Inside the Windows Vista Kernel, Part 2

BuR4N writes "Mark Russinovich takes a look at the Windows Kernel and the changes made in Vista. In this second part he describes the workings of the features SuperFetch, ReadyBoost, ReadyBoot, and ReadyDrive and how they improve system performance."

Comments

Vista seems quite slow to me

[El Cubano]

Some friends were visiting last night and they had recently purchased a new HP laptop (1.6 GHz CPU and 1 GB RAM with 80 GB HDD). I was struck by how abysmally slow Vista was. The thing had Vista Home Premium on it. Putting a blank CD entailed a wait of anywhere from 15 to 25 seconds before the stupid dialog came up asking if I wanted to burn something to the blank disc. Connecting to a wireless network was a complete disaster. My wireless network is setup to not broadcast its SSID, so I had to enter the setting manually along with the WPA password. As soon as I was done, the thing would take the dialog away and then not connect. It took me 30 minutes of hunting to find the listing that had the wireless networks I had manually entered in (as opposed to the networks which were broadcasting). To top if off, the system kept prompting to allow things that it really seemed I should not need to be asked. I am seriously not trying to troll here, this is just
my first impression of vista.

Re:Vista seems quite slow to me

[TheRaven64]

Anything would be slow on that laptop (well except for dos or stripped down linux distros). Excuse me? Anything slow on a 1.6GHz machine? What on earth are you smoking? The only things that should come close to taxing a 1GHz or faster x86 chip are:

1. Video editing.
2. Large compile jobs.
3. Some resource-intensive games.

My two backup machines are a 1.5GHz G4 and a 1.2GHz Celeron M, and they are both acceptably fast for 90% of what I do. For the vast majority of users a 1GHz Pentium III is more than adequate. You don't need to run 'dos or a stripped down linux distro,' Windows 2000, XP, or a full *NIX distrubution will be perfectly happy. I have a 500MHz UltraSPARC IIi on my desk, and it has no problems with a full install of Solaris 10, a desktop environment, and a few apps.

You can run DOS or a small *NIX distribution quite happily on a low-end 486 (a 286 lets you run a lot of DOS apps pretty fast). You certainly don't need anything like a 1.6GHz machine.

My pentium M 1.8ghz I bought a year ago runs XP slow Either you have a tiny amount of RAM, or a huge amount of malware. Windows XP was released in 2001. The Pentium M was released in 2004. You are running XP on a CPU two Moore-generations newer than the fastest CPU available at the time of XP's launch.

Re:Inside the kerne;l

[Saffaya]

I wish I had some mod points today to give you some +funny.
Thanks for the laugh :)

WTF

[abradsn]

Windows Vista uses the same boot-time prefetching as Windows XP did if the system has less than 512MB of memory, but if the system has 700MB or more of RAM, it uses an in-RAM cache to optimize the boot process.

Okay, so I just wanted to nitpick a sentence here. What happens between 512 and 700. I presume it does the same thing at XP would have. But this sentence is confusing, and perhaps implies that perhaps Ms. PacMan will get launched in this scenario.

Overall though, an interesting series. Kudos to the author.

Re:Slowing down over time?

[gad_zuki!]

I have found the exact opposite to be true. I hate reinstalling and almost never do. I have installs that are years old that are pretty damn fast. The most obvious culprits are users who install any app they can find, have dozens of system tray icons, dozens of startup objects, dozens of unecessary services, spyware, etc. They complain about 'slowness' then do a reinstall which only removes all this unecessary software and blame MS. I'm more than a little skeptical fo these claims.

Granted, the OS could be doing moer to assist these users, but for the part its just poor user maintenance. Of course, all OS's will develop crust over time, but that doesnt mean a noticable performance drop. My aging system at home, which gets some serious abuse, produces the same FPS in bf2 as someone with a fresh install with the same hardware. It encodes video just as quickly. It feels as reponsive. It runs graphical benchmarks just as well. If windows performance degrades over time than I have some mystical power to be immune from this. Or more likely, the crap most users do to their computer just piles up, they dont bother to examine their systems, do a reinstall, and just bitch about MS.

Re:Where's the Beef?

[QuietLagoon]

With all these performance-improving things, shouldn't performance actually, you know, be improved?

Yes, it should.

The fact that performance has not improved is the reason behind articles like this in which Microsoft is talking about how great Vista is, when it really is disappointment.

Where's the room for incremental improvement?

[ivan256]

If they pick the best names the first time around, they won't have any room to innovate new fancy names for these technologies in the next Windows.

Really the title of this article should be "Microsoft Implements Fresh New Names for Existing and Obvious Technology in Vista Kernel."

Re:Is this secure

[Slashcrap]

That depends on who your neighbor is: a pimple faced, socially awkard, teenager that spends his spare time turning toasters into Linux boxen or a grandmother who can't seem to access voice mail on her new "modular" telephone? Considering that encryption is AES 128, it may be compromised by one of the two types of neighbors.

If you replaced "pimply Linux geek" with "respected crypto analyst with large amounts of computing resources at his disposal", you'd still be just as wrong.

Did you confuse AES128 with 56bit DES or something?

If you do have an example of someone breaking AES128, don't post it - publish it.