Yea, yea, I know. I too have a Compute Stick that Microsoft has updated with an incompatible driver and now I see the flashing window issue (Grrrrr ). Block these updates or replace them with a working version after the fact is all that you can do.
You can get this issue escalated far faster (and with higher priority) if it is done through (by) the board/system vendor. I am not saying don't post the issue here; I am saying make sure you (also) post through the board/system vendor.
I'm not sure how to this nicely but I'll try,
It seems that with your suggestion and based on the fact that INTEL provides mostly drivers and over 90% of these drivers are used in machines NOT built by Intel, this forum becomes meaningless. Most of the posts that relate to issues have units manufactured by other companies.
Could you please think about the above before you respond further on the same path that you've done so far?
If Intel had not been dropping the ball so frequently these days, we wouldn't be in this situation. But I guess standards have fallen.
Be brutally honest; I am. I am a big boy and I can take it.
The problem is that Intel has tested their drivers on a representative set of systems. The BIOS on the motherboards out in the world may not match up with this representative set. In addition, in the case of laptops - and especially those that have hybrid graphics (i.e. utilize both the Intel subsystem and an add-in solution from ATI/AMD, NVIDIA, etc.), the vendor may be modifying the Intel driver to support switching, etc. with their hybrid solution. We push folks to the vendor's web site because the drivers that they post there are the ones that have been tested on the hardware, with the latest BIOS, etc. If you want my opinion, graphics drivers shouldn't be distributed via Windows Update. This creates way, way more problems that it addresses.
While I am retired and not directly in the line of fire, I still take exception to your statement about Intel dropping the ball. Intel didn't and hasn't drop the ball.
First of all, we are talking about vulnerabilities that have existed for quite some time (more than 10 years in some cases). They are only now being discovered because the sophistication of the attack vectors being utilized have progressed to the point where they have been exposed. Intel is in the process of delivering microcode updates for literally hundreds of processor designs and are on schedule delivering them. Intel committed to a schedule and is holding to it.
When a vulnerability like this is discovered, the announcement of its discovery is supposed to be held back until the fixes for the vulnerability are ready. Announcing the vulnerability ahead of time is completely irresponsible. It opens a window for the nefarious to take advantage of this vulnerability before anyone has a fix for it. The problem is that the discoverer doesn't want to miss his five minutes of fame and the media scum want their scoop. I have some choice words for the media scum which, unfortunately, I cannot use here.
The third problem, specific to this situation, is that the motherboard vendors simply do not want to provide support. They want to take their boards, throw them over the wall and move on to the next one. BIOS updates? Forget it! They would rather cheaply hire some company - who don't know a thing about the products initially - to provide this support for them -- and then give them explicit instructions to punt every issue they can to Intel (the less issues they process, the less it costs them overall). It is deplorable. Yes, support standards have fallen - and Intel is as guilty of this as anyone else. There is a difference, however; Intel does not compromise on security issues. If a security fix is necessary, it is delivered for every product that it applies to, even those long EOLed. How many board vendors will you find these days that meet this benchmark? They are refusing to update older products right and left, even those only a few years old.
That's my position,
P.S. As it turned out, the problem I was seeing when Windows Update upgraded my video driver was the result of a bug in the BIOS. If I had been a little proactive in installing BIOS updates, I wouldn't have seen an issue at all.
Well several points that you raised so I'll try to address them, hopefully in order.
1. First off, don't think you addressed the main point I raised - "If one is directed to board/unit manufacturer instead of Intel, then what is the purpose of these boards? Only for retail product owners? If that is so then most of the posts are at the wrong place since most posts that I have seen are from people with products from Lenovo, IBM, Dell, HP etc etc."
2, At the end of the day, the drivers carry Intel's name not the board/unit manufacturer and hence the questions that people ask on this site and their frustration. With the scenario you mentioned, there is no ownership and no responsibility and no accountability. Every company in the chain just shrugs off (while each one has made money at the same time) any issue.
3. And, consumers have to start somewhere and who better than the original manufacturer of the component whose name is on the driver? Now if Intel comes back and can state that it is not their driver causing the issue then of course we can get the manufacturers involved. Looking at the several versions of the drivers that Intel has released recently, that has not seemed to be the case so far.
4. I've had to roll back Intel Graphics and Wireless drivers to get my machine to work as it did before, though I'm still looking at wifi performance issues right now. And looking at other forums and other sites online, issues are not just with the above 2 drivers but are also with Intel chipset drivers etc. Hence my comment about standards being lowered. I think you might be referring to the processor-related security error but that wasn't even on my radar when I made the comment. I was and still am, only concentrating on component drivers right now.
5. Of course, once Intel Management Drivers are updated to fix this security issue, a significant performance drop would certainly frustrate consumers. Like getting a car with 500 hp and with a fix to the engine, one now gets 400 hp.
6. Glad that you agree that standards have fallen all across. Just because I mentioned Intel explicitly does not mean that I have a very high opinion of standards at manufacturers. At Lenovo, on one of their forums, the company asked people to revert their drivers to older versions as a fix, though even that did not result in a fix. Such standards should not be acceptable in this day and age. And I have posted on Lenovo forums about Lenovo's standards as well.
[I edited to add numbering so I can respond more easily...S]
My responses (by the numbering I added to your message)...
- Let me be clear: Your primary point of support is *always* supposed to be your board/system manufacturer. A portion of the selling price for their product is supposed to be for support. Unfortunately, for the reasons I detailed earlier (but primarily cost), they punt to Intel if they can. When I push back on an issue, it is because the issue in question can very easily - and most often is - the result of problems in some other component like the BIOS (which Intel has absolutely no control over). As I pointed out, for example, I have seen lots of issues reported against the Graphics driver that are actually the fault of the BIOS, not the driver. In those places where I push back, I try to provide as much information as I can to help them directly, but will not waste too much of my time on these types of postings (I am retired and a volunteer). If a post is not one that I have the knowledge to answer, I leave them for the ICS agents to triage (after 21 years with Intel, I know a *lot* more than the ICS agents, but I don't have the escalation paths available to me that they have).
- I say again, your board/manufacturer is *always* supposed to be your primary point of contact for all support. Regardless, if they cannot help, Intel does not ever decline to (at least attempt to) provide support for all issues. That does not mean that they (and volunteers like myself) are not going to push back where appropriate. There are only so many resources available to handle issues.
- Yea, I know that folks think this way. They are wrong and I am willing to tell them that.
- I agree; Intel's driver packages leave a lot to be desired (witness the issues that the ICS team is having trying to get a update tool to work properly). I can't do anything about this.
- First of all, the remediation for the vulnerabilities in the ME (and AMT) do not affect performance at all. Neither does the remediation introduced via microcode. The cost is in the workarounds that the O/Ss are having to do (but Intel accepts responsibility for this as well). The drops are not significant for most folks. Unfortunately, the workloads used in many of the benchmarks are heavily skewed towards the workarounds (when there is I/O bounding, for example). For some people, they are going to complain regardless of what they see. This will always be the case if they think they can get something out of it. The few places where big delays are being seen are being looked into.
- Hell yes. This has been happening for a long, long time. All shareholders care about are profits. Maximize profits. Employees means nothing. If profits are down, get rid of some employees is always the first response. Quality is a perception and it has been proven that spending more to increase this perception never works. I have seen cases where increasing the price of a product did more to increase the quality perception than anything else ever could.
Look, we both have spent way too much time on this discussion. I have received notifications of two dozen new posts while I was typing this (sigh!).