This is a simple and quick poll to determine how much performance you lost to the updates, this poll will handle all systems, Mac Linux, Windows, Android, etc.
It was said by intel that the average performance drops in modern chips is 6% but can easily range from 2% - 20%
This is to cover the broad spectrum, because I know a lot of people still use older chips, we will get the average of user drop, instead of controlled drop from the company to hide face if any. If you don’t know then you can default to 6% (or the equivalent). There is no need to mention BIOS versions or updates as not all even will update. (example, this Laptop I am using now had a 20% drop on skylake but a BIOS update upped performance 18% of the gap, it can’t be assumed that everyone will take a corporate head’s advice). Now, on to the poll…
Each step is 5%, so to figure out the current average, multiply the number by 5
Mitigating the damage done by an update. Instead of having the processor second guess itself so it doesn’t speculate, it would instead have the feature fully disable so it processes it as normal (minus the speed boost). With the patch, the processor does things twice for some tasks. (unless they just re-enable the function regardless of the risk…)
Not to mention that the BIOS could have the processor or some device locked. AN example is an old Dell I had, never hit 100% CPU, was locked to .6 GHz for some reason. Updated BIOS, fully unlocked the processor, etc.
It all depends on the manufacturer. Or they just push a BIOS that forcefully overclocks the components so they seem snappier, like an OC’ed iGPU (would not recommend in a laptop).
There are other examples but a lot of them feel outdated by now.
Edit:
Though, if your manufacturer had to put out a BIOS to improve performance, it is likely they capped the hardware or limited it in some way, unless they just made lazy/bad firmware
Intel Core i7 x990 6 core here. Under both Linux and Windows, I’d estimate just under 10% hit on the high end. Overall, in terms of frame rate and responsiveness, I think it gets lost in the noise given enough ram to keep things out of swap.
By unlocking or removing restrictions capping clock, voltage, or other overclocking parameters available to end users. By enabling the use of an extra core on chip that might normally be disabled due to semiconductor fabrication reasons (see also PS4, for example).
So the new chip works faster, but after a few years , is designed to run slower.
That means that when users who buy those chips 1 or 2 steps behind expect the same speed difference, they would still be left with lesser performance, compared to the new High-price scheme chip.
Hmm, how much is it to design and sell those chip that works actually?
No, they are not designed to do that. The current performance hit is because of a new class of attacks which was discovered and some of the performance improvement features of modern CPUs had to be disabled/modified to mitigate the security issues.
If this was by design that would mean it’s planed obsolescence and that is illegal in certain countries.
lol, I meant that other systems are more efficient for military application.
The system of the intel chip which is potentially slowed down to 20% of it’s normal rate can in fact be slowed down to over 50% of that , even 99%.
To mitigate the powers that be may not solve the security issue.
It is rather insulting.
I did not verify how to solve the problem, but it may improve the performance on my system if there are any of those solutions to improve performance.
Of course, if that proved to be false, it may not be worth it either.
I do plan on getting more hardware and other hardware, including AMD chip to compare the effectiveness and method of operation.
i still don’t get what military applications have to do with all of this. Do you mean because of the security issues? Because that is a problem for everyone and not just for a specific domain.
I did not apply the patches so far on my personal machines. I run free software almost exclusively and I don’t really think that my applucations suddenly start to hack eachother. However, and I think this is more important: Block ■■■■■■■ javascript in the browser and only enable code from sources you trust.
Why should military application have any to do with it?
Military application are being used for security yes, that is a system for everyone, not just for a specific domain.
I did not apply the patches, nor download any on my personal machine.
What is the :
" Block ■■■■■■■ javascript "
you mentioned?
Just because the chip and symptons or conditions is not the best for military application does not mean that military application have anything to do with it.
In fact, it rather means that it does not have anything to do with it, since military application use other hardware which may or may not be prone to this kind of attacks.
They use the same chips everyone else uses. There is no special line of CPUs just for military use. Even the NSA uses Intel chips, they famously asked for the switch to disable the ME in their processors.
This issues have absolutely nothing to do with military or not. Everyone is basically in the same boat.
Many sites on the web use javascript to run active code directly in your browser. The spector variants of the new vulnerabilities can be used to read out your whole memory via a javascript program and send it all back to an attacker on the net. There are plugins like noscript which block javascript and let you select what to run and what not.
That’s not true.
The processors used on electronic systems are not same.
For one thing, some are used for electronic warfare, which civilian systems do not have the right to and power to.
The NSA will not go and shoot a missile on your house, or get a jet fighter or laser satellite to intercept it, each with their own electronic system and processors.
If any one had those, they could hack it.
Try to find the patent to top secret systems, for some reasons, they are not listed for public view.
Just because a military system would be more efficient than a non-military one does not mean that it is not or that it has anything to do with.
It can be more efficient to not have anything to do with, as much as it can be more efficient if it did have anything to do with it.
There are parameters on which those efficiency comparison are based on.
That is how they are researched and developed .
It’s not that hard to trace the attacker who gets the data or fake data.
Those affect vulnerable systems more, like if someone was made to need the attacker id.
That is also why satellites going to mars or other have to be secured enough against those vulnerabilities or risk failure.
That is also why I rather design and make my own to prevent this.
It’s easier to run 15 computers with only 1 or 2 connected to the internet and the rest running as data security systems which do not have to rely on JavaScripts or 3rd party programs or hardware.
Of course, it is more like a proprietary system, and many of those are not marketed in part to prevent those security risk.
Sure there are still security risk and problems to solve and damage to repair.
It’s also more cost efficient to apply for patents.
Edit:
The 4004 microprocessor is one of 4 chips constituting the MCS-4 chip-set, which includes the 4001 ROM, 4002 RAM, and 4003 Shift Register. With these components, small computers with varying amounts of memory and I/O facilities can be built. Three other CPU chip designs were done at about the same time: the Four-Phase System AL1, done in 1969; the MP944, completed in 1970 and used in the F-14 Tomcat fighter jet (not to confuse with BobCat) ; and the Texas Instruments TMS-1000 chip, announced in September 17, 1971. Both the AL1 and the MP944 use several chips for the implementation of the CPU function.
Handy
posted on Mar, 5 2005 @ 09:30 AM link
Modern fighter jets generally don’t use off the shelf items. Instead, low ‘speed’, larger ‘scale’ RISC processors with advanced real time operating systems are used. The problem in most aircraft is not speed but durability and survivability. What’s the point of having the worlds most advanced jet if it crashes when it flies above a satelite link.
The Shuttle uses a complex of 5 IBM AP-101B flight computers to negotiate the flight. These systems are slow, well known systems based off 60’s technology. With the recent upgrades the human/machine interface was updated with digital equipment run by a number(3 i believe) off the shelf 80386s.
https://www.quora.com/Lockheed-Martin-F-22-Raptor-How-fast-is-the-CPU-of-an-F-22-of-the-latest-model
Andy Meecham, Worked in military sensor design and OA for nearly 20 years
Answered Mar 14, 2016 · Upvoted by Doug Hanchard, worked at Aviation
When you ask “how fast is the CPU?” If you’re asking how complex and/or capable is the avionics system, taken as a whole, then the answer is very. Almost certainly more so than any other system flying now or in history - and there’s a significant probability that it will be the most complex avionics system that will ever be developed.
Like virtually all “5th generation” fighters it is practically impossible for the pilot to fly the aircraft without the aid of the computer based flight control system, but in terms of sheer processing “grunt” it doesn’t need to crunch lots of numbers very quickly. It just needs to process a limited number of inputs - from the instrumentation and control inputs - very quickly, very accurately and very reliably and never crash (excuse the pun!).
Many PC/server/laptop manufacturers deliver processor microcode updates with the BIOS. CPU microcode (“ucode”) is specific to different product classes and implements (or changes) CPU features and MSRs. ucode updates can also be applied directly by an OS if it has that capability - Intel CPUs at least are updatable without a restart. The microcode driver of the OS just temporarily halts the core, updates it, resumes it, and then continues scheduling threads onto it. ucode updates are not permanent - they must be applied every time the core/socket is reset (like, during a reboot or after the machine has been turned off) so this is why they’re commonly packaged with and applied by the BIOS of a system.
In the case of Meltdown/Spectre, Intel released a ucode update recently that adds some MSRs which allow OS makers to implement certain countermeasures - the ucode features themselves don’t solve the Meltdown/Spectre issue, they just implement some glue that enables some countermeasure methods to work in the OS.
Yes, I know about the microcode updates, but since the kernel is able to update (and therefor it is possible to deliver them as regular OS updates) what else would a BIOS update be needed for? It’s probably just some other firmware which gets loaded before handing everything to the CPU. I should probably read some more about this stuff one day
