Asus X299 - Support

[izo1]

Well, I do not know if it will be false or not, the truth is that the 9980xe has not been what was expected, the general scores of the 9980xe in mac do not approach those of the 7980xe, that is not false ... I do not speak of geekbench 5 I don't like it either. Hopefully you are right and the 10980xe is much better than its successor, with 5-10% it is not enough, because that is the performance of the 7980xe, the price if I like it, now that I have bios 2002, if the results are good I will buy one, although the problem of these processors with so many cores is that, for basic applications, Photoshop, After FX and others, they do not go very fast, a 9900k of € 456 overcomes them. If you use 3D, Cinema 4D, Lighwave or applications that use all cores, then yes, but most applications are not optimized, sadly, for use with these processors, the attached files are not false, I assure you ...

The thing is applications such as After Effects and Photoshop and even Premiere Pro thrive on high speed single core performance. This is why you sometimes see the 9900k win in certain scenarios.

AE for example has horrible multi-processor support, but you can use a background renderer to max out all of your CPU cores.

The 10980XE is not going to be that much better than the 9980XE, especially if you overclock the latter.

I'm glad they dropped the price in half, but Intel seriously needs to move to 7nm asap.

By no means is the 18 core Intel a slouch. I personally love this setup especially after overclocking.

Multicore processors is the future and app developers will be forced to optimize it one way or another.

 

[izo1]

Hello everyones and especially X299 SAGE 10G users.
What did you do after EEPROM modding to make tech 10g ethernet ports work ?
Thanks in advance.

I used the original kgp thread to flash under Ubuntu. It worked perfectly for both ports.

iMac Pro X299 - Live the Future now with macOS 10.14 Mojave [Successful Build/Extended Guide]

Up and successfully running macOS Mojave 10.14.4 (18E226) on iMacPro Skylake-X/X299 This afternoon @DSM2 advised me that yesterday Apple seems to have added a second special build of MacOS 10.14.4 (18E2034) on swcdn.apple.com, likely for new devices. Abstract and Introduction: This...

www.tonymacx86.com

Section E.8.2.3)

Do it carefully and it will be ok.

 

[sambirch19]

Maybe he needed something now, without waiting for official release of new 10XXX series
We know it's not a great deal to buy older CPUs today, but when you have work to do...
sometimes, some people will earn more money in few days than cpu price difference...
and time to send back some project or something to client worth more than 2 new cores and 10% more power etc
(it's just an example of many reasons - why, we simply can't wait, even if it's not good choice)

My partner needed a computer "yesterday" per se and I got it for well under 50% of RRP

 

[oli.mathieu]

I used the original kgp thread to flash under Ubuntu. It worked perfectly for both ports.

iMac Pro X299 - Live the Future now with macOS 10.14 Mojave [Successful Build/Extended Guide]

Up and successfully running macOS Mojave 10.14.4 (18E226) on iMacPro Skylake-X/X299 This afternoon @DSM2 advised me that yesterday Apple seems to have added a second special build of MacOS 10.14.4 (18E2034) on swcdn.apple.com, likely for new devices. Abstract and Introduction: This...

www.tonymacx86.com

Section E.8.2.3)

Do it carefully and it will be ok.

thanks
That what I did .... I feel dumb.
I give another shot !!!

 

[oli.mathieu]

You need to change the Sub-System ID not the Sub-Vendor ID. Offsets are next word, 0x242 and 0x243. The order of Sub-System ID and the Sub-Vendor ID are switched for X520 and X550.

Thank you
I will do it again and pay more attention.

 

[mgregrs]

I deleted last prepared post because @iso wrote faster exactly what I wanted to say

So @oli.mathieu - believe me - at the beginning first moths, every time I read KGP guide and posts I felt dumb,
sometimes ... even x2...
- and even today, sometime when I have some thing more complecated I feel like that, until moment when I understand how it works etc ... of core it's 300% less problems than before, but it's nothing bad... it only need more research of info/examples, tests etc

 

[oli.mathieu]

I deleted last prepared post because @iso wrote faster exactly what I wanted to say

So @oli.mathieu - believe me - at the beginning first moths, every time I read KGP guide and posts I felt dumb,
sometimes ... even x2...
- and even today, sometime when I have some thing more complecated I feel like that, until moment when I understand how it works etc ... of core it's 300% less problems than before, but it's nothing bad... it only need more research of info/examples, tests etc

Okay I gave another shot
lspci -nn -vvv | grep Ethernet :

I read 8086: 1563
so I'm looking for 63 15 86 80 with
sudo ethtool -e enp226s0f0 | less :

I only find 86 80. I think it's an issue may be i'm wrong.
Nothing close to what I'm looking for at 0x242 and 0x243.
Further down there no 63 15 86 80.
Thanks for any support

 

[ydeng]

Okay I gave another shot
lspci -nn -vvv | grep Ethernet :View attachment 431744
I read 8086: 1563
so I'm looking for 63 15 86 80 with
sudo ethtool -e enp226s0f0 | less :View attachment 431745
I only find 86 80. I think it's an issue may be i'm wrong.
Nothing close to what I'm looking for at 0x242 and 0x243.
Further down there no 63 15 86 80.
Thanks for any support

0x260, 0x261 is the PCI vendor id, 0x8086, you don't want to change that. Because you already changed offset 0x240, 0x241 to 0x0a00, it indeed shows 0x0a00. This is correct, Intel data sheet says SubSystem and PCI vendor ids are 0x20 bytes apart. If you really want to be sure, you may read the word at 0x12, that is the word address pointing to the start of PCIe general configuration. The SubSystem is at 0x11 word address (0x22 + value at 0x12 should be 0x242). Same as X520 case https://www.tonymacx86.com/threads/...mall-tree-drivers.156135/page-16#post-1364958 .

Look at this https://www.tonymacx86.com/threads/...mall-tree-drivers.156135/page-28#post-1864722: you need to change 0x1287 to 0x0a00, not 0x4310. Device id is loaded from a different place not together with vendor id (0x8680).

 

[oli.mathieu]

0x260, 0x261 is the PCI vendor id, 0x8086, you don't want to change that. Because you already changed offset 0x240, 0x241 to 0x0a00, it indeed shows 0x0a00. This is correct, Intel data sheet says SubSystem and PCI vendor ids are 0x20 bytes apart. If you really want to be sure, you may read the word at 0x12, that is the word address pointing to the start of PCIe general configuration. The SubSystem is at 0x11 word address (0x22 + value at 0x12 should be 0x242). Same as X520 case https://www.tonymacx86.com/threads/...mall-tree-drivers.156135/page-16#post-1364958 .

Look at this https://www.tonymacx86.com/threads/...mall-tree-drivers.156135/page-28#post-1864722: you need to change 0x1287 to 0x0a00, not 0x4310. Device id is loaded from a different place not together with vendor id (0x8680).

I've just read several times your reply. it's hard to understand. But i think I get it .Thanks
I know now how to change 0x1287 to 0x0a00.
Do I need to change back the word before (0x0a00 to 0x4310) ?
Why "lspci -nn -vvv | grep Ethernet" doesn't show any subsytem ID ?
Thanks again

 

[ydeng]

I've just read several times your reply. it's hard to understand. But i think I get it .Thanks
I know now how to change 0x1287 to 0x0a00.
Do I need to change back the word before (0x0a00 to 0x4310) ?
Why "lspci -nn -vvv | grep Ethernet" doesn't show any subsytem ID ?
Thanks again

0x4310 is the Sub-vendor id. The driver does not use it. Change it back let lspci show correct vendor name.

Subsystem is on the next line:

"lspci -nn -vv|less" and type "/ Ether" to search for the card:

09:00.1 Ethernet controller [0200]: Intel Corporation Ethernet Connection X553/X557-AT 10GBASE-T [8086:15c8] (rev 11)
Subsystem: Super Micro Computer Inc Ethernet Connection X553/X557-AT 10GBASE-T [15d9:15c8]