System Light Dark

xentr_theme_editor

xentr_page_setup xentr_appearance xentr_sidebar

  • xentr_page_width

    xentr_toggle_page_width

    xentr_page_width_description

    xentr_color_pickers

    xentr_toggle_color_picker

    xentr_toggle_color_picker_description

    xentr_typography

    xentr_select_the_font_you_want_to_use

    xentr_node_layout

    xentr_grid_layout

    xentr_grid_layout_description

  • Style variation

    xentr_select_the_style_variation_you_want_to_use
    System Light Dark

  • xentr_sidebar

    xentr_toggles_sidebar

    xentr_sidebar_description

    xentr_sticky_sidebar

    xentr_sticky_sidebar_description
    xentr_sidebar_position
    Left
    Right
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

Quad processor G34 build

BIOS settings are same for CPU, here are the others. Linux is also set up per the Musky Mix, TheKraken is installed, not Langouste, and -bigadv -smp 32 are the flags. I don't know if my Hyper Transport settings are fine.

7034f0f96b33fe0a.jpg


7034f0f96b341b80.jpg


7034f0f96b34342a.jpg
 
xentr_thread_starter
3.0charlie said:
NUMA enabled. Gotta love IPMI.
One more S8812 in the House. Folding for 54196 I hope? My 4p (see sig) hovers around 247k ppd on a 6903. Here's mine.
Click to expand...
3.0charlie said:
HOW MUCH?!? My rig sits at 223k ppd for a 6903...
Click to expand...

If you say so. My bios settings are the same as yours but I hadn't realised I could change the settings under ECC Config with standard DDR3. Mine was set to [Basic] so I changed it to [Disabled] like you have because that sounds like the faster option.
My flags are -smp -bigadv -bigbeta but I don't think these matter as long as the client gets a 6903.

My tpf on the current unit seems to be about 22:45 on average so "245k ppd" according to the calc. My first ppd number of 335k is from when I had PowerNow [Enabled]. It seams we're even so that's great.

EDIT: I'm getting 22:54 tpf on the same unit as before so I'm switching ECC back on. It could be something else like dynamic load balancing but I think ECC [Basic] was a bit faster than [Disabled]. I'm trying the [Good] setting now just for fun.
 
Last edited:
Zink said:
If you say so. My bios settings are the same as yours but I hadn't realised I could change the settings under ECC Config with standard DDR3. Mine was set to [Basic] so I changed it to [Disabled] like you have because that sounds like the faster option.
My flags are -smp -bigadv -bigbeta but I don't think these matter as long as the client gets a 6903.

My tpf on the current unit seems to be about 22:45 on average so "245k ppd" according to the calc. My first ppd number of 335k is from when I had PowerNow [Enabled]. It seams we're even so that's great.

EDIT: I'm getting 22:54 tpf on the same unit as before so I'm switching ECC back on. It could be something else like dynamic load balancing but I think ECC [Basic] was a bit faster than [Disabled]. I'm trying the [Good] setting now just for fun.
Click to expand...
Keep us on the loop, I always had ECC disabled. But Powernow, that's interesting. Could it be the 'Turbo' kicking in? As far as tpf and ppd, both rigs are almost the same (22:40, 238k per HFM).

edit: I don't think Powernow makes any difference - see AMD PowerNow!. My question is, why add this feature on a 4p server board when the article talks about notebooks and battery power savings? Weird.
 
xentr_thread_starter
3.0charlie said:
Keep us on the loop, I always had ECC disabled. But Powernow, that's interesting. Could it be the 'Turbo' kicking in? As far as tpf and ppd, both rigs are almost the same (22:40, 238k per HFM).

edit: I don't think Powernow makes any difference - see AMD PowerNow!. My question is, why add this feature on a 4p server board when the article talks about notebooks and battery power savings? Weird.
Click to expand...
I think the recommendation is to disable PowerNow to increase performance slightly. PowerNow lets the CPUs drop bellow 2Ghz when they are not fully loaded I would guess and then they take some time to get back to full frequency which impacts performance.
 
xentr_thread_starter
3rd 6903 is almost done. 22:40-22:50 tpf which is 243k ppd before upload and dl times bring it back down to about 230k.
Power use is 160W idle, 470W while working a 6903 = 510ppd/w. Most extras on the board are disabled but IPMI is still on.

Using wondershaper to regulate download bandwidth so it doesn't completely kill the connection for people gaming on my dsl.
Uploads are capped at 75% of the maximum to keep the ping down, done directly in langouste using the rate limiting feature that you an learn about in the readme.
QoS also set to low on the router now, before it was off.

I haven't tested the bandwidth limiting yet but I know without it my connection had very high latency and very slow page loads for the hour it takes to upload 200MB of results.

EDIT: these setting work great during wu uploads, barely any difference in ping or dl speeds. There is little upload bandwidth for other purposes but that is how it is intend.
 
Last edited:
Zink said:
3rd 6903 is almost done. 22:40-22:50 tpf which is 243k ppd before upload and dl times bring it back down to about 230k.
Power use is 160W idle, 470W while working a 6903 = 510ppd/w. Most extras on the board are disabled but IPMI is still on.

Using wondershaper to regulate download bandwidth so it doesn't completely kill the connection for people gaming on my dsl.
Uploads are capped at 75% of the maximum to keep the ping down, done directly in langouste using the rate limiting feature that you an learn about in the readme.
QoS also set to low on the router now, before it was off.

I haven't tested the bandwidth limiting yet but I know without it my connection had very high latency and very slow page loads for the hour it takes to upload 200MB of results.

EDIT: these setting work great during wu uploads, barely any difference in ping or dl speeds. There is little upload bandwidth for other purposes but that is how it is intend.
Click to expand...


does the transfer from fah really affect the link that much?

-joe
 
It's the aggregate time between download and upload that determines the bonus points, so yes, the transfer can have a large impact on the final points value, if you have a really slow upload like me.
 
You must log in or register to reply here.

Latest posts

xentr_legal_notice_title

xentr_legal_notice_description

Back
Top