2016+ CPU/GPU News

[Bikerdude]

nVidia will stop supporting x86(32bit) operating systems after the 390.xx driver -

After Release 390, NVIDIA will no longer release drivers for 32-bit operating systems for any GPU architecture. Later driver release versions will not operate, nor install, on 32-bit operating systems. Driver enhancements, driver optimizations, and operating system features in driver versions after Release 390 will not be incorporated back into Release 390 or earlier versions.

 

This decision affects all supported versions of Microsoft Windows -- Windows 7, Windows 8.1 and Windows 10 -- as well as Linux and FreeBSD.

 

Nvidia will release critical security updates for drivers until January 2019 however.

- http://www.dsogaming.com/news/nvidia-will-stop-supporting-32-bit-operating-systems-after-its-geforce-390-drivers/

[Oktokolo]

Finally a good decision from nVidia!

It is about time to let legacy x86 die. All the devs that had to support the ancient architecture (despite all consumer CPUs being fully AMD-x64-compatible since a decade) will thank nVidia for getting the last nostalgic users to finally switch to an x64 OS. Less compatibility overhead means more time for polishing the x64 versions.

[Bikerdude]

Upcoming security fix may decrease performance on Intel’s CPUs by up to 30%

 

We all know that Intel’s CPUs run extremely well all modern games, and way better than AMD’s offerings. However, things may change as an upcoming security fix may decrease overall performance on Intel’s CPUs by up to 30%.According to reports, a fundamental design flaw in Intel’s processor chips has forced a significant redesign of the Linux and Windows kernels. Microsoft will soon roll out these changes in an upcoming update and according to early benchmarks, there will be an impact of 5-30%, depending on the task and the processor model. What’s really interesting here is that this security bug/issue affects a lot of Intel’s CPUs. In fact, this bug is present in all Intel processors that were produced in the past 10 years.

 

As TheRegister reported, this security issue allows normal user programs to discern to some extent the contents of protected kernel memory.

It allows normal user programs – from database applications to JavaScript in web browsers – to discern to some extent the contents of protected kernel memory. The fix is to separate the kernel’s memory completely from user processes using what’s called Kernel Page Table Isolation, or KPTI. At one point, Forcefully Unmap Complete Kernel With Interrupt Trampolines, aka FUCKWIT, was mulled by the Linux kernel team, giving you an idea of how annoying this has been for the developers. Whenever a running program needs to do anything useful – such as write to a file or open a network connection – it has to temporarily hand control of the processor to the kernel to carry out the job. To make the transition from user mode to kernel mode and back to user mode as fast and efficient as possible, the kernel is present in all processes’ virtual memory address spaces, although it is invisible to these programs.

 

When the kernel is needed, the program makes a system call, the processor switches to kernel mode and enters the kernel. When it is done, the CPU is told to switch back to user mode, and re-enter the process. While in user mode, the kernel’s code and data remains out of sight but present in the process’s page tables.The downside to this separation is that it is relatively expensive, time wise, to keep switching between two separate address spaces for every system call and for every interrupt from the hardware. These context switches do not happen instantly, and they force the processor to dump cached data and reload information from memory. This increases the kernel’s overhead, and slows down the computer.

 

Furthermore, it appears that this security fix will only affect Intel’s CPUs. AMD claimed that its CPUs are not subject to these types of attacks.

AMD processors are not subject to the types of attacks that the kernel page table isolation feature protects against. The AMD microarchitecture does not allow memory references, including speculative references, that access higher privileged data when running in a lesser privileged mode when that access would result in a page fault.

Microsoft will issue this security fix next week, so we’ll be sure to benchmark some games in order to see how much this fix will affect the performance on our Intel CPU.

Edited January 3, 2018 by Bikerdude

[jaxa]

Careful with this subject. There is a ton of FUD out there.

 

But it does look like it is serious and that Intel's CEO sold all the stock he was allowed to before it came to light.

[jaxa]

Intel Responds to Security Research Findings

 

^ Translation

Edited January 4, 2018 by jaxa

[stumpy]

it supposedly will only effect virtual machines and cloud servers, pcgamer found a site that did some tests and its a loss of 1 fps on most games, due to games running through gpu and not cpu.

 

http://www.pcgamer.com/serious-intel-cpu-design-flaw-may-require-a-windows-patch-but-probably-wont-affect-gaming-performance/

Edited January 4, 2018 by stumpy

[jaxa]

it supposedly will only effect virtual machines and cloud servers

 

Impact on gaming performance is expected to be minimal, but this part of your statement is obviously not true. Certain operations will be made slower on consumer hardware.

[lowenz]

Patch is here: https://support.microsoft.com/en-us/help/4056892/windows-10-update-kb4056892

[lowenz]

 

Impact on gaming performance is expected to be minimal, but this part of your statement is obviously not true. Certain operations will be made slower on consumer hardware.

Using the old (fixed) Quake 2 soft renderer - ah, the 1997 - there's a 2.5% performance hit (i3 6300) @1280x960

Edited January 4, 2018 by lowenz

[lowenz]

Using LLVMpipe with the yamagiQ2 OpenGL 1.4 renderer (OpenGL using the CPU as GPU) there's NO performance hit (0%)

 

So guys turn off the red alert

[peter_spy]

Is this present in win 7 as well?

[lowenz]

Yes.

 

Here the patches!

 

win 10: https://support.microsoft.com/en-us/help/4056892/windows-10-update-kb4056892

win 8.1: https://support.microsoft.com/en-us/help/4056898/windows-81-update-kb4056898

win 7 SP1: https://support.microsoft.com/en-us/help/4056897/windows-7-update-kb4056897

[Bikerdude]

Well the intel issue isnt going away -

 

Update from intel for affected processors

http://www.fudzilla.com/news/processors/45310-intel-rushes-out-spectre-and-meltdown-updates

 

Old and new gen consoles also effected

http://www.dsogaming.com/news/old-gen-and-current-gen-consoles-amd-cpus-and-smartphones-are-also-vulnerable-to-the-spectre-threat/

Edited January 5, 2018 by Bikerdude

[Bikerdude]

Here is the list of all the Intel processors that are currently being affected by Spectre and Meltdown -

• Intel® Core™ i3 processor (45nm and 32nm)
• Intel® Core™ i5 processor (45nm and 32nm)
• Intel® Core™ i7 processor (45nm and 32nm)
• Intel® Core™ M processor family (45nm and 32nm)
• 2nd generation Intel® Core™ processors
• 3rd generation Intel® Core™ processors
• 4th generation Intel® Core™ processors
• 5th generation Intel® Core™ processors
• 6th generation Intel® Core™ processors
• 7th generation Intel® Core™ processors
• 8th generation Intel® Core™ processors
• Intel® Core™ X-series Processor Family for Intel® X99 platforms
• Intel® Core™ X-series Processor Family for Intel® X299 platforms
• Intel® Xeon® processor 3400 series
• Intel® Xeon® processor 3600 series
• Intel® Xeon® processor 5500 series
• Intel® Xeon® processor 5600 series
• Intel® Xeon® processor 6500 series
• Intel® Xeon® processor 7500 series
• Intel® Xeon® Processor E3 Family
• Intel® Xeon® Processor E3 v2 Family
• Intel® Xeon® Processor E3 v3 Family
• Intel® Xeon® Processor E3 v4 Family
• Intel® Xeon® Processor E3 v5 Family
• Intel® Xeon® Processor E3 v6 Family
• Intel® Xeon® Processor E5 Family
• Intel® Xeon® Processor E5 v2 Family
• Intel® Xeon® Processor E5 v3 Family
• Intel® Xeon® Processor E5 v4 Family
• Intel® Xeon® Processor E7 Family
• Intel® Xeon® Processor E7 v2 Family
• Intel® Xeon® Processor E7 v3 Family
• Intel® Xeon® Processor E7 v4 Family
• Intel® Xeon® Processor Scalable Family
• Intel® Xeon Phi™ Processor 3200, 5200, 7200 Series
• Intel® Atom™ Processor C Series
• Intel® Atom™ Processor E Series
• Intel® Atom™ Processor A Series
• Intel® Atom™ Processor x3 Series
• Intel® Atom™ Processor Z Series
• Intel® Celeron® Processor J Series
• Intel® Celeron® Processor N Series
• Intel® Pentium® Processor J Series
• Intel® Pentium® Processor N Series

And here is a test done by EpicGames on one of thier servers after they patched, the chart below shows the significant impact on CPU usage -

 

[Anderson]

Celeron potatoes have been taken care of.

[Bikerdude]

If you have an AMD based laptop/.desktop, do not install the Meltdown/Spectre patch - as you may end up with an un-bootable machine -

 

http://www.dsogaming.com/news/microsoft-halts-distribution-of-spectre-meltdown-fixes-on-amd-systems-after-reports-of-unbootable-pcs/

 

Which is ironic as I have had 3 macbook's become unbootable after the user did an update from the appstore.

[nbohr1more]

The latest Nvidia driver has a Spectre fix.

 

I didn't notice any substantial performance loss but some folks are reporting issues.

[lowenz]

It's not a fix of the driver.

It'a a fix of the compiler used to build the driver

 

So no big performance impact is expected, it's something about the execution of the driver on the CPU/OS side.

Edited January 10, 2018 by lowenz

[jaxa]

STOP PATCHING

 

https://www.theverge.com/2018/1/22/16919426/intel-advises-pause-deployment-of-spectre-patch

 

https://techcrunch.com/2018/01/22/linus-torvalds-declares-intel-fix-for-meltdown-spectre-complete-and-utter-garbage/

 

https://lkml.org/lkml/2018/1/21/192

[Bikerdude]

Oh for the love of christ!

[lowenz]

It's a real.....MELTDOWN

[lowenz]

If there weren’t detail-oriented, no-BS, old-school coders out there watching out for the likes of you and me, the great complacent unwashed out here in userland, we would have to take whatever Intel and the others hand us and thank them in our ignorance. I for one am glad to have people smarter and more uncompromising than myself fighting on our behalf, however “shouty” they may be.

 

Well said.

Edited January 22, 2018 by lowenz

[Oktokolo]

The honest thing to do would be a full-scale product recall. Neither Intel nor any other vendor will do that though. They would go bankrupt on that. So we will get some (more or less lame) mitigations by microcode, new compiler behaviour and explicit changes in some exposed code and will still be vulnerable until we buy a new CPU (likely from AMD instead of Intel), that has been fixed on the hardware level.

 

But when retpoline finally arrives in the stable gcc on my distribution - i certainly will recompile my Kernel and entire userland immediately. Until then, the web contains a lot less JavaScript for me than before i knew about Spectre. uMatrix reduces my attack surface a lot.

[jaxa]

uMatrix reduces my attack surface a lot.

 

Good on you for using the most powerful fork of uBlock.

[Oktokolo]

Always thought that uBlock misses a lot of opportunities when it comes to granularity of control and completeness. uMatrix is the first thing i saw, that i would actually call a browser firewall.

And we really need those in a time where browser makers keep on adding "features" consisting of millions of lines of code without even considering the implications of all the bugs that naturally come with the growth of any code base.

 

Too bad, uMatrix still can't protect against the threats to privacy caused by diverse browser fingerprinting techniques on the hand-selected sites i do allow to run some JavaScript...