ROG Crosshair VI overclocking thread - Page 3262 - Overclock.net - An Overclocking Community

Forum Jump: 

ROG Crosshair VI overclocking thread

Reply
 
Thread Tools
post #32611 of 43397 (permalink) Old 01-10-2018, 04:13 PM
New to Overclock.net
 
NotAgain's Avatar
 
Join Date: Feb 2011
Posts: 404
Rep: 35 (Unique: 29)
Quote:
Originally Posted by ressonantia View Post

I have yet to read a credible source that says otherwise, unless you're referring to the change request in the SUSE codebase which suggests so but has since been debunked?
From your link:

"The AMD change-log does note this AMD microcode update is indeed for CVE-2017-5715, a.k.a. SPECTRE."

Their description of the microcode completely disabling branch prediction is wrong. The fact that there is microcode required to mitigate Spectre Variant 2 is not.
It is not critical right now, as AMD's branch predictor has not been reverse-engineered like Intel's has, but there is always that possibility.
NotAgain is offline  
Sponsored Links
Advertisement
 
post #32612 of 43397 (permalink) Old 01-10-2018, 04:26 PM
Linux Lobbyist
 
ressonantia's Avatar
 
Join Date: Oct 2016
Posts: 142
Rep: 7 (Unique: 6)
Quote:
Originally Posted by NotAgain View Post

From your link:

"The AMD change-log does note this AMD microcode update is indeed for CVE-2017-5715, a.k.a. SPECTRE."

Their description of the microcode completely disabling branch prediction is wrong. The fact that there is microcode required to mitigate Spectre Variant 2 is not.
It is not critical right now, as AMD's branch predictor has not been reverse-engineered like Intel's has, but there is always that possibility.

Maybe you know something that the I don't, but I have not yet read anywhere that their branch predictor is vulnerable the same way Intel's is. If you have sources, I would very much like to have a look at them. For all we know (again not saying that I do know), their mitigation might be reducing the "near zero risk to zero". Reading through the whitepaper, Ryzen is mentioned as being vulnerable to Variant 1 but nowhere is it mentioned for Variant 2.
ressonantia is offline  
post #32613 of 43397 (permalink) Old 01-10-2018, 05:28 PM
New to Overclock.net
 
NotAgain's Avatar
 
Join Date: Feb 2011
Posts: 404
Rep: 35 (Unique: 29)
Quote:
Originally Posted by ressonantia View Post

Maybe you know something that the I don't, but I have not yet read anywhere that their branch predictor is vulnerable the same way Intel's is. If you have sources, I would very much like to have a look at them. For all we know (again not saying that I do know), their mitigation might be reducing the "near zero risk to zero". Reading through the whitepaper, Ryzen is mentioned as being vulnerable to Variant 1 but nowhere is it mentioned for Variant 2.
Again: it's not currently vulnerable because their branch predictor has not been reverse-engineered yet.
See here for details on the new microcode that adds Indirect Branch Predictor Barrier (IBPB) instructions to mitigate CVE-2017-5715 (Spectre Variant 2) on Zen (17h) CPUs.

https://launchpad.net/ubuntu/+source/amd64-microcode/3.20171205.1
NotAgain is offline  
Sponsored Links
Advertisement
 
post #32614 of 43397 (permalink) Old 01-10-2018, 06:41 PM
Linux Lobbyist
 
ressonantia's Avatar
 
Join Date: Oct 2016
Posts: 142
Rep: 7 (Unique: 6)
Quote:
Originally Posted by NotAgain View Post

Again: it's not currently vulnerable because their branch predictor has not been reverse-engineered yet.
See here for details on the new microcode that adds Indirect Branch Predictor Barrier (IBPB) instructions to mitigate CVE-2017-5715 (Spectre Variant 2) on Zen (17h) CPUs.

https://launchpad.net/ubuntu/+source/amd64-microcode/3.20171205.1

Well then, I think you have just answered your own question wink.gif
ressonantia is offline  
post #32615 of 43397 (permalink) Old 01-10-2018, 07:34 PM
New to Overclock.net
 
LightningManGTS's Avatar
 
Join Date: Mar 2017
Location: United States
Posts: 247
Rep: 9 (Unique: 8)
Quote:
Originally Posted by dual109 View Post

Is the DRAM calculator still relevant with the 3101 beta bios? I have some M-Die Hynix ram Corsair Vengeance LPX CMK16GX4M2B3200C16 (2x8GB) which I'm trying to get running at rated speeds 3200Mhz, no mem presets available for these in bios.



Thanks

why wouldn't it be? unless there's a microcode update that messes with the mathematics behind sub timming calculation it should always be accurate to the point it normally is regardless of the board, bios, or agesa version. weather or not you have access to specific feature or if the bios is correctly configured to handle certain tweaks on the other hand is a completely different matter and is why you should be using a full release bios and not a beta bios. on that note beta bios 3101 is just a microcode jump from 3008 to agesa 1.0.10 so there is probably something elmor and his team need to do to it to make sure its fully compatible with the bios features they coded in independently to any changes made in agesa versions. both for the current chips now and for pinnacle ridge chips when they are released in April as agesa 1.0.10 is the microcode that adds support for theses processors.

LightningManGTS is offline  
post #32616 of 43397 (permalink) Old 01-10-2018, 07:40 PM
New to Overclock.net
 
janreiviardo's Avatar
 
Join Date: May 2017
Posts: 12
Rep: 0
Hi, is there something wrong with my settings and results? Cause comparing it to Usmus results below, mine is so much worse than his but the settings are almost same. I used Usmus Calculator for and used the 3333 fast setting.

My Sepcs:
Ryzen 1700
Crosshair Hero VI
G.Skill - Trident Z 16GB (2 x 8GB) DDR4-3200 CL14 (F4-3200C14D-16GTZ)

Also I'm just getting around 1650-1660 score in Cinebench15 with CPU overclocked to 3800. What's bugging me is that, when doing the benchmark, cinebench will sometimes hang for like 3 secs but the mouse is moving smoothly. The results with few hangs and without hangs are the same. 1650ish.

In cenebench, just recently my CPU will keep showing it is 3ghz when I am overclocked to 3.8ghz. Same for the Asus software, it says 3ghz but in task manager and CPU-Z it is 3.8.

Expert's help is so gretly appreciated. I'm newbie.



Quote:
Originally Posted by 1usmus View Post

@harrysun

try this smile.gif

3300MHZ



CAD_BUS 20 20 20 20
procODT 60
RTT 7 3 1
PLL 1.9

BCLK 103.0 , PCI work fine in GEN 3 mode

______________________________________

DR 3432

CLDO 1000mv
DRAM R Tune 61
procODT 68.6
CAD_BUS 30 30 40 60

janreiviardo is offline  
post #32617 of 43397 (permalink) Old 01-10-2018, 07:54 PM
New to Overclock.net
 
LightningManGTS's Avatar
 
Join Date: Mar 2017
Location: United States
Posts: 247
Rep: 9 (Unique: 8)
Quote:
Originally Posted by janreiviardo View Post

Hi, is there something wrong with my settings and results? Cause comparing it to Usmus results below, mine is so much worse than his but the settings are almost same. I used Usmus Calculator for and used the 3333 fast setting.

My Sepcs:
Ryzen 1700
Crosshair Hero VI
G.Skill - Trident Z 16GB (2 x 8GB) DDR4-3200 CL14 (F4-3200C14D-16GTZ)

Also I'm just getting around 1650-1660 score in Cinebench15 with CPU overclocked to 3800. What's bugging me is that, when doing the benchmark, cinebench will sometimes hang for like 3 secs but the mouse is moving smoothly. The results with few hangs and without hangs are the same. 1650ish.


1usmus is using a dual rank kit which has extremely different timings and tolerances, use the stilits 2 dimm SR presets as a base and then use 1usmus calculator and figure out where your kit falls between so to speak. at no point will the claculator give you settings you can just run without tweaking other thinks like vpp_mem or the termination so feel free to hit up the advance tab

LightningManGTS is offline  
post #32618 of 43397 (permalink) Old 01-10-2018, 10:10 PM - Thread Starter
New to Overclock.net
 
elmor's Avatar
 
Join Date: Mar 2014
Posts: 835
Rep: 335 (Unique: 147)
New beta BIOS 3501

- Update to latest AGESA with support for 2nd gen Ryzen processors. Version number is reset, this release is 1.0.0.0a.
- Supports additional DRAM ratios which might help you get closer to your max DRAM frequency without using reference clock. New ratios: 2733, 2866, 3000, 3133, 3266, 3400, 3533, 3666, 3800, 3933, 4066, 4133, 4200.
- Still has the S3 resume TSC frequency bug. Enable HPET or don't use sleep mode for now.
- Custom P-state overclocking seems to disable P1/P2 states

http://www.mediafire.com/file/e07ivw2lstf0933/CROSSHAIR-VI-HERO-ASUS-3501.zip (sha256 9a3147f22908139f0150eab07abbeabc1a8040b541592879ad215ec723eac958)
http://www.mediafire.com/file/e2ng10fjf4gu7fj/ROG-CROSSHAIR-VI-HERO-WIFI-AC-ASUS-3501.zip (sha256 c4bce19fd6d6a96b6487983d886a1ed61b6ac77163f1c87c552c7698f3397e1b)
http://www.mediafire.com/file/3j2fjp2e1r2xe8v/ROG-CROSSHAIR-VI-EXTREME-ASUS-3501.zip (sha256 738dbd3e33406845d6f52e1401e62205376254e5d805233bb5a566a95eaeafd4)
elmor is offline  
post #32619 of 43397 (permalink) Old 01-10-2018, 10:22 PM
New to Overclock.net
 
LightningManGTS's Avatar
 
Join Date: Mar 2017
Location: United States
Posts: 247
Rep: 9 (Unique: 8)
Quote:
Originally Posted by elmor View Post

New beta BIOS 3501

- Update to latest AGESA with support for 2nd gen Ryzen processors. Version number is reset, this release is 1.0.0.0a.
- Supports additional DRAM ratios which might help you get closer to your max DRAM frequency without using reference clock. New ratios: 2733, 2866, 3000, 3133, 3266, 3400, 3533, 3666, 3800, 3933, 4066, 4133, 4200.
- Still has the S3 resume TSC frequency bug. Enable HPET or don't use sleep mode for now.
- Custom P-state overclocking seems to disable P1/P2 states

http://www.mediafire.com/file/e07ivw2lstf0933/CROSSHAIR-VI-HERO-ASUS-3501.zip (sha256 9a3147f22908139f0150eab07abbeabc1a8040b541592879ad215ec723eac958)
http://www.mediafire.com/file/e2ng10fjf4gu7fj/ROG-CROSSHAIR-VI-HERO-WIFI-AC-ASUS-3501.zip (sha256 c4bce19fd6d6a96b6487983d886a1ed61b6ac77163f1c87c552c7698f3397e1b)
http://www.mediafire.com/file/3j2fjp2e1r2xe8v/ROG-CROSSHAIR-VI-EXTREME-ASUS-3501.zip (sha256 738dbd3e33406845d6f52e1401e62205376254e5d805233bb5a566a95eaeafd4)

Amazeballs, I honestly was not expecting anything until after ces so this is a nice surprise! I'll take a look at it afterwork. Is this just another microcode update with the ram speeds or was there anything extra done to get everything to play nice?

LightningManGTS is offline  
post #32620 of 43397 (permalink) Old 01-10-2018, 10:53 PM
New to Overclock.net
 
VPII's Avatar
 
Join Date: Dec 2014
Location: Hout Bay, Cape Town, South Africa
Posts: 611
Rep: 8 (Unique: 7)
Quote:
Originally Posted by elmor View Post

New beta BIOS 3501

- Update to latest AGESA with support for 2nd gen Ryzen processors. Version number is reset, this release is 1.0.0.0a.
- Supports additional DRAM ratios which might help you get closer to your max DRAM frequency without using reference clock. New ratios: 2733, 2866, 3000, 3133, 3266, 3400, 3533, 3666, 3800, 3933, 4066, 4133, 4200.
- Still has the S3 resume TSC frequency bug. Enable HPET or don't use sleep mode for now.
- Custom P-state overclocking seems to disable P1/P2 states

http://www.mediafire.com/file/e07ivw2lstf0933/CROSSHAIR-VI-HERO-ASUS-3501.zip (sha256 9a3147f22908139f0150eab07abbeabc1a8040b541592879ad215ec723eac958)
http://www.mediafire.com/file/e2ng10fjf4gu7fj/ROG-CROSSHAIR-VI-HERO-WIFI-AC-ASUS-3501.zip (sha256 c4bce19fd6d6a96b6487983d886a1ed61b6ac77163f1c87c552c7698f3397e1b)
http://www.mediafire.com/file/3j2fjp2e1r2xe8v/ROG-CROSSHAIR-VI-EXTREME-ASUS-3501.zip (sha256 738dbd3e33406845d6f52e1401e62205376254e5d805233bb5a566a95eaeafd4)

Thank you elmor, I'll give it a run now to see how it works.

AMD
(9 items)
CPU
AMD Ryzen 2700X
Motherboard
Asus Crosshair VII Hero
GPU
GALAX RTX 2080 TI OC
RAM
G.Skill f4-3200c14d-16gtzr tridentz rgb
Hard Drive
Crucial MX300 M.2 SSD
Power Supply
Corsair HX1200i
Cooling
CoolerMaster MLW-D24M-A20PC-R1 MasterLiquid ML240L 240mm RGB Liquid CPU Cooler
Operating System
Windows 10 Pro
Monitor
Dell U2717D 27" UltraSharp InfinityEdge LED Monitor
▲ hide details ▲
VPII is offline  
Reply

Tags
Bios5 , profiles to usb

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: 16 (2 members and 14 guests)
gtgambit , Sbb Kbb
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