Intel Gags Customers from Publishing Performance Impact of Microcode Updates

mvmiller12 said:
The new version isn't really any better, stating that you are forbidden from publishing any before/after patch performance benchmarks. That is 100% bullshit on a product I purchased with my money. This is not an NDA pre-release embargo kind of thing - this is a flat-out non-expiring restriction on what I can do with my hardware. Fuck that.
Click to expand...

Ummm, where does the new version you cannot do benchmarks and compare them before and after? To me the new one is just bunch of legalese that has bearing for us ordinary citizens.
 
MaZa said:
Ummm, where does the new version you cannot do benchmarks and compare them before and after? To me the new one is just bunch of legalese that has bearing for us ordinary citizens.
Click to expand...

From the front page, specifically referring to the updated terms:

The company has hence updated the license terms governing the microcode update distribution to explicitly forbid its users from publishing comparative "before/after" performance numbers of patched processors.
 
mvmiller12 said:
From the front page, specifically referring to the updated terms:

The company has hence updated the license terms governing the microcode update distribution to explicitly forbid its users from publishing comparative "before/after" performance numbers of patched processors.
Click to expand...

That still refers to the old one. Click the new link below it and there is not a single reference to benchmarking.
 
The new license:

Copyright (c) 2018 Intel Corporation.
All rights reserved.

Redistribution.

Redistribution and use in binary form, without modification, are permitted, provided that the following conditions are met:

  • Redistributions must reproduce the above copyright notice and the following disclaimer in the documentation and/or other materials provided with the distribution.
  • Neither the name of Intel Corporation nor the names of its suppliers may be used to endorse or promote products derived from this software without specific prior written permission.
  • No reverse engineering, decompilation, or disassembly of this software is permitted.
“Binary form” includes any format that is commonly used for electronic conveyance that is a reversible, bit-exact translation of binary representation to ASCII or ISO text, for example “uuencode.”
 
notarat said:
You will not, and will not allow any third party to (i) use, copy, distribute, sell or offer to sell the Software or associated documentation; (ii) modify, adapt, enhance, disassemble, decompile, reverse engineer, change or create derivative works from the Software except and only to the extent as specifically required by mandatory applicable laws or any applicable third party license terms accompanying the Software; (iii) use or make the Software available for the use or benefit of third parties; or (iv) use the Software on Your products other than those that include the Intel hardware product(s), platform(s), or software identified in the Software; or (v) publish or provide any Software benchmark or comparison test results."

Intel is 1st party
Mobo Mfgr is 2nd party
You are 3rd party (so it applies to you, as well.)

Do not update your Microcode until the license terms are changed or you are in violation of the license terms (if you bench and publish the results)are open to legal issue. It's that simple.

Edited to clarify - (if you bench and publish the results)
Click to expand...

You've been trained too well if you really think Intel could take legal action against a person based on what's in this stupid EULA.
 
Knurrus said:
Intel are sugarcoating the new license that they are committed to open source, and it is true, they have and are contributing a lot to open source with the appropriate open source licenses. But the above license is far from an open source license, and the last clause is unenforceable at least in several EU member countries.
Click to expand...
Exactly. I imagine it was more of a copy paste error than anything else. It is 100% non-enforceable. Laughable even.
 
mvmiller12 said:
The new version isn't really any better, stating that you are forbidden from publishing any before/after patch performance benchmarks. That is 100% bullshit on a product I purchased with my money. This is not an NDA pre-release embargo kind of thing - this is a flat-out non-expiring restriction on what I can do with my hardware. Fuck that.
Click to expand...

The license terms are for the software 'patch' that fixes their hardware security vulnerability. You can do whatever you want with the CPU you bought, just not publishing benchmarks with their software update installed.....

Either way, it's completely stupid on their part and make me wish I invested in AMD a few months back. The hit to DB/Virtualization performance will move a lot of enterprises away from intel during the next hardware refresh. Who the hell is going to spend a few 100k+ to upgrade their datacenter with an intel CPU knowing you HAVE to install a patch that will reduce performance by 10-30%+ on your main workloads.... Yea intel is gonna feel the pain from this over the next 1-2 years, and AMD is going to increase market share a TON.
 
At least they caught their over zealous-ness. Is that a word? Anyways, take note NVIDIA.
 
Wait, their clarification here really didn't help, or help their cause.

So, you have to agree to the license terms in order to get the microcode, and in order to do so you have to agree to not benching before and after?

I wonder how that would hold up in a court of law?
 
GoodBoy said:
Was dumb of them to try and restrict benchmark results, especially since they published their own:
https://www.intel.com/content/www/us/en/architecture-and-technology/l1tf.html

Majority of workloads are unaffected or show +-1%
postgres database takes 18% performance hit.
virtualization takes up to 31% hit.
webserver workload took a 7% hit.
There were a few others I didn't recognize what they were, that were about 10%.

I doubt most of us will see any impact from these changes...
Click to expand...

Reading more of the notes on those benchmarks they published, it compares it to the prior microcode patches, so it is only showing the additional performance hits from the second set of fixes. The first set of fixes from January are what those results consider as the baseline... which means the total performance impact of this is probably 50% or more for the most impacted workloads (virtualization), and 35 to 40% on the database workloads... that's really significant.

Hopefully, for most normal end user use-cases, we will see relatively little impact.

But now I am hoping that [H] can do some cpu benching with a before and after, where the before is prior to any of these fixes at all, vs after. So something at a december 2017 firmware and os patch level, vs something with latest bios and os patches. Then we can get an idea of how fucked up this mess is..
 
Got KB4100347 last night on win 10, spectre patch, made my pc crash with 60 seconds or so of login. Completely borked my pc. Finally was able to get to control panel and uninstall it, then use the MS hide or show utility to prevent this patch from installing again
 
Lifelite said:
Fucking hell. Intel is just one screw up after another lately. This is what happens when you are practically the only choice for so long.

Great timing on AMD's part to step up there game. I still think they need to go a bit further so they can get back to the point of having near as much performance as Intel Processors for a much lower price, like they used to be in the 2000s.
Click to expand...

All AMD has to do is wait for Intel to release one or two more microcode updates. Once that happens AMD will probably have the performance crown again.
 
choppedliver said:
Got KB4100347 last night on win 10, spectre patch, made my pc crash with 60 seconds or so of login. Completely borked my pc. Finally was able to get to control panel and uninstall it, then use the MS hide or show utility to prevent this patch from installing again
Click to expand...

Gotta love the MS way of doing things. lol

I have an older Intel chip and I do turn off Meltdown and Specter patches.

The difference is as a Linux user. If I am doing something important. I can reboot and turn them on/off at will. lol
 
dgz said:
You've been trained too well if you really think Intel could take legal action against a person based on what's in this stupid EULA.
Click to expand...
Intel has the budget to initiate legal action and drag you to court even if they can't win.
 
jpm100 said:
Intel has the budget to initiate legal action and drag you to court even if they can't win.
Click to expand...

They also have the budget to hire a "cleaning service" and disappear you. Doesn't mean they are actually going to do it.

Dragging you into court over this would take a really corrupted judge and a severe case of stupid on their own side. There would be nothing to gain, and the risk is super high.
 
dgz said:
They also have the budget to hire a "cleaning service" and disappear you. Doesn't mean they are actually going to do it.

Dragging you into court over this would take a really corrupted judge and a severe case of stupid on their own side. There would be nothing to gain, and the risk is super high.
Click to expand...

Risk is only high if you have good representation (and it's either sponsored or you can afford it). There are many MANY examples of cases in the US that seemed like they should be a no-brainer in a particular direction which where lost/abandoned because of lack of good representation and/or $$$ to properly fight. Make no mistake, the system is rigged against the player with the least $$$ - and the larger the discrepancy, the harder the fight will be for that party. This is why we have pro bono initiatives like the EFF around to help the little guy (but their resources are finite - they can't help with EVERY case).
 
Intel needs to watch themselves. I could still see using my old i7-920 for day to day it would be fine. For gaming my 2600k does fine still. It isn't like the latest 90s where you get massive gains anymore. We really have hit the spot where current software (daily productivity software) doesn't need more. Games even have mostly what they need with a 3+ year old i5/7 and rely on the video card you have.

Computing habit changes with the move to mobile is what will get Intel if they don't pivot more/faster IMHO.
 
mvmiller12 said:
Risk is only high if you have good representation (and it's either sponsored or you can afford it). There are many MANY examples of cases in the US that seemed like they should be a no-brainer in a particular direction which where lost/abandoned because of lack of good representation and/or $$$ to properly fight. Make no mistake, the system is rigged against the player with the least $$$ - and the larger the discrepancy, the harder the fight will be for that party. This is why we have pro bono initiatives like the EFF around to help the little guy (but their resources are finite - they can't help with EVERY case).
Click to expand...

Good thing you mentioned EFF. Turns out there's also the European Digital Rights (EDRi) and EFF is listed as on the many members. Apparently good representation is likely to be provided if Intel goes after a US/EU based person over this one., which I think they wont.

Sure they have money but they're likely to lose the case, meaning they get to cover all legal fees. They have nothing to gain. It's a very specific text for their partners only
 
You must log in or register to reply here.
Back
Top