How much of a leg up have Intel's security issues given Ryzen? - Page 2 - Overclock.net - An Overclocking Community

Forum Jump: 

How much of a leg up have Intel's security issues given Ryzen?

Reply
 
Thread Tools
post #11 of 17 (permalink) Old 06-16-2019, 05:38 PM
New to Overclock.net
 
SwitchFX's Avatar
 
Join Date: Jan 2007
Posts: 726
Rep: 14 (Unique: 11)
Right. Not that I know much about uarchs, because I know little, but would it have been possible for Intel to still get those performance strides, mainly with Core in 06, then Sandy in '11 and so forth with security in mind or would the major booms after slower cadences been a nothingburger?

SwitchFX is offline  
Sponsored Links
Advertisement
 
post #12 of 17 (permalink) Old 06-16-2019, 08:07 PM
New to Overclock.net
 
Caffinator's Avatar
 
Join Date: Apr 2014
Posts: 138
Rep: 2 (Unique: 2)
Unfortunately it's a chicken and egg problem. This algorithm(although implemented through hardware logic) is simply branch prediction - and if you can characterize the hardware algorithm, you can manipulate the outcome. So then you write code that produces the outcome you desire - triggering branch prediction to load protected memory regions into a space which can be read by your program.

If you want to software patch this, it cost CPU cycles to ensure protected region cannot be loaded.

Or, you can disable the functionality.

Both reduce performance by hardware design, because you are disabling performance-enhancing features. That is all.
Caffinator is offline  
post #13 of 17 (permalink) Old 06-17-2019, 10:48 AM
AMD Overclocker
 
Alastair's Avatar
 
Join Date: Mar 2012
Location: Benoni (South Africa)
Posts: 6,021
Rep: 259 (Unique: 162)
Quote: Originally Posted by F3ERS 2 ASH3S View Post
Intel will only put a cover on it.. mainly because the full root cause is a lot of the cores initial arch.. see below



correct the security should have been fixed but they where making too much money and that wasn't a focus



see below



For the most part this.. it isn't a leg up unless you are referring to PR

(ps ya missed me? well im back looks like I has no more reps anymores though le sigh lol)



This.. the cores issues are a result in security flaw from the derived arch.. with that being said that it was been improved on and the flaws are now mitigated through other security software patches however on a hardware level the only "patch" that can be provided would be a driver patch reroute etc the handling of threads.. the issue is that Intel isn't going to go full circle on it because that would cause a huge overhead that would degrade performance.. so what you will see is a way to make it look better in publics eyes as they tried to fix it.. same old Intel and thats how business is.. they made far more money doing wrong than was worth spending to do right and if they have to pay for it.. its pennys on the dollar for them.
Hey man good to see ya!









Back to the conversation at hand. I think with Core's design at the core still being heavily influenced by P6 is one of the reasons they have so many security flaws. Because they jumped straight back to P6 after netburst failed hard. So I am guessing a lot of these flaws have to do with P6's aging foundations within the core uArch.

If You Can Cool It! You Can Clock It! (AMD's Vishera)
GHOST rev 3.5
(29 items)
Rura Penthe
(12 items)
CPU
AMD FX-8370 @ 4.95GHz 1.5V 2700MHz NB/3000MHz HTT
Motherboard
ASUS M5A99FX PRO R2.0
GPU
Sapphire R9-Fury 3840
GPU
Sapphire R9-Fury 3840
RAM
Corsair Vengeance 2133 4x4GB @ 2000 9-9-10-27 1CR
Hard Drive
Samsung 850 PRO 1TB
Hard Drive
WD Blue 500GB
Hard Drive
WD Blue 1 TB
Hard Drive
Samsung Spinpoint HD502HI
Power Supply
Antec High Current Pro Platinum 1300W
Cooling
XSPC D5 Bay Res
Cooling
1x CoolerMaster Storm Force 200
Cooling
2x EK-FC Fury X fullcovers
Cooling
XSPC Raystorm CPU Block
Cooling
EKWB Coolstream PE360
Cooling
6x CoolerMaster Jetflo's 120mm
Cooling
Laing D5 Vario
Cooling
EK-FC Terminal Dual Parallel
Cooling
2x EK-FC Fury X Backplates
Cooling
EKWB Coolstream CE280
Cooling
6 Corsair ML140's
Case
NZXT Phantom 820 Black
Operating System
Windows 10 Pro 64-Bit
Monitor
Dell SE2717H
Keyboard
Roccat Ryos MK Pro
Mouse
Mionix Naos 7000
Mouse
Mionix Alioth
Audio
GX Gaming SW-G2.1 3000
Audio
Kingston Hyper X Cloud Core
CPU
AMD Athlon 860K
Motherboard
Asus A88X-Pro
GPU
MSI R6850 Cyclone Power Edition
GPU
MSI R6850 Cyclone Power Edition
RAM
G-Skill Ares 2133 CL10 8GB (2x4GB)
Hard Drive
Seagate ST1000VX003
Optical Drive
Samsung HD502HI
Power Supply
Seagate M12II 620W
Cooling
Arctic Cooling Freezer A30
Case
Corsair Carbide SPEC-02
Operating System
Windows 10 Home Edition 64bit
Keyboard
Razer Lycrosa
CPU
Intel Celeron G1840
Motherboard
MSI Z97-Guard Pro
GPU
Gigabyte RX580 Gaming 4G
GPU
Gigabyte RX570 Gaming 4G
GPU
XFX RX480 reference edition 8G
GPU
Gigabyte RX570 Gaming 4g
GPU
Gigabyte RX570 Gaming 4G
RAM
G-Skill Ares F3-1600C9D
Hard Drive
Samsung 750 Evo 120GB
Power Supply
Gigabyte XP1200M
Cooling
Stock
Operating System
Windows 10 Pro
▲ hide details ▲
Alastair is offline  
Sponsored Links
Advertisement
 
post #14 of 17 (permalink) Old 06-17-2019, 02:22 PM
New to Overclock.net
 
F3ERS 2 ASH3S's Avatar
 
Join Date: Oct 2012
Location: Austin, TX
Posts: 4,375
Rep: 132 (Unique: 89)
Quote: Originally Posted by Alastair View Post
Hey man good to see ya!









Back to the conversation at hand. I think with Core's design at the core still being heavily influenced by P6 is one of the reasons they have so many security flaws. Because they jumped straight back to P6 after netburst failed hard. So I am guessing a lot of these flaws have to do with P6's aging foundations within the core uArch.
Now if Intel already knew that this was an issue... do you think that they have a hidden uarch that they will bring out? expecially since AMD is now got the security and pushing the performance?

Warning: Spoiler! (Click to show)
Asus sabertooth rev 2.0 BIOS 1503
Coolermaster Haf Stacker
FX 8350 @5.1 @1.7v :O
XSPC Raystorm RS240 + Swiftec D5 pump and 360 Rad
2x AMD XFX 280x DD @ 1210/1700
Rosewill Capstone 750w PSU


post-flame-small.gif5 GHz Overclock Club post-flame-small.gif
F3ERS 2 ASH3S is offline  
post #15 of 17 (permalink) Old 06-19-2019, 03:22 PM
New to Overclock.net
 
doritos93's Avatar
 
Join Date: Oct 2009
Location: Montreal
Posts: 2,303
Rep: 119 (Unique: 84)
Quote: Originally Posted by F3ERS 2 ASH3S View Post
Now if Intel already knew that this was an issue...
Every other chip maker on the planet knew it was an issue and safeguarded against it

There is zero doubt that Intel knew about this and let it be for the sake of performance



doritos93 is offline  
post #16 of 17 (permalink) Old 06-19-2019, 03:49 PM
⤷ αC
 
AlphaC's Avatar
 
Join Date: Sep 2012
Posts: 10,715
Rep: 874 (Unique: 577)
https://www.pugetsystems.com/labs/ar...yCapture-1461/
A single patch, to address a couple of CPU vulnerabilities, caused Intel's Core i9 9900K processor to take 2 to 11% longer to process image sets in these photogrammetry applications.

► Recommended GPU Projects: [email protected] , [email protected] (FP64) (AMD moreso) ► Other notable GPU projects: [email protected] (Nvidia), GPUGrid (Nvidia) ► Project list


AlphaC is online now  
post #17 of 17 (permalink) Old 06-19-2019, 04:28 PM
Not a linux lobbyist
 
rluker5's Avatar
 
Join Date: Feb 2014
Location: Wisconsin
Posts: 1,643
Rep: 43 (Unique: 34)
Quote: Originally Posted by AlphaC View Post
https://www.pugetsystems.com/labs/ar...yCapture-1461/
A single patch, to address a couple of CPU vulnerabilities, caused Intel's Core i9 9900K processor to take 2 to 11% longer to process image sets in these photogrammetry applications.
That's already outdated. It is tested on 1809 and I doubt the pervious security patch has been disabled to allow for the version 1903 retpoline mitigation to minimize performance loss.

The whole spectre mitigation being done by intel and windows is really a mess. Once I had the mitigation update double installed and it doubled my performance penalty.

I've found the retpoline mitigation still has a significant performance penalty on random crystaldiskmark values, but it is a fraction of the penalty Windows implemented in 1809. At least on my Haswell system I tested it on, which performs differently from Skylake from what I hear.

There is a penalty, but the mitigation implementation seems too inconsistent across cpu performance reviewers to know how big.

L5
(18 items)
Lea2
(11 items)
L7
(11 items)
CPU
5950hq
Motherboard
z97 Classified
GPU
Aorus 1080ti Waterforce
RAM
16 G Gskill Trident @ 2400,cas10,1.575v
RAM
16 G Team Extreme @ 2400,cas10,1.575v
Hard Drive
2xSamsung 840 EVO 250G
Hard Drive
seagate barracuda 3T
Hard Drive
Optane 900p 480G OS
Optical Drive
Asus BW-16D1HT
Power Supply
EVGA Supernova 1300 G2
Cooling
Cooler Master MasterLiquid Pro 120 (cpu)
Cooling
2 140mm case fans, 2 120mm
Case
Fractal Design R4 (no window)
Operating System
W10 64 pro
Monitor
panasonic TC-58AX800U
Audio
Focal Elear
Audio
SoundbasterX AE-5
Other
Megatron
CPU
4770k
Motherboard
Asus Z87 Deluxe
GPU
Fury Nitro
RAM
8Gb patriot 1600mhz
Hard Drive
ROG Raidr 240Gb pcie
Hard Drive
1Tb WD blue
Power Supply
Pc Power&Cooling silencer Mk2 950w
Cooling
Deepcool Lucifer V2
Case
DIYPC P48-W
Operating System
W10 64 pro
Monitor
40"tv
CPU
4980hq
Motherboard
Asus H81T/CSM
RAM
16GB 1600 generic
Hard Drive
Samsung 850 evo 120gb
Power Supply
Skyvast 90w brick for hp pavilion something
Cooling
SilverStone Tek Super Slim
Case
SilverStone Tek PT13B
Operating System
W10 64 pro
Monitor
50" samsung plasma 720p
Keyboard
Logitech K400+
Other
Intel wifi ac card and noname antennas
▲ hide details ▲
rluker5 is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off