Intel Slashes Computer Startup Times

An anonymous reader writes "At Intel's Developer Forum in Taiwan, Intel introduced a new Non-volatile caching technology called 'Robson'." The new Robson cache technology allows computers to start up almost immediately and load programs much faster. Intel declined to comment on the specifics of how the technology works only saying that 'More information will be revealed later'.

Comments

I wonder

[venya]

Hmm... I hope this doesnt require big changes to computer architecture...

The REAL reason

[cdrdude]

The real reason more informatin will be revealed later is that their computers are still booting up!

Re:The REAL reason

[haraldm]

For me, this sounds like a company trying to prevent customers from going elsewhere soon. Announce early, release late if at all, keep customers. It's not that we didn't see this before. Is anybody else known to work on something like that?

Hmmm.......

[_Sharp'r_]

So are we going to all be expected to hibernate our Robson's now?

Why does this sound like a CowboyNeal joke to me?

Re:Hmmm.......

[DigiShaman]

You sound as though you make an inclination to a flaccid penis.

Ohhh come on!! It's a joke! Laugh ;)

I hope this is real

[Monstard]

I hope this is real and not vapor-ware. I've been waiting for instant start for 20 years.

Re:I hope this is real

[JoeCommodore]

Man, you need to upgrade your hardware if it takes THAT long to boot your system!

Re:I hope this is real

[Sloppy]

I've been waiting for instant start for 20 years.

Uh.. 20 years ago, you probably had it. My VIC-20 took less than a second from powerup to the READY prompt. My Amiga 500 was slower, but still a lot faster than anything modern. My Amiga 3000 was even slower, because I had it do more. My 5-year-old Linux boxes are even slower to boot than that.

Computers just keep getting slower. I'm afraid to see how slow a new dual-Opteron machine is.

Re:I hope this is real

[AuMatar]

Most of the boot time of a camera isn't memory loading, its hardware initialization and calibration.

If Transmeta licensed it...

[Black Parrot]

Would we get a Robson Crusoe?

Apple?

[great throwdini]

FTFA: "It's up to the [equipment manufacturers] to decide how it will be implemented. My guess is that enterprise users will likely see it first," [Mooly Eden, VP and GM of Intel's mobile platform group] said.

S.Jobs: "Oh, yeah?"

...one can dream.

If this kind if thing is a concern

[cxreg]

you're booting too often

Re:If this kind if thing is a concern

[ChrisGilliard]

Actually, maybe the reason we're _not_ rebooting too often is because this technology has not existed. No one will wait 5 minutes for a computer to startup, but this might make it more reasonable to do so. With the current energy crisis, I like this idea.

Re:If this kind if thing is a concern

[bypedd]

Agreed, but perhaps that's the Unix background peeking through. Windows in general needs too much rebooting, I say. More time should be spent on making it more dynamic and flexible so you don't have to restart your computer every time you uninstall a program or update windows.

Re:If this kind if thing is a concern

[martinX]

If you'd posted with your name, I'd now know more about you than I ever really wanted to.

That would be bad enough, but since you posted AC, I now suspect everyone...

Re:If this kind if thing is a concern

[Comatose51]

Windows XP user, can't help it.

I kid, I kid.

I reboot maybe once a week. However, I also work in IT and a reboot really does solve a great majority of problems on the platform. It's not so much the OS as other programs. The worst part is that Windows doesn't have an unconditional kill so some process just never dies and never lets go of all the files and handles. So when you go to restart the program, it fails because a previous instance is still hanging onto the files/handles. So we have to reboot the machine.

The thing that drags when we reboot our big Dell workstations isn't so much loading the OS as loading other programs and the SCSI detection process. Then there's the log in script that runs. Robson will only really help with a small chunk of the total boot-up time. As our computers get more networked, I expect network lag to drag us down as well during boot-time.

Re:If this kind if thing is a concern

[quantum bit]

The worst part is that Windows doesn't have an unconditional kill so some process just never dies and never lets go of all the files and handles.

Yes it does. Killing a process from task manager is the same thing as kill -9. When the process dies it unconditionally releases all file handles, mutexes, and any other resources that it had open.

The only time that won't work is if the process is stuck in a system call somewhere (i.e. in the kernel). That usually means buggy device drivers which unfortunately are all too common in the Windows world. It could also be a bona-fide kernel bug, though those are fairly rare (but I do know of one way to cause a vfs lockup on any version of NT -- including fully patched 2k3 server -- without admin rights).

I see the same thing happen all the time on Linux. For example if a process is stuck trying to read a file that's on an nfs server that has become unreachable, not even kill -9 will get rid of it. Even *BSD sometimes gets unkillable processes in cases where the underlying hardware has gone to lunch. I see it sometimes with flaky CD burners, for example.

Re:If this kind if thing is a concern

[hazem]

My box at work is a win2k. I come in, turn it on, and then go make a liter of tea. When I come back, it's just popping up to leg me log in. I log in, then go down to the vending machine 4 floors down (walking both ways) to get a snikers or something and come back. Right about that time, it's finally doing all it's post-login stuff and I'm ready to work. That's a good 10 minutes out of my day.

I don't know what I'll do if they make the damn thing boot up immediately. My boss would probably expect me to start working too.

Not all progress is a good thing!

Re:If this kind if thing is a concern

[MSZ]

Yes it does. Killing a process from task manager is the same thing as kill -9.

ROTFL.

Unix "kill -9" will terminate any process, regardless of the process' attempts to keep going. Windows task manager will kill almost any, but not all processes, most of the time. But when you really need it, it turns out you hit that "almost" part.

The difference is, in unix type systems, SIGTERM and SIGKILL are handled by the OS and the process is only informed of them (so it can try to shut down properly), in Windows, the process is being asked nicely to close. Windows process is free to ignore these events.

Thus it's quite easy to end up with unkillable process. Not to mention that some processes are considered system (or something) and task manager will refuse to kill them flat out. On Linux you can kill even init if you like (not a wise thing to do - but you can).

There is no unconditional forced kill in Windows. Even MS admits that.

Re:If this kind if thing is a concern

[quantum bit]

The difference is, in unix type systems, SIGTERM and SIGKILL are handled by the OS and the process is only informed of them (so it can try to shut down properly), in Windows, the process is being asked nicely to close. Windows process is free to ignore these events.

No, no, and no. It's true that in the "Tasks" or "Applications" tab, hitting End Task will send a request (WM_QUIT) to exit. That's not what I'm talking about. I mean the Processes tab. That is handled by the OS; it routes through Win32 first but ends up at NtTerminateProcess (ZwTerminateProcess). Go read the API reference or even the DDK if you don't believe me. Maybe in the dark days of Win9x that was true, but the NT kernel is a real OS, no matter what other crap you layer on top of it.

There are only three states a process can be in where it's unkillable.

1. "Access denied". This happens on some system processes because they run as the user SYSTEM (equivalent to root), where your task manager process is not. The security descriptor on those processes is set so that nobody except SYSTEM (not even Administrators) can kill them. They can be killed by running task manager as SYSTEM. There are various methods to run a process under the system account; the easiest is by using the "at" command to have the scheduler service start it. Newer versions of task manager also have a list of processes it will refuse to kill, but you can still kill them by using pskill or some other third-party utility that has no such restrictions.

2. Process is stuck in the kernel somewhere. Happens when system calls never return, which isn't supposed to happen. Often due to bad drivers -- even with flaky hardware it SHOULD timeout eventually. I've seen add-on firewall software that hooks the TCP stack and can cause this condition. Sometimes you can get one unstuck by kicking the kernel in the head (i.e. removing or stopping the offending device), otherwise a reboot is the only way to clear it. Unless you're running a checked build with a remote serial debugger, but not many people outside of driver developers do that.

3. Process has a debugger attached. In this case, simply kill the debugger instead.

News for Nerds!

[dj245]

Stuff that will be revealed at a later date, if market conditions warrant its release.

From TFA

[mincognito]

The laptop with Robson also opened Adobe Reader in 0.4 seconds, while the other notebook required 5.4 seconds.

Presumably, the other notebook was running Intel's next generation CPU with sixteen cores.

Should Boost Battery Life a Lot

[BondGamer]

The biggest application for this will probably be laptops. If the computer has 1GB of space for a page file and other stuff, then it will spend a lot less time accessing the hard drive. Less hard drive spinning means longer battery life.

Re:Should Boost Battery Life a Lot

[swf]

Just think of the performance increase from putting the page file into RAM!

Re:The big secret...

[macklin01]

While cute, that's not entirely accurate. A well-maintained WinXP installation with antivirus installed still boots in the 30 second range on a P4 with a decent amount of RAM. It's the extra stuff that can really slow it down. (OpenOffice or MS Office, taskbar goodies, etc.)

Just like a really good Gentoo installation can boot up very quickly, but it can take awhile to go through the process if it isn't so well-optimized. Out-of-the-box on a dual boot P4, it's been my experience that WinXP boots faster than out-of-the-box Linux. (But I'm not enough of a linux guru to trim it down.) -- Paul

I don't get the "instant-on" craze

[ZombieRoboNinja]

I've been hearing this touted for over a decade, now. "In the future, your PC will turn on as quickly as your TV!"

The thing is, I don't care how long my computer takes to boot. With decent sleep and hibernate modes, I don't need to boot more than a couple times a month anyway - and that's usually rebooting for software updates. (If you're wondering, this is on a PowerBook G4 laptop).

It takes my computer under a second to wake up from sleep mode. How much more "instant" does it need to get?

Now, those quick-loading programs, on the other hand, do sound appealing...

Faster to?

[SWroclawski]

If I were using Windows, I could start my computer in seconds.

And have it then crash in... seconds.

Has anyone RTFA?

[Ricardo]

This doesnt seem to be about start up times at all (except from Hibernation). All it is, is a large HDD cache. This will do nothing to make PCs "Start up" Faster. It only has affect in the Article [aparrently] because the "slower" laptop had put its HDD to sleep.
I think PC Hardware and Software manufacturers really do need to work on the glacial boot times that PCs have. Unfortunately, this is only a solution to some of the minor problems, and not the main ones.

This will probably only be for Windows PCs

[Chrismith]

Windows User: Hooray! Now that my computer boots in six seconds, my productivity will be way up!

Linux User: Boo...ting? Oh...that thing I had to do when I first plugged it in. Gotcha.

Instant Startup Isn't All New

[Phat_Tony]

I read that instant startup was supposed to be a feature of the Apple Lisa, and I thought I'd heard before that this had been implemented, but I couldn't quickly Google up any references to it.

At any rate, the theory behind instant startup isn't too hard, it's just an engineering implementation.

All you do is make it so that, following shutdown procedures, the computer immediately switches to startup, except keeping track of the fact it was "shut down," not "restarted." When it finishes restarting, it writes the startup RAM state to disk, then turns itself off.

Upon being turned on, the computer just writes the stored RAM state back from the disk to RAM, and presto! It's just like starting up the computer, except really fast. At least, that was the theory. I've been sort of surprised not to see this implemented, it seems like everyone would like to see fast startups, but hardly anyone cares how long it takes to shut down (especially with soft power)- you're done with he computer anyway. I've heard that a lot of work goes into decreasing boot times for Windows and OSX. It seems like a lot less work to implement an "instant startup" plan, and then not have to care much if startup takes forever, than to carefully track, fiddle with, and optimize everything that happens during startup.

Of course, with this system, restarting after a crash would not be instant, it would take just as long as ever. So it might work to greater advantage on some operating systems than others, depending on why you usually restart.

Re:Instant Startup Isn't All New

[Just Some Guy]

Upon being turned on, the computer just writes the stored RAM state back from the disk to RAM, and presto! It's just like starting up the computer, except really fast. At least, that was the theory.

The reality ain't so hot. In the meantime, your network connections have dropped. Your Kerberos ticket or domain login has expired. Your clock has drifted but your NTP client hasn't noticed that it hadn't been running in hours.

There's nothing earth shattering that can't be explicitly dealt with, but the problem is that there are a million and one little things that you'd never think of that have to be accounted for. It'd be like you waking up from a year-long coma, and realizing that you'd lost your job and your girlfriend even though it only felt like you'd been away for five minutes.

DOH!

[Spy der Mann]

From TFA:

"Chipmaker demonstrates 'Robson' flash memory to boost laptop startup speeds."

Mystery solved.

If they just took the crap out...

[Animats]

The laptop with Robson also opened Adobe Reader in 0.4 seconds, while the other notebook required 5.4 seconds.

This isn't a load time problem. It's a load crap problem.

"Loading and verifying WebBuy.api" (does anyone ever use WebBuy, Adobe's DRM system for PDF documents?)
"Checking for updates" (Adobe might have changed the format of PDF again.)
Loading ad content for toolbar. (Sigh.)

And then all the crap that's being downloaded has to be scanned for viruses. It's all that junk that's the problem.

Of course, OpenOffice isn't all that great on launch time either. And no, loading it at boot time isn't the answer.

Re:If they just took the crap out...

[NotAnotherReboot]

Someone feel free to correct me if I'm wrong, but Acrobat 7 finally fixed the issue of loading a bazillion plugins at startup that almost no one uses. I believe it now dynamically loads them as needed.

Load times for Acrobat 7 vs. Acrobat 6 are clearly far less. The fix often mentioned is to delete/move non-key plugins from the Acrobat plugin folder, but their solution finally fixes the problem in an elegant way.

"DANGER, DANGER, WILL ROBSON!"

[Caspian]

(The message to be displayed when the cache gets corrupt...)

*dodges tomatoes*

Boot times

[ashground]

Actually, yesterday I took a stopwatch to my computers to compare their boot times (comparing my new PowerBook G4 1670 with 1gb of RAM and my old Athlon 1700+ desktop with 768mb of RAM).

Both computers are running a similar load of software at boot. The PC boots with Palm Desktop, Rainlender, and a web server (Abyss) while the Mac boots with Quicksilver and a web server (Apache). Other than that, everything else is pretty standard--audio drivers, video drivers, tablet drivers, and so on. Most of these things are present on both computers. The Mac is a month or two old, the PC hasn't been formatted in two years or so.

Everything timed at home with a stopwatch.

First up--the amount of time it takes from pushing the power button until you have a usable login screen.
Mac--139 seconds
PC--38 seconds

Next--the amount of time it takes from entering your password until you have an idle workspace (on Windows, this was when things stopped loading in the system tray, on OSX this was when the Finder menu appeared completely).
Mac--50 seconds
PC--9 seconds

So, complete boot time (plus whatever time it takes to enter a username and/or password)...
Mac--189 seconds
PC--47 seconds

Finally--the amount of time from the time you click "shutdown" until your computer is powered off.
Mac--53 seconds
PC--11 seconds

So, the time it takes to do a complete reboot...
Mac--242 seconds
PC--58 seconds

Instant-on would be fantastic if it could recover from crashes. There's nothing more frustrating than waiting three minutes for my laptop to boot.

Re:Boot times

[amadeusb4]

To reduce boot, login and shutdown times, upgrade to Tiger (10.4). Here are times for mine(PB G4 with 768MB running 10.4.0):

• boot to login = 29 sec.
• login = ~25 sec (extended by startup items like iCal and stickies)
• shutdown = 11 sec

These numbers are a huge improvement to 10.3.9 running on my cube but then again the cube is nearly 5 years old.

Regarding the non-volatile booting, I would like to point out that my C-64 was already doing that.

Re:Boot times

[nick this]

Yes... osx boot times completely suck. That's why I only reboot after updates.

Time from opening lid of powerboot to idle desktop: 2-3 sec.
Time from closing lid of powerbook to glowing white "sleep" light: 4-5 seconds, but doesn't much matter.

I wouldn't want to be called an apologist or anything, but my laptop seems *way* faster to me than my xp box, just because my pb is essentially instant-on, what with the quick sleep times. It is annoying when you have to do a full boot though. Although 10.4 is some better in this regard than 10.3. Guess it's dictated by usage. Perhaps they spent time optimizing sleep times, not boot times, in that they expected people to sleep more often. Dunno.

Re:Boot times

[TimmyDee]

I'm sorry, I realize this is a non-scientific comparison, but this is anything but scientific. You're comparing a laptop, which likely has a 5400 rpm drive, to a desktop, which likely has a 7200 rpm drive. One of the major limiting factors at bootup is the disk. With so much seeking and reading going on, a difference in disk speed will be a huge hit to the laptop -- any laptop.

Also, comparing Mac OS X 10.3 to XP SP2 (as you failed to note in your original post) is also a bit bunk. Apple completely rewrote the boot sequence for 10.4 and, as a result, has dramatically decreased boot time.

Oh, and one more thing, you PowerBook has more memory. Memory tests happen at startup. The extra 256 MB of RAM may add a few seconds to the boot sequence.

I admit I'm an Apple fan, but we need to have a level playing field if we're going to compare these things.

Dual Boots

[whogben]

I think this could be most interesting with regard to dual boots - especially with regard to Mactels with Windows onboard. You can switch operating systems without rebooting, or without going through all the loading and calculation involved with rebooting. Virtual PC has a "Save PC state on shutdown" option which already does this. When you quit virtual PC, everything its doing is simply stored, and recalled very fast when you reopen this. Implementing this for x operating systems on your Mactel can't be impossible. Each OS stores itself and then restores the one you want before terminating itself - you could switch from Linux to OS-X to XP in as little as 20 seconds each change!

Re:What do you mean, "one can dream"?

[jellomizer]

I don't know why Intel is working so hard to try to make Microsoft look good. Improvements to hardware can't fix shitty software.

Because when Microsoft looks good Intel Looks good. Most people do not know the difference between the OS and the hardware. When the OS is slow they get new hardware, figuring their computer is just old and slow. While in the short term this may sound good but what will probably happen people will be frustrated with the intel system (Figuring it is a peace of junk) and Go with AMD or what ever else. And by chance they may go with a PC manufacture that doesn't pre-load the computer crap so they get a computer that seems extremely fast so people my not go with Intel again.

Windows is even more embarassingly beaten when you compare OS X Server with Windows 2000 or 2003 Server. Those fuckers take FOREVER to reboot.
This again may point to the hardware. A lot of time when I see a window server boot a bulk of the time is before it gets to the OS Level it is just probing for SCSI devices or doing a detailed check on all the ram (The issues TFA is saying it improved) If you want to see slow take a look at a Sun Enterprise system, they can take 5 minutes before they show you anything on the screen. The reason for this slowness is the fact that because these systems should go down often they need a full check on the hardware to make sure nothing is wrong after month/years of uptime.
Also the issue with Windows vs. OSX Server is that Windows can run on Any Box so it needs to check for as many possibilities as possible. While OSX knows what to do when it asks for the hardware configuration and the hardware responds XServe G5 32gb RAM. You can fault windows on a lot of thing, But I give them credit for being able to run on all the crap it does.

IME, OS X boot times beat the living shit out of Windows boot times. I've seen years-old, sub-1GHz G4s boot faster than home-built (i.e. lacking all the extra, cycle-eating horseshit programs that hobble your average Dell or HP PC) 2.0+GHz Wintel boxes with fresh installs of XP.
I don't know I have seen XP having a rather snappy boot up time, it is about on par with OS X. The only real difference is XP tends to still boot after the start button appears, allowing you to access the interface. While OS X takes a little longer in the Splash screen.

To the OP - misunderstanding cameras, Doh! etc.

[new500]

You complain about the zoom extending. Uhu, have you looked at compact multi-element zoom designs? 12 or more elements, many or even most of which geared independantly is not uncommon. The longer the zoom - comsumer guys want optical 10 times zooms, which would be unheard of in a professional lens for many other considerations (predominantly aperture speed and distortion characteristics) and that means even more complicated designs, even allowing small lenses are easier and simpler to design. Now try shifting all those elements _accurately_ with a tiny low voltage low torque servo (see why it's low torque here - too fe2w turns possible in such a small space to get a focus throw long enough to try to do this quickly and accurately and repeatably*). This is why my piezo-wave-effect ring-motor driven Nikkor zoom is several times more expensive of itself than almost any digicam.

Got the idea?

To the above poster - i sure hope there's not much calibration going on when i boot my Nikon. Unless it's to compensate for working temperature effects, if i've spent time and effort having a lens tuned to how i like it (yes this doesn't just happen, it's common) i want it to be left alone at that spec. Now that even modest digicams such as the Fuji F10/11 boot instantly and respond extremely quickly, there's simply no excuse for slow electronics and (electronic) shutter save at the real budget segment.

* even some (sadly many) professional photogs insist on continuing the myth that because the lens / sensor is small, everything remains sharp because the DOF (depth of field) is greater in those conditions. Er, DOF is a psychological effect which is a function of the print enlargement factor, print size, viewing distance and airy dic resolving limit - so the assumption is not true at equivalent apertures, hence the need even in very small "format" cameras to _still_ focus accurately, in OP's case, sadly, slowly too. The effect observed is anecdotally true however at small print sizes like 6" by 4".

About time we got back to the mid nineties.

[Goth Biker Babe]

Up to about 1996 my regular computer was one that booted virtually instantaniously. It's just that it didn't run Windows, Mac OS or Linux. RISC OS (as mentioned on Slashdot a few days ago) was/is in ROM/FLASH and was there the moment the machine started. I held off moving over to a PC/Linux basically because of boot times. Admittedly with linux you just leave the machine on so it's not an issue but Windows was/is a real pain.

Yaawwnn

[Stumbles]

Well LinuxBIOS has been getting boot up times in the 3 second range for a while. Nothing new move on. http://www.linuxbios.org/index.php/Main_Page