qubit
Overclocked quantum bit
- Joined
- Dec 6, 2007
- Messages
- 17,866 (3.00/day)
- Location
- Quantum Well UK
System Name | Quantumville™ |
---|---|
Processor | Intel Core i7-2700K @ 4GHz |
Motherboard | Asus P8Z68-V PRO/GEN3 |
Cooling | Noctua NH-D14 |
Memory | 16GB (2 x 8GB Corsair Vengeance Black DDR3 PC3-12800 C9 1600MHz) |
Video Card(s) | MSI RTX 2080 SUPER Gaming X Trio |
Storage | Samsung 850 Pro 256GB | WD Black 4TB | WD Blue 6TB |
Display(s) | ASUS ROG Strix XG27UQR (4K, 144Hz, G-SYNC compatible) | Asus MG28UQ (4K, 60Hz, FreeSync compatible) |
Case | Cooler Master HAF 922 |
Audio Device(s) | Creative Sound Blaster X-Fi Fatal1ty PCIe |
Power Supply | Corsair AX1600i |
Mouse | Microsoft Intellimouse Pro - Black Shadow |
Keyboard | Yes |
Software | Windows 10 Pro 64-bit |
I saw this on vr-zone and couldn't believe what I found! This is a real, genuine conspiracy theory that Intel designed USB so that the CPU needed to poll the device, be slowed down and encourage CPU upgrades. Heck, it even comes with a disclaimer that this isn't vr-zone's general opinion!
I doubt that it's true, but it's better read while wearing your best tinfoil hat...
vr-zone
I doubt that it's true, but it's better read while wearing your best tinfoil hat...
[Disclaimer: The following views are solely the author's and do not necessarily represent those of the VR-Zone team]
Blast from the Past - USB polling induced performance slowdowns
Reported by Nebojsa Novakovic on Thursday, July 12 2012 6:51 pm
You all remember how USB, even when used by the keyboards and mice, tend to ever so slightly slow down the system, due to the often unnecessary CPU polling? Well, once even these key input devices standardize on USB3 and Thunderbolt, that sorry story should be over...
The ancient (some 25 years old by now) PS/2 keyboard and mouse interface, developed by IBM for the same-named PC series in the mid eighties, is pretty much adequate for those two input devices, unless you're talking about extreme gamer-style super-responsive mice that send their movements like 1,000 times a second or so. The little Intel 8042 keyboard controller chip handled all the work, not bothering the CPU much there. However, these connectors are, despite the still wide device availability, increasingly rare on the modern mainboards nowadays. Why?
Well, conspiracy theorists say that when Intel first pushed their Universal Serial Bus, one major issue for them then was that the CPUs weren't fully loaded by then-typical OS and apps, so there had to be ways found to coerce buyers to part with their money for ever faster CPUs any possible way. One way was to load the CPUs with the I/O processing burdens, such as 'software sound' instead of Sound Blaster, 'software modems', and, yes, USB. The trick with USB was that it constantly had to be polled by the CPU, even if the peripheral connected did absolutely nothing, like the case of a keyboard without any key pressed. Polling introduces interrupts, task switching and such overhead, which of course bogs down the CPU, introducing the 'need for more speed'. Sometimes, even PC benchmarks were, and still are, slightly but still negatively affected by using the USB keyboard and mouse instead of the PS/2 one. Not to mention the battery drain on mobile devices...
Fast forward to today and USB 3.0. It is the first USB update to finally remove the need for polling. After all, the bloatware of today knows well how to load any CPU, even multi core ones, so too-probing I/O interfaces become a liability. Now, if you have a super-speed mouse that needs polling to get its response speed, it can have it - but a keyboard with a few letters typed per second, doesn't need it that much, obviously.
Here's the problem though - there are no USB 3 keyboards. Simply, they don't need USB3 extra speed, so why add the cost? Then comes the Thunderbolt, essentially an extended PCIe v2 lane combined with a DisplayPort link onto a single connection. Now, this goes, via the controller, directly to the chipset or CPU PCIe lanes, with minimal complexity - therefore less CPU load as well. So, both USB3 and Thunderbolt peripherals should, where possible, reduce or eliminate the unnecessary CPU I/O monitoring load left from the ancient USB times.
Again, you're not likely to see these two interfaces on keyboards and mice. So, to complete the picture, the old PS/2 should, in my mind, make the comeback - and yes, both connectors, just like in the good old days without the artificially induced slowdowns.
vr-zone
Last edited: