Contribute
Register

Speed Xeon E5-26xx v3

Joined
Feb 14, 2013
Messages
117
Motherboard
Asus ROG Strix x99 Gaming
CPU
E5-2696v3
Graphics
RX 480
Mac
  1. MacBook Pro
  2. Mac Pro
Mobile Phone
  1. iOS
is there someone with a X99 in combintaion with E5-26xx V3?
I have 2 machines one with a E5-2683 v3 and one with a E5-2696 v3 and both have trouble with speed.
Geekbench gives me 20000 and 27000 of multi core score's. However on Geekbench score's i see people getting 40.000 on the E5-2696 v3.
I tried al XCPM patches and without.
Not getting higher values.
Anybody can help?
 
Joined
Jun 9, 2013
Messages
10,781
Motherboard
Gigabyte GA-H77N-WIFI
CPU
i7-3770
Graphics
HD 4000
Mac
  1. MacBook
  2. Mac mini
Mobile Phone
  1. Android
is there someone with a X99 in combintaion with E5-26xx V3?
I have 2 machines one with a E5-2683 v3 and one with a E5-2696 v3 and both have trouble with speed.
Geekbench gives me 20000 and 27000 of multi core score's. However on Geekbench score's i see people getting 40.000 on the E5-2696 v3.
I tried al XCPM patches and without.
Not getting higher values.
Anybody can help?

Have you taken a look at https://www.tonymacx86.com/threads/...broadwell-ep-haswell-ep-cpus-on-10-12.204468/
 
Joined
Feb 14, 2013
Messages
117
Motherboard
Asus ROG Strix x99 Gaming
CPU
E5-2696v3
Graphics
RX 480
Mac
  1. MacBook Pro
  2. Mac Pro
Mobile Phone
  1. iOS
Yes this is one of the things i tried but i get better results without xcpm but also sometimes my machine freezes.
On my i7 4770 i had better benchmarks on osx then on windows now with the xeon windows gives me 15%-20% more speed.
I hoped one of the high scored people is on Tonymacx86...
For both systems i have a other brand motherboard and tried also to switch the processors but it does not matter... (have not tried hs yet...)
 
Joined
Jun 9, 2013
Messages
10,781
Motherboard
Gigabyte GA-H77N-WIFI
CPU
i7-3770
Graphics
HD 4000
Mac
  1. MacBook
  2. Mac mini
Mobile Phone
  1. Android
Yes this is one of the things i tried but i get better results without xcpm but also sometimes my machine freezes.
On my i7 4770 i had better benchmarks on osx then on windows now with the xeon windows gives me 15%-20% more speed.
I hoped one of the high scored people is on Tonymacx86...
For both systems i have a other brand motherboard and tried also to switch the processors but it does not matter... (have not tried hs yet...)

You will get more performance with the i7 4770. A Genuine Mac has had that CPU. Without Native Powermanagement implemented you will get poor performance you will need to match the hardware to a SMBIOS.
 
Joined
Feb 14, 2013
Messages
117
Motherboard
Asus ROG Strix x99 Gaming
CPU
E5-2696v3
Graphics
RX 480
Mac
  1. MacBook Pro
  2. Mac Pro
Mobile Phone
  1. iOS
I just ran this on Linux using a 2683 v3 on an X99 Taichi if you need some comparison. Note that Geekbench 4.1 seems a little higher score than 4.0

http://browser.geekbench.com/v4/cpu/3717593
how?? Do you use the bios mod for the microcode hack?
I even not getting that on my E5-2696v3 (which is a 18 core's 2,3Ghz turbo to 3,8Ghz...) man that is high! I am doing something wrong... maybe i have to set something in bios?? i now running with Load Optimized settings getting 27000 multicore really disapointing. Also if i look at the CPU monitor it never goes into reds keeps half way when running geekbench
 
Joined
Jun 9, 2013
Messages
10,781
Motherboard
Gigabyte GA-H77N-WIFI
CPU
i7-3770
Graphics
HD 4000
Mac
  1. MacBook
  2. Mac mini
Mobile Phone
  1. Android
how?? I even not getting that on my E5-2696v3 (which is a 18 core's 2,3Ghz turbo to 3,8Ghz...) man that is high! I am doing something wrong... maybe i have to set something in bios?? i now running with Load Optimized settings getting 27000 multicore really disapointing. Also if i look at the CPU monitor it never goes into reds keeps half way when running geekbench

The Link shows its in Linux not macOS.
 
Joined
Aug 16, 2017
Messages
2
Yes, modified BIOS to run all cores at 30x except AVX2 which drops one bin to all cores at 29x. Would like a little more even but held back by the manufacturers constraints.

The gain should only be 20% to 25% compared to stock so something seems amiss unless you are running the 32-bit bench instead of the 64-bit one.
 
Top