Personal computing discussed

Moderators: Flying Fox, morphine

 
Robotics
Gerbil
Topic Author
Posts: 59
Joined: Fri Aug 26, 2016 5:33 pm
Location: Chapel Hill/Istanbul
Contact:

Spectre-Meltdown barricade

Mon Aug 13, 2018 9:28 pm

When does cpu's spectre-meltdown problems solved. I want to upgrade new system but this problem blocked me. Do you have any prediction time about this?
 
just brew it!
Gold subscriber
Administrator
Posts: 51858
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Spectre-Meltdown barricade

Mon Aug 13, 2018 10:29 pm

Side channel information leakage is very hard to avoid in any architecture that uses advanced techniques to optimize performance (SMT, speculative execution, multi-level caches, etc.)

Just because CPU vendors claim to have implemented hardware mitigations for the currently known exploits doesn't mean they are immune to future side-channel exploits.

Also keep in mind that:

1. Meltdown only poses a serious risk if you're running a VPS in a cloud environment where "your" hardware is shared with other potentially malicious users; or (in a non-VPS scenario) if an attacker has already compromised your system in a way that allows them to run arbitrary code.

2. The risk from Spectre is primarily via attacks on the sandboxing between different web browser sessions, and mitigation is a cooperative hardware/software thing. You're relying on the browser vendors to modify their JIT compilers to insert "barrier" instructions (which already exist in current CPUs) at points in the code which are vulnerable to speculative execution attacks, and to segregate different sites into separate processes (Chrome already has support for this, though I am not certain it is enabled by default).

Cliff Notes version of the above: Unless you're a Cloud VPS hosting provider, I wouldn't put off a CPU purchase just because you're waiting for better hardware mitigation of side-channel attacks. Practicing "safe computing", and (especially) keeping your OS and web browser updated are much more important.
Nostalgia isn't what it used to be.
 
Topinio
Gerbil Jedi
Posts: 1562
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 2:44 am

Chrome always (IIRC) had 1 process per tab, but has had strict site isolation available since v63 in December last year, now enabled by default as of v67 in May.

Firefox doesn't have it yet, as Google had a head start in knowing about these vulns, but FF also has the issue that it's already too memory heavy (particularly on Linux and Mac, where heavy FF usage makes 8GB machines non-viable in a work context!). Details.

Microsoft took a different approach, pre-Spectre, of trying to prevent some types of remote code execution and have better sandboxing. I haven't seen any update since Spectre, on whether they've thought again.

Safari, I don't know what Apple's up to. Which is stupid, as I'm posting this post using it :roll:
Desktop: E3-1270 v5, X11SAT-F, 32GB, RX Vega 56, 500GB Crucial P1, 2TB Ultrastar, Xonar DGX, XL2730Z + G2420HDB
HTPC: i5-2500K, DH67GD, 6GB, RX 580, 250GB MX500, 1.5TB Barracuda
Laptop: MacBook6,1
 
Chrispy_
Maximum Gerbil
Posts: 4462
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 7:37 am

There are multiple software fixes for Spectre and Meltdown. The cumulative slowdown of any processor from the software fixes is single-digit percentage.

After the first major wave of patches, performance dropped by 3-5% on Intel systems. Two new waves of patches have come since, and even if you take the worst-case scenario, a modern processor is still delivering at least 90% of its original performance.

If you're upgrading a PC, the chances are that you're not doing it for a mere 10% of performance increase, so go ahead and upgrade; It's highly likely that Intel won't have Spectre-immune products on the market until 2020 and even AMD processors are affected by some of the meltdown variants.
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
notfred
Maximum Gerbil
Posts: 4516
Joined: Tue Aug 10, 2004 10:10 am
Location: Ottawa, Canada

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 8:04 am

If you want to get really worried you can look up Net Spectre.
 
Topinio
Gerbil Jedi
Posts: 1562
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 3:36 pm

I'm convinced that the main effect of this is going to be to drive new PC (Mac) sales due to browser mitigations which increase memory usage.

@notfred, thanks. Now thought really does need to be given to using separate machines for separate use cases :wink:
Desktop: E3-1270 v5, X11SAT-F, 32GB, RX Vega 56, 500GB Crucial P1, 2TB Ultrastar, Xonar DGX, XL2730Z + G2420HDB
HTPC: i5-2500K, DH67GD, 6GB, RX 580, 250GB MX500, 1.5TB Barracuda
Laptop: MacBook6,1
 
just brew it!
Gold subscriber
Administrator
Posts: 51858
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 3:44 pm

Topinio wrote:
I'm convinced that the main effect of this is going to be to drive new PC (Mac) sales due to browser mitigations which increase memory usage.

We were already moving towards the "process per site" model anyway. Perversely, this may actually help counteract the other negative impact of Spectre mitigation (higher CPU load), by allowing the browser to take better advantage of lots of cores.
Nostalgia isn't what it used to be.
 
DancinJack
Maximum Gerbil
Posts: 4116
Joined: Sat Nov 25, 2006 3:21 pm
Location: Kansas

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 3:55 pm

Honestly, Safari is really the only usable browser on macOS with <=8GB or RAM already. It's insane. Google and Mozilla both need to do some serious work on Chrome and FireFox for macOS. It pains me deeply to use Safari.
i7 6700K - Z170 - 16GiB DDR4 - GTX 1080 - 512GB SSD - 256GB SSD - 500GB SSD - 3TB HDD- 27" IPS G-sync - Win10 Pro x64 - Ubuntu/Mint x64 :: 2015 13" rMBP Sierra :: Canon EOS 80D/Sony RX100
 
Topinio
Gerbil Jedi
Posts: 1562
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 4:10 pm

Yep, this thing has 4GB and FF is no go. Same on 8GB machines, Safari or go home. Kinda :evil: as I've been using Firefox since it was Netscape Navigator in 1994 or so.
Desktop: E3-1270 v5, X11SAT-F, 32GB, RX Vega 56, 500GB Crucial P1, 2TB Ultrastar, Xonar DGX, XL2730Z + G2420HDB
HTPC: i5-2500K, DH67GD, 6GB, RX 580, 250GB MX500, 1.5TB Barracuda
Laptop: MacBook6,1
 
ludi
Lord High Gerbil
Posts: 8109
Joined: Fri Jun 21, 2002 10:47 pm
Location: Sunny Colorado front range

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 4:18 pm

DancinJack wrote:
Honestly, Safari is really the only usable browser on macOS with <=8GB or RAM already. It's insane. Google and Mozilla both need to do some serious work on Chrome and FireFox for macOS. It pains me deeply to use Safari.

Apple requires all browsers on iOS to use WebKit rendering, and Safari is married very tightly into it. Google forked from WebKit in 2013 for Windows & Linux development, and IIRC Chrome and other third-party browsers on iOS are basically a shell and extensions riding on top of Safari's WebKit engine?
Abacus Model 2.5 | Quad-Row FX with 256 Cherry Red Slider Beads | Applewood Frame | Water Cooling by Brita Filtration
 
DancinJack
Maximum Gerbil
Posts: 4116
Joined: Sat Nov 25, 2006 3:21 pm
Location: Kansas

Re: Spectre-Meltdown barricade

Tue Aug 14, 2018 4:23 pm

I was under the impression it uses Blink on macOS, but I could be totally wrong.

edit: Derp, you said iOS and we were talking macOS. Yeah, I think on mobile it's Webkit or bust for iOS. Though that's not all there is to Chrome on iOS. I know there are lot of networking changes they made in addition so they aren't necessarily a 1:1 comparison other than the Webkit engine.
i7 6700K - Z170 - 16GiB DDR4 - GTX 1080 - 512GB SSD - 256GB SSD - 500GB SSD - 3TB HDD- 27" IPS G-sync - Win10 Pro x64 - Ubuntu/Mint x64 :: 2015 13" rMBP Sierra :: Canon EOS 80D/Sony RX100
 
Amiga500+
Gerbil
Posts: 54
Joined: Wed Sep 21, 2016 2:10 am

Re: Spectre-Meltdown barricade

Thu Aug 23, 2018 12:55 pm

https://perens.com/2018/08/22/new-intel ... cceptable/

Very shady practice from Intel (again).

You will not, and will not allow any third party to
...
(v) publish or provide any Software benchmark or comparison test results.


I'd have doubts its even legal. They should (likely will) be roasted with a lawsuit for this - be it from shareholders,journalists or customers.
 
Captain Ned
Gold subscriber
Global Moderator
Posts: 27388
Joined: Wed Jan 16, 2002 7:00 pm
Location: Vermont, USA

Re: Spectre-Meltdown barricade

Thu Aug 23, 2018 12:56 pm

Amiga500+ wrote:
https://perens.com/2018/08/22/new-intel-microcode-license-restriction-is-not-acceptable/

Very shady practice from Intel (again).

You will not, and will not allow any third party to
...
(v) publish or provide any Software benchmark or comparison test results.


I'd have doubts its even legal. They should (likely will) be roasted with a lawsuit for this - be it from shareholders,journalists or customers.

Already under discussion here:

viewtopic.php?f=2&t=121239
Humans sleep soundly in their beds because rough cats stand ready in the night to visit violence on those who would do us harm.
 
Amiga500+
Gerbil
Posts: 54
Joined: Wed Sep 21, 2016 2:10 am

Re: Spectre-Meltdown barricade

Thu Aug 23, 2018 1:06 pm

Captain Ned wrote:
https://techreport.com/forums/viewtopic.php?f=2&t=121239


I noticed that *after* I posted.

D'oh! :oops:
 
credible
Gerbil XP
Posts: 494
Joined: Fri Oct 07, 2011 7:47 pm

Re: Spectre-Meltdown barricade

Thu Aug 23, 2018 3:08 pm

Handy little tool for some.

https://www.grc.com/inspectre.htm
 
DPete27
Silver subscriber
Grand Gerbil Poohbah
Posts: 3617
Joined: Wed Jan 26, 2011 12:50 pm
Location: Wisconsin, USA

Re: Spectre-Meltdown barricade

Fri Aug 24, 2018 8:16 am

Just read this article that Intel's latest Spectre patch included legal verbiage BANNING public benchmarking of Intel chips with this patch. Apparently this Spectre ordeal is getting pretty dark.

Intel has since removed the outlaw on benchmarking.

Could this possibly be why the 9th-gen i7 will not include hyperthreading?
Main: i5-3570K, ASRock Z77 Pro4-M, MSI RX480 8G, 500GB Crucial BX100, 2 TB Samsung EcoGreen F4, 16GB 1600MHz G.Skill @1.25V, EVGA 550-G2, Silverstone PS07B
HTPC: A8-5600K, MSI FM2-A75IA-E53, 4TB Seagate SSHD, 8GB 1866MHz G.Skill, Crosley D-25 Case Mod
 
just brew it!
Gold subscriber
Administrator
Posts: 51858
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Spectre-Meltdown barricade

Fri Aug 24, 2018 9:05 am

DPete27 wrote:
Could this possibly be why the 9th-gen i7 will not include hyperthreading?

That would not surprise me. The very nature of SMT allows threads running in the same physical core to affect each other's execution timing in very complicated ways, and this means it is a potential minefield when it comes to side channel timing attacks.
Nostalgia isn't what it used to be.
 
Krogoth
Gold subscriber
Gerbil Elder
Posts: 5518
Joined: Tue Apr 15, 2003 3:20 pm
Location: somewhere on Core Prime
Contact:

Re: Spectre-Meltdown barricade

Fri Aug 24, 2018 10:33 am

DPete27 wrote:
Just read this article that Intel's latest Spectre patch included legal verbiage BANNING public benchmarking of Intel chips with this patch. Apparently this Spectre ordeal is getting pretty dark.

Intel has since removed the outlaw on benchmarking.

Could this possibly be why the 9th-gen i7 will not include hyperthreading?


It is more like due to marketing shenanigans if anything else. They want to make i9 brand the new "i7" while "i7" will become the new mid-tier range. The other lesser stuff will be degraded further. HT's vulnerabilities to side-channel attacks being a factor in that decision is entirely coincidental.
Ivy Bridge i5-3570K@4.0Ghz, Gigabyte Z77X-UD3H, 2x4GiB of PC3-12800, Sapphire RX Vega 64, Corsair CX-600 and Fractal Refined R4 (W). Kentsfield Q6600@3Ghz, HD 4850 2x2GiB PC2-6400, Gigabyte EP45-DS4P, OCZ Modstream 700W, and PC-7B.
 
BIF
Gold subscriber
Minister of Gerbil Affairs
Posts: 2309
Joined: Tue May 25, 2004 7:41 pm

Re: Spectre-Meltdown barricade

Fri Aug 24, 2018 11:07 pm

I don't care what anybody says, the slowdowns caused by BIOS+Windows updates are KILLING me with any CPU i5, or i7.

I call BS on anybody who says "it's only" some small percentage of speed penalty. I call BS on anybody who says it shouldn't even be noticeable with "modern" processors. I mean it man, I'm dying here. All of my computing devices are a lot slower now and I'm seeing red.

Punishment should be stiff for virus and malware makers. Whatever we're doing now, it's not enough.

Right now the "cure" (disabling CPU speed/execution logic via BIOS and Windows updates) is no cure at all. It's WORSE than the disease, and I'm feeling it every day. Windows is slow, browsers are slow, fricken' Quicken is slow, and hell, even getting a response from the Windows key is slower than baking bread! Very frustrated, and I wish I could back out these so-called "fixes". I wish I could track down whoever made these latest viruses.

Very much looking forward to a new i7 or i9 architecture that can leave this horrible nightmare behind. And don't tell me it can't be done, because that will just make me madder. :x
 
ludi
Lord High Gerbil
Posts: 8109
Joined: Fri Jun 21, 2002 10:47 pm
Location: Sunny Colorado front range

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 1:01 am

BIF wrote:
I call BS on anybody who says "it's only" some small percentage of speed penalty. I call BS on anybody who says it shouldn't even be noticeable with "modern" processors. I mean it man, I'm dying here. All of my computing devices are a lot slower now and I'm seeing red.

All of my computing devices have been trivially affected at most, and that includes an i7-4600U laptop I'm using right now. What else do you have on "all your computing devices" that makes them feel so slow?
Abacus Model 2.5 | Quad-Row FX with 256 Cherry Red Slider Beads | Applewood Frame | Water Cooling by Brita Filtration
 
Topinio
Gerbil Jedi
Posts: 1562
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 4:02 am

BIF wrote:
I don't care what anybody says, the slowdowns caused by BIOS+Windows updates are KILLING me with any CPU i5, or i7.

I call BS on anybody who says "it's only" some small percentage of speed penalty. I call BS on anybody who says it shouldn't even be noticeable with "modern" processors.
[...]
Very much looking forward to a new i7 or i9 architecture that can leave this horrible nightmare behind. And don't tell me it can't be done, because that will just make me madder. :x

Hear, hear. The solution to your pain might already be out there, though, if AMD CPUs are really unaffected. I hear that Ryzen and Threadripper are quite good, and am planning for the HTPC to go that route next build, would be looking at them for my main machine too if it weren't only 2 years old.

Edit: just came across Theo de Raadt's post from Thursday evening:
Solving these bugs requires new cpu microcode, a coding workaround,
*AND* the disabling of SMT / Hyperthreading.

SMT is fundamentally broken because it shares resources between the two
cpu instances and those shared resources lack security differentiators.
[...]
A few months back, I urged people to disable hyperthreading on all
Intel cpus. I need to repeat that:

DISABLE HYPERTHREADING ON ALL YOUR INTEL MACHINES IN THE BIOS.

Also, update your BIOS firmware, if you can.
[...]
So please try take responsibility for your own machines: Disable SMT in
the BIOS menu, and upgrade your BIOS if you can.

I'm going to spend my money at a more trustworthy vendor in the future.
Desktop: E3-1270 v5, X11SAT-F, 32GB, RX Vega 56, 500GB Crucial P1, 2TB Ultrastar, Xonar DGX, XL2730Z + G2420HDB
HTPC: i5-2500K, DH67GD, 6GB, RX 580, 250GB MX500, 1.5TB Barracuda
Laptop: MacBook6,1
 
BIF
Gold subscriber
Minister of Gerbil Affairs
Posts: 2309
Joined: Tue May 25, 2004 7:41 pm

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 8:57 am

ludi wrote:
BIF wrote:
I call BS on anybody who says "it's only" some small percentage of speed penalty. I call BS on anybody who says it shouldn't even be noticeable with "modern" processors. I mean it man, I'm dying here. All of my computing devices are a lot slower now and I'm seeing red.

All of my computing devices have been trivially affected at most, and that includes an i7-4600U laptop I'm using right now. What else do you have on "all your computing devices" that makes them feel so slow?
Stop right there.

Me: Working fine...
Me: Apply BIOS and update Windows...
Me: Working SLOW!

"What else" do I have? Well, Windows 10, Office, Firefox, IE, and Chrome, and different other software. No two computers are precisely the same, but each got slow immediately after I rebooted after applying the so-called "fix".

And it happened all on different days because I didn't do all this to all of my machines on the same day, but it did happen for each machine immediately after I applied the "cure".

This is clearly caused by the medicine, not by some other illness.
 
BIF
Gold subscriber
Minister of Gerbil Affairs
Posts: 2309
Joined: Tue May 25, 2004 7:41 pm

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 9:01 am

Topinio wrote:
BIF wrote:
I don't care what anybody says, the slowdowns caused by BIOS+Windows updates are KILLING me with any CPU i5, or i7.

I call BS on anybody who says "it's only" some small percentage of speed penalty. I call BS on anybody who says it shouldn't even be noticeable with "modern" processors.
[...]
Very much looking forward to a new i7 or i9 architecture that can leave this horrible nightmare behind. And don't tell me it can't be done, because that will just make me madder. :x

Hear, hear. The solution to your pain might already be out there, though, if AMD CPUs are really unaffected. I hear that Ryzen and Threadripper are quite good, and am planning for the HTPC to go that route next build, would be looking at them for my main machine too if it weren't only 2 years old.

Edit: just came across Theo de Raadt's post from Thursday evening:
Solving these bugs requires new cpu microcode, a coding workaround,
*AND* the disabling of SMT / Hyperthreading.

SMT is fundamentally broken because it shares resources between the two
cpu instances and those shared resources lack security differentiators.
[...]
A few months back, I urged people to disable hyperthreading on all
Intel cpus. I need to repeat that:

DISABLE HYPERTHREADING ON ALL YOUR INTEL MACHINES IN THE BIOS.

Also, update your BIOS firmware, if you can.
[...]
So please try take responsibility for your own machines: Disable SMT in
the BIOS menu, and upgrade your BIOS if you can.

I'm going to spend my money at a more trustworthy vendor in the future.


Thank you, I shall investigate this. Even in my 3 year old i7 machine, I'd rather have 6 fast cores than 12 unresponsive cores.
 
NovusBogus
Silver subscriber
Graphmaster Gerbil
Posts: 1305
Joined: Sun Jan 06, 2013 12:37 am

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 12:38 pm

just brew it! wrote:
Just because CPU vendors claim to have implemented hardware mitigations for the currently known exploits doesn't mean they are immune to future side-channel exploits.

...not to mention the side-channel attacks that already exist but are only known to the spooks and black hats. Staying current on security updates is always a good thing but one must remember that the stuff the media is hyperventilating about is just a slice of what's really out there, so losing sleep about whether X hardware is immune to Y vulnerability is an exercise in futility. Practice safe browsing on any physical systems that you care about, and if it's really important the only way to be reasonably sure it's safe is to physically disconnect it from the Internet. Mobility, stealth, and good operational security always outperform static defenses.
 
CScottG
Graphmaster Gerbil
Posts: 1150
Joined: Fri Dec 01, 2006 9:53 pm

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 2:36 pm

just brew it! wrote:
Side channel information leakage is very hard to avoid in any architecture that uses advanced techniques to optimize performance (SMT, speculative execution, multi-level caches, etc.)

Also keep in mind that:

2. The risk from Spectre is primarily via attacks on the sandboxing between different web browser sessions, and mitigation is a cooperative hardware/software thing. You're relying on the browser vendors to modify their JIT compilers to insert "barrier" instructions (which already exist in current CPUs) at points in the code which are vulnerable to speculative execution attacks, and to segregate different sites into separate processes (Chrome already has support for this, though I am not certain it is enabled by default).



-I think is largely tied to use.

About 5 years ago I decided to segment my computer-use, and as a result: no longer have a "general purpose" *PC that does everything. (..and even before then I used 2nd tier (hyper-visor) VM's for browsing.)

Now I've got a separate machine for *most* web-use, and like TAIL's it's a system memory-only build with regular power-offs. (..the one area of web-use that isn't "amnesiac" is fire-walled-accepted gaming like STEAM on a more traditional Windows system.)


I tend to think that most people would do reasonably well with a modest Linux system for web-use on something as simple as this:

https://libre.computer/products/boards/roc-rk3328-cc/

https://developer.arm.com/support/arm-s ... nerability

If you need power for certain applications or Windows, then make a more traditional computer that doesn't normally use the web (or open-up un-secure programs..).




*..it's funny/sad, but I say "PC" as if most people browse the web like this now, but that's not the case anymore - phones are now the interface of choice for most, and while that's fine - it's unfortunately being used in that "general" like nature where all sorts of sensitive information is stored on the phone (..just ripe for an attack from more vectors than a typical PC).
 
Ryu Connor
Gold subscriber
Global Moderator
Posts: 4322
Joined: Thu Dec 27, 2001 7:00 pm
Location: Marietta, GA
Contact:

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 2:49 pm

BIF wrote:
Thank you, I shall investigate this. Even in my 3 year old i7 machine, I'd rather have 6 fast cores than 12 unresponsive cores.


Just disable the mitigations.

The risk profile doesn't really change for a home user.
All of my written content here on TR does not represent or reflect the views of my employer or any reasonable human being. All content and actions are my own.
 
swaaye
Gold subscriber
Gerbil Team Leader
Posts: 275
Joined: Mon Apr 21, 2003 4:45 pm
Contact:

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 4:52 pm

I've seen that the worst performance losses are storage related so I decided to test my Core i5 6500 machine here. It is a HP business class machine and has a very recent UEFI with all the updates. I used Inspectre to disable Windows mitigations. AS SSD bench actually showed no significant change for the HP M700 SSD in it. Scores of 694 vs 702 (it varies a few points each run anyway too).
 
Topinio
Gerbil Jedi
Posts: 1562
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 4:55 pm

BIF wrote:
Thank you, I shall investigate this. Even in my 3 year old i7 machine, I'd rather have 6 fast cores than 12 unresponsive cores.

NP; I didn't actually expect that turning off Hyper-Threading would help with the sucky performance issue from the patches (I thought it was just another way to reduce the attack surface).

But I tried it on seeing your post, and it helps (Skylake-DT) ! 10 hours since I did so, the machine is more responsive and feels less bogged down. Sucks that it's only 4-way now, and it's supposed to have another 3+ years to run, but one can't have it all.
Desktop: E3-1270 v5, X11SAT-F, 32GB, RX Vega 56, 500GB Crucial P1, 2TB Ultrastar, Xonar DGX, XL2730Z + G2420HDB
HTPC: i5-2500K, DH67GD, 6GB, RX 580, 250GB MX500, 1.5TB Barracuda
Laptop: MacBook6,1
 
BIF
Gold subscriber
Minister of Gerbil Affairs
Posts: 2309
Joined: Tue May 25, 2004 7:41 pm

Re: Spectre-Meltdown barricade

Sat Aug 25, 2018 7:55 pm

Topinio wrote:
BIF wrote:
Thank you, I shall investigate this. Even in my 3 year old i7 machine, I'd rather have 6 fast cores than 12 unresponsive cores.

NP; I didn't actually expect that turning off Hyper-Threading would help with the sucky performance issue from the patches (I thought it was just another way to reduce the attack surface).

But I tried it on seeing your post, and it helps (Skylake-DT) ! 10 hours since I did so, the machine is more responsive and feels less bogged down. Sucks that it's only 4-way now, and it's supposed to have another 3+ years to run, but one can't have it all.


I'm not surprised that you feel the responsiveness improvement, and I'm looking forward to trying it on my own machines. My newer i7 was supposed to go another 3+ years too, but the older i7 (a laptop) is just waiting for money for a replacement; then it will become a web browser for my guest room. I have my heart set on a Eurocom, but will probably wait now for an AMD cpu version or else wait for a new Intel CPU without the issue to be made available. So no rush now on a new laptop. It's funny how priorities can change right after a bogus update (which I think this is).

And finally, I won't be able to do it on the company i5 laptop ever. That machine is outside of my control, at least until I'm due for a hardware upgrade...which will be next year sometime, I think. So yeah, "can't have it all" for sure. But as the song goes, two out of three ain't bad!
 
Robotics
Gerbil
Topic Author
Posts: 59
Joined: Fri Aug 26, 2016 5:33 pm
Location: Chapel Hill/Istanbul
Contact:

Re: Spectre-Meltdown barricade

Sun Aug 26, 2018 1:28 am

BIF wrote:
Topinio wrote:
BIF wrote:
Thank you, I shall investigate this. Even in my 3 year old i7 machine, I'd rather have 6 fast cores than 12 unresponsive cores.

NP; I didn't actually expect that turning off Hyper-Threading would help with the sucky performance issue from the patches (I thought it was just another way to reduce the attack surface).

But I tried it on seeing your post, and it helps (Skylake-DT) ! 10 hours since I did so, the machine is more responsive and feels less bogged down. Sucks that it's only 4-way now, and it's supposed to have another 3+ years to run, but one can't have it all.


I'm not surprised that you feel the responsiveness improvement, and I'm looking forward to trying it on my own machines. My newer i7 was supposed to go another 3+ years too, but the older i7 (a laptop) is just waiting for money for a replacement; then it will become a web browser for my guest room. I have my heart set on a Eurocom, but will probably wait now for an AMD cpu version or else wait for a new Intel CPU without the issue to be made available. So no rush now on a new laptop. It's funny how priorities can change right after a bogus update (which I think this is).

And finally, I won't be able to do it on the company i5 laptop ever. That machine is outside of my control, at least until I'm due for a hardware upgrade...which will be next year sometime, I think. So yeah, "can't have it all" for sure. But as the song goes, two out of three ain't bad!


I do same thing. And only solution for me;

Ryu Connor wrote:

Just disable the mitigations.

The risk profile doesn't really change for a home user.

Who is online

Users browsing this forum: No registered users and 2 guests