this looks like authorities should be involved?
"The proposed move is especially interesting, given that 32-bit Vista and XP cannot access 4 GB of memory."
So they would opt-in for 64bit Vista OEM or XP Pro 64bit edition. Lack of software compatibility? It would take a week, no more for Adobe come with 64bit Flash in case a large manufacturer calls them and tells the situation. Unfortunate but true.
Yeah, that would be why I posted a second comment immediately after that one ;)
No, no, "Real people" don't even know that there's a difference based on context. They probably don't even know that "1GB" is a HUGE number.
Also, it isn't always "1000-based" for hard-disks, some are "1GB = 1,000,000 kB, 1kB = 1024 bytes". It's also only 1000-based when talking about the total size; filesizes, used/free space, partition size etc. are shown as 1024-based in all OSs I've used.
"Total disk size" is the only use of base-1000 measures of bytes. Disk being hard-disk, floppy-disk, cd, dvd or whatever else. Some people point out "connection speeds" (eg 1Mbps), but those are kilo/mega/giga "bits per second" not bytes.
I'm honestly surprised they haven't tried to sell disks measured in "bits" yet. Get your 2Tb hard-disk here!
32-bit Flash and Java work perfectly well on 64-bit Windows.
The problem is game DRM that uses 32-bit "device" drivers, which simply doesn't work.
Dell is not an OEM? Is not Dell selling laptops with Ubuntu on it? Last I checked Ubuntu ran Linux...
For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
Yes. The hardware has been here, and we never really got derailed from the 64-bit track.
Now we are waiting for "OS" vendors and peripheral companies to get off their asses and move to 64-bit, like we would have been doing anyways, had the multi-core 'revolution' never happened.
I quote "OS" because of the fact that all major operating systems EXCEPT Microsoft's shit have been ready and waiting for the 64-bit changeover for a while now. Yes, I rather strongly dislike Microsoft, it's activities, and it's products. I will not hide it.
For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
Also note that there's another config option that allows one to change that 3G/1G split for NOHIGHMEM mode, if desired. It's normally hidden, but available if one activates EXPERIMENTAL and I believe EMBEDDED.
That 3-level paging above 4 gig is a bit of a performance hit, as the kernel shifts its 4 gig window around in that 64 gig frame, tho if one runs the sort of apps that actually use that sort of memory, it's less of a performance hit than going to swap would be. Still, going 64-bit Linux isn't such a big deal any more, if your CPU supports it, and it's MUCH more efficient since multiple terabytes of memory can be directly accessed.
There's another factor at play here as well, however, and this applies to ALL OSs on 32-bit x86 and most or all on 64-bit x86 as well. It's a PCI hardware issue more than a software issue. Many old PCI devices were designed for 32-bit only operation, and their hardware can't address memory above the 32-bit 4 gig memory barrier. When memory was running less than a gig, this didn't matter much, and it became customary to reserve the virtual space at the top of the 32-bit address pool, 3.5-4 GB, for PCI device i/o access. As real memory expanded into that area, it runs into the reserved area, and the real memory behind it can't ordinarily be directly accessed.
Folks who've been around for awhile will likely recall a similar issue back at the 1 MB barrier, and how it was resolved using a "memory hole". The same technique is used here. With a BIOS setup to do so, one can configure a "memory hole" at the 3.5-4GB location, and the BIOS will remap the affected memory up above the 4 GB barrier.
This explains the complaints about Apple and MS platforms also having 4 GB look like 3 or 3.5 GB. I'm not sure if their 32-bit kernels can cope with that remapping or not -- they won't be able to if they can't address more than the 4 gigs anyway, but even if they can, the BIOS must be configured to map the hole as well.
Meanwhile, while addressing memory above the 4 gig line shouldn't be a problem for 64-bit kernels, the BIOS must still be able to do the remapping as well -- and the kernel must understand and deal with the hole. 64-bit Linux has suitable config options to do so, but I've not the foggiest how binary platform systems shipping a single binary kernel for all users deals with this. Primarily binary Linux distributions generally ship a number of different kernels, including an enterprise
Duncan
"Every nonfree program has a lord, a master,
and if you use the program, he is your master."
R Stallman
You are right - we were all being told how wonderful Longhorn was going to be so expected it would support the extensions that came out with the Pentium Pro more than five years proir. The only people that veered of course were in Redmond with the Longhorn/Vista situation.
No it doesn't. It doesn't even apply to all versions of Microsoft Windows 32 bit. PAE was built into the Pentium Pro and later to extend the address space to 64GB and is supported in MS Server 2003, linux and most likely a host of others.
proprietary piece of hardware in the Intel processor. (Platform lock-in...love it!)
so get a 64 bit processor. There's nothing inherently wrong with supporting the latest and greatest features of a CPU. Torvalds originally wrote his system because Tanenbaum was reluctant to support the advanced feature set of the 386. The ports to PowerPC, Alpha, Sparc, etc came later.
Really? Linux works flawlessly in all my light sockets. I guess the system whereby engineers ask for specs from the people who want to sell them light sockets gets the job done.
-FL
I finally got a new dual core AMD 6k machine with 3.1 Gb ram and discovered that I really didn't care for eclipse. Vista runs pretty good on it and x64 Linux is wicked-fast
Apocalypse Cancelled, Sorry, No Ticket Refunds
The trouble with that observation is that in comtemporary chipsets, there is no 32-bit mode (they all support 36-40 bits of address in normal operation, ignoring any A20 legacy crap). Since all popular processors have supported 36 bits for quite a while (including remapping the 1G hidden by PCI to upper address ranges) this is a software issue. PAE legacy support hardware on the cpu allow user mode processes to happily play in a 4GiB virtual sandbox while the OS/kernel could have juggled these inside a 36-bit address space. Of course windows and linux elected to require some shared mapping so that was restricted to 2-3GiB tops. On top of that windows has it's own issues with more than 3GiB since it wanted to play with 32-bit code natively in the kernel (didn't have to do that, but of course everything is easier that way). Linux just wants you to get x86-64 everything to get around those type of kernel issues...
I can run FCP and Motion and Shake at the same time with very little slow down. The only applications I have where I notice any difference between my MacBook Pro and my QuadCore PowerMac G5 with 8GB of RAM is Lightwave and Blender 3D. And that has to do with the amount of RAM in the video card versus the system.
I give XP Pro 1.5GB when I launch it via Paraellels and it plays any games I want to play without any slow down issues. (Which the only game I play is Falcon 4.0 Allied Force and do some work on the Homeworld 2 Battlestar Mod)
I've been working around 64-bit chips since DEC Alphas in the late 1990's. The old adage back then was double the ram on the 64-bit processors. (I remember that a 1-GB Ram module in those days cost almost as much as my Quadcore G5).
The thing is, I'm not an average user. I work in video production. Mostly it's editing for a couple videographers doing boring things like weddings. (Hey it pays the bills). But every couple months someone needs help with some 3D animation for a corporate project, or a law firm wants to model an accident for a court case, or an indy film with a small FX budget wants some work done, etc. But time is money. The faster I can render something, the sooner I can move to the next project.
And I'll admit that 4GB (Especially the 8GB on the Quadcore) is even overkill for me 95% of the time. I edited most videos on an old 1Ghz Titanium Powerbook with 1GB of Ram with FCP 4.5HD just fine. Sure it lacked real time rendering, but if my other machines were busy rendering a final product, I was still being productive.
Does it improve workflow being able to keep FCP, motion, and Shake open at the same time? A little. I can take a strip, sent it to shake, animation some effects in Motion, import into Shake, render the composite and send the final back to FCP with a few mouse clicks. Sure, it's nice, when I do it. But that's maybe once every couple weeks. On a day to day basis, I have FCP open or DVD Studio with iTunes playing in the background and that's about it.
Now I don't know, maybe with some of the games out on the market, they require the gobs of extra Ram.
"The problem with socialism is eventually you run out of other people's money" - Thatcher.
Strange, Java web start worked just fine on 64bit Solaris.
Agreed for flash tho, flash is the biggest thorn in the side of 64bit linux users. Virtually everything else can be recompiled for 64bit these days.
http://spamdecoy.net - free throwaway anonymous email - avoid spam!
It seems just plain wrong that the virtual address space is smaller than the physical address space. Shouldn't it be the other way around?
Okay, now can someone explain why there is a 32-bit limitation when x86 operates off of selector:offset addressing? I haven't coded x86 in years, but my recollection is that there were 32-bit selector registers and 32-bit offset registers. That allows for 2^64 addressing. I thought the limitation was no more than 4GB per selector, not 4GB total. Is the 32-bit limitation a function of the GDT?
A standard XP install only has 2GB available to user mode program, Server versions have access to 3GB but only if the program was specifically write to take advantage of this mode. Vista probably has this special paging mode build into all versions but the programs still need to be compiled to take advantage of it. Only programs I know compiled for it are databases.
But windows can take advantage of all that extra memory for disk cache and won't need to touch the first two gigs, leaving it completely available to user programs. Also buggy drivers with memory leaks are only likely to bother you for a reboot every couple days.
AMD and Intel both went with AMD's x86_64 architecture, back in the Pentium 4 days, before any multi-core x86 parts existed. In fact, the first multi-core x86 parts were x86_64 through and through, not any less capable of running 64-bit applications than the single core.
I'm guessing you are confused because of the Intel Core Duo line that was prominent before Core 2 released. The Intel Core line was released after 64-bit P4s not because of inherent multi-core advantages, but because they realized how the NetBurst architecture was not working out, particularly in low TDP mandated environments like laptops (where they currently were using Pentium-M now, derived from Pentium-III). They released Core in an effort to have a more consistant offering, with lower TDP and better per-clock performance, forsaking 64-bit until Core 2 (except the Xeon family, which stuck with NetBurst until 64-bit was available via Core 2). It had nothing to do with multi-core and would have played out that exact same way if it was just single cores.
XML is like violence. If it doesn't solve the problem, use more.
As long as we are at it, here's the analysis of how the freebsd hackers handles the lb problem
so say if dell started shipping laptops with 4 gb of ram, would a stock ubuntu (32 bit) kernel use all 4gb? or would dell have to tweak something? gee, I wish I had 4gb!
Try running Gentoo on it. The software compiles in no time!
It is plain wrong. The reason this happened is because people the instruction set in use, 32-bit x86, was not designed to be used with this much memory. The solution is to move to a 64-bit architecture, like x86_64.
vi ~/.emacs # I'm probably going to Hell for this.
640KB should be enough for ten years.
That's what he said he though (or wrote) in the times of 64k, when he was proven wrong because only 7 years had passed and computers already needed more than 640k.
We are Turing O-Machines. The Oracle is out there.
What about Linux? Can the kernels shipped with current 32-bit Linux systems use the full 4 GBs?
Of course, I guess, on these machines you'd just use a 64-bit installation.
Please correct me if I got my facts wrong.
25 watts is good enough for anybody!
So yes, there's a lots of parts that people don't necessarily understand. Besides, facts would get in the way of a good flame fest.
My other car is a cons.
So.. Are you saying they should just take any old light bulb and just jam it in there, regardless of how well it fits? Or are you saying they should buy all their light bulbs from the socket supplier, thereby implicitly locking them into an indefinite future lighting contract?
Can you be Even More Awesome?!
One former co-worker's comment was the world would have been a much better place had IBM gone with the 68k running something like OS-9.
A Shadeless room is a brighter room.
Thanks twitter! I see you haven't stopped following me around.
You might want to hold your horses on the insults though, seeing as I made the second light bulb joke on this thread, but people without a sense of humour decided to mod it down.
"It does not do to leave a live dragon out of your calculations, if you live near him." - Tolkien
I agree with you spywhwere about regular pc consumers having to inevitably move to vista, but bittorrent/p2p users will 'get' XP on our new computers until drivers are not made for the new hardware. At that point, I might kill myself(j/k), or move to linux, or just run vista in the Classic visual mode.
There is no sanctuary. There is no sanctuary. SHUT UP! There is no shut up. There is no shut up.
This is a limit of a 32-bit system with PCI devices, not a Windows limit. You have a 4GB virtual address space. However there are PCI devices that use memory addresses to do what they need to do. That means that if you have 4GB of physical memory, some of it is going to be robbed for other things and you can't access all of it directly.
It isn't that they don't "play well" it is just that you don't get full benefit from the memory. Some of it, how much depends on the hardware you have, will be eaten up by PCI devices. It doesn't crash or anything.
The answer is quite simply to go to a 64-bit OS. Then there's no problem, it'll address as much memory as you can stick in your computer.
I'm typing this on an HP Compaq 8710w, running Vista Business x64.
Many of us have been using 64-bit versions of windows for years, in both XP pro and 2003 server form.
Do you mean PAE?
PAE has been supported in all NT versions of Windows since 2000.
This is how 32-bit versions of windows can use up to 64GB of RAM.
The only exception to this was XP SP2 where this was for all intents and purposes disabled.
This information is in the article you're replying to.
Your post is inaccurate. You only skimmed the KB article, and missed paragraph 4 under the "CAUSE" heading.
You're confusing per-process limits and system limits. Some versions of XP (and apparently Vista) cannot support more than 3.1GB of total system RAM. AT ALL. This is because PAE is not supported.
Linux can support the full 64GB of total system RAM. It does support PAE. Full 64 bit is preferred because using PAE means performance losses as the kernel maps pages around to access them, and has to use bounce buffers for 32 bit PCI devices since a 32 bit PCI bus doesn't support the 36 bit addressing.
Each process is limited to 2 or 3 (depending on your config options) GB because, oddly enough, the virtual address space is smaller than the physical on a system using PAE, and the kernel takes part of that space. However, if 32 processes each want 2-3GB or RAM actually in use (minus a few meg for the kernel), it will do that for you.
and 64 bit vista is there. The problem is however that the drivers are not that reliable as on 32 bit vista. I don't know why, but on 64 bit vista there is a whole new set of problems that were solved in XP long ago.
There is no technical reason for this. Just the fact that this technology is rather new compared to 32 bit OS.
-FL
The limit is in kernel address space and the inability of the kernel to address memory beyond 4GB. With video cards taking up 512M or more, plus the rest of the memory space taken up by other devices in the system, it's not uncommon for PC chipsets to simply map out the top half gig or so for I/O. The only way to access the memory that's been mapped out is by addressing it over the 4GB limit, or (if the chipset supports it) remapping it sequentially through some window. If the kernel doesn't support memory over 4GB and doesn't include a driver to do something useful with the extra 512M of RAM that it can't address directly, it might as well not be there.
1GB RAM for most everybody is PLENTY
2GB RAM for most gamers/heavy application users is PLENTY
because anything past that will see almost zero net gain in performance, end of freakin story.
See Raymond Chen's explanation:
http://blogs.msdn.com/oldnewthing/archive/2006/08/14/699521.aspx
I really don't care what comes standard with my notebook, but I do care very much about how much I can expand it to later on.
According to my recent enquiry to Dell they do not have a single notebook that will take more than 4 Gig RAM. Not even with 64bit vista can you order a notebook with more than 4 Gig. So now is the standard also going to be the maximum or will I finally be able to get enough RAM to reasonably run all the VM's I want at once?
Someone needs a nap. Or meds. Or maybe a spanking. In any case case my benchmark programs yield almost the same performance in terms of responsiveness between an old IBM PL30GL with a Pentium 3 550 & 384MB RAM running W2K and my Lenovo N3000-768 laptop running Vista. And Vista really doesn't bring THAT much more to the table. Unless you get wood over new 3D effect and a BRAND NEW menu bar.
Oh wait, maybe you do - BECAUSE YOU'RE A FUCKING ASSHOLE. So - take your nap, or your Paxil or have your top beat you, whatever the fuck it is you need. Thank you and have a nice day.
They may have used the hardware MS was already using (PAE), but they certainly didn't do what MS had done.
For a start, the Linux version works.
"I've got more toys than Teruhisa Kitahara."
There are actually two reasons this doesn't work.
First, segments have nothing to do with physical addresses. A logical address specifies a segment and a 32-bit offset; after a bunch of gymnastics, this gets converted to a 32-bit physical address (assuming PAE is not in use). So even though there are a lot more than 2^32 logical addresses, you only get 2^32 physical addresses, which limits the amount of RAM.
It's actually even worse than that, though; the virtual address space is still only 2^32. You are probably confused by memories of real mode, where the point of segmentation was to increase the available address space. In protected mode, though, segementation exists only to carve up the address space in funny ways. A bunch of checks are done against the segment (checking privilege level, comparing the offset to the segment limit, etc.), and then if it's all ok the 32-bit offset is added to the segment base to produce a 32-bit linear address (which then goes through the usual paging to produce the physical address). So you are still limited to a 32-bit linear address space for virtual addresses; segments just allow you to access that address space in different ways.
http://en.wikipedia.org/wiki/Segment_selector
He never said it.
I just don't think the man's credible at all. I didn't believe him when he originally said it, and I don't believe him when he says he didn't say it. If he were to say he said it, then I wouldn't believe him either.
My other car is a 1984 Nark Avenger.
Here's an equally blunt response for you: You're wrong, 32-bit OSes can access up to 64GiB of RAM using a feature called PAE. It's a bit of a hack and has some downsides - for example, no individual 32-bit program can access more than 4GiB, and some drivers aren't compatible with it (which is why it isn't enabled by default in client versions of Windows). However, the hardware has been capable of it since the Pentium Pro CPUs (mid 90s)
In reference to the GP, there are typically two variants of standard Linux kernels available for a x86 system: Default (or similar) and Big, BigSMP, or similar. The SMP stands for Symmetric MultiProcessing (ability to use multiple CPUs, CPU cores, execution paths, etc.) and has been integrated into the Default kernel for some time now. The "big" kernels also support PAE. This is not in the default due, I believe, to the risk that some kernel modules such as drivers don't handle PAE correctly (the Wikipedia page also mentions that PAE-enabled kernels won't run on non-PAE-capable CPUs, though this is hardly a concern on any modern machine).
There's no place I could be, since I've found Serenity...
No, I'm saying it was a joke and that some people might want to lighten up a bit ;)
"It does not do to leave a live dragon out of your calculations, if you live near him." - Tolkien
I hope you don't need to change a linux lightbulb - cause you can't find them for sale anywhere, only XP/Vista and Mac bulbs. ;)
I hate Vista, use Linux for most of my computing time and have vowed to stay with XP for the bits Linux cannot do. However, a number of my close buddies who have never even touched Linux also hate Vista and are sticking with XP.
Likewise, the original poster could well be running BSD, FreeDOS or countless other OSes.
Certain members of the Windows community may wish to promote the idea of a "Windows v Linux" war where in reality most of we Linux users use it because it does some jobs we need it to do very well - end of story.
So stop with the bullshit - if you personally like Vista then go for it...
Gentoo Linux - another day, another USE flag.
Too bad you didn't read that he was saying Warcraft 3, not WOW. Given the fact that WarCraft 3 required hardly ANYTHING to really run, there was OBVIOUSLY something wrong with the system, and I doubt it was the hardware he removed. I'll bet $$$ it lies in the other running processes that were not required. As noted by the system requirements:
* 400 MHz Pentium II or equivalent, or a 400 MHz G3 processor or better.
* Windows 98, Windows ME, Windows 2000, or Windows XP. Warcraft III requires Mac OS® 9.0 or higher, or Mac OS X® 10.1.3 or higher
* Warcraft III requires 128 MB of RAM. Virtual Memory should be enabled on computers running pre-Mac OS X versions of the Mac OS®.
* A keyboard and mouse are required.
* A hard drive with at least 550 MB of free space
* 8MB 3D video card (TNT, i810, Voodoo 3, Rage 128 equivalent or better) with DirectX 8.1 support. For Mac OS® systems, a video card consisting of an ATI Technologies or nVidia chipset with at least 16 MB of memory is required.
* A DirectX-compatible 16-bit sound card is recommended. Warcraft III will work with the built-in sound features of the Mac OS®.
Recommended System Requirements:
* 600 MHz processor or better
* 256 MB of RAM
* 32 MB 3D Video card
512 megs of RAM should have been perfectly fine. Anyone saying the OP was wasting time going thru the system in such a thorough fashion is likely just talking out of their ass.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
1) dual channel chipsets require both channels operational.
2) More than 2 slots takes space.
3) There are no 1.5GB dimms.
Of course not everyone uses MS-OS And have a proper chipset for utilisation of 4GB.
Or this can be seen as manufacturers trying to create upgrade wave. If new machines have 4GB of ram the software will adapt to fill it within a year or two. Then OLD machines with 1GB or less require atleast new RAM. But while your at it why not buy entirely a new PC?
©God
3 GB/s is at the low end of system RAM bandwidth, so you must be talking about bytes.
But 1280x1024x32x75=3145728000 bits/second, about 3.1x10^9 bits/second, or 393x10^6 bytes/second.
For most systems this would be about 4 to 10% of memory bandwidth. This article from Tom's Hardware seems to show that the effect on non-3D applications is minimal.
It was too close to the truth perhaps. Truth dosn't go down too well around these parts.
throw new NoSignatureException();
So.. it being a joke lets you off the hook for wantonly insulting Linux developers? No come-backs, counter-claims, or parodies allowed?
Can you be Even More Awesome?!