Overclock.net › Forums › AMD › AMD CPUs › Linux bulldozer Benchmarks Trickling In "update with more benchmarks 10-20-2011"
New Posts  All Forums:Forum Nav:

Linux bulldozer Benchmarks Trickling In "update with more benchmarks 10-20-2011" - Page 6

post #51 of 58
Quote:
Originally Posted by SCollins View Post
Here are results of fx4100 vrs A8 quad core and a few modern intel 4core cpus 2400 2500 etc.

This one should drive the IPC crowd crazy with the results. LOL.

http://openbenchmarking.org/result/1...LI-1110171AR89
fx-4100 didn't do half bad in these test, in most, better than the fx-8150, but that apu that was in there, was doing quite well as well, maybe there is something to die shrinking the thuban to 32nm or coming out with a octo-core phenom II.
Bazinga Punk
(12 items)
 
ooh shiny!
(6 items)
 
 
CPUMotherboardGraphicsRAM
Intel Xeon 3440 AsRock P55 extreme Evga 8800 GT 512 MB Gskill Ripjaws 
Hard DriveCoolingOSMonitor
Western Digital Blue Antec Khuler 620 Ubuntu 11.10 Asus vw264H 
KeyboardPowerCaseMouse
GIGABYTE KM7600 CORSAIR TX 650 Cooler Master 590 GIGABYTE GM-M6800 
CPUMotherboardGraphicsRAM
Intel Core I5 6500 Gigabyte z170xp-SLI Nvidia 970gtx Corsair 16gb ddr4 2666mhz  
Hard DriveOS
250gb Samsung Evo 850 Windows 10 & Ubuntu 15.10 
  hide details  
Reply
Bazinga Punk
(12 items)
 
ooh shiny!
(6 items)
 
 
CPUMotherboardGraphicsRAM
Intel Xeon 3440 AsRock P55 extreme Evga 8800 GT 512 MB Gskill Ripjaws 
Hard DriveCoolingOSMonitor
Western Digital Blue Antec Khuler 620 Ubuntu 11.10 Asus vw264H 
KeyboardPowerCaseMouse
GIGABYTE KM7600 CORSAIR TX 650 Cooler Master 590 GIGABYTE GM-M6800 
CPUMotherboardGraphicsRAM
Intel Core I5 6500 Gigabyte z170xp-SLI Nvidia 970gtx Corsair 16gb ddr4 2666mhz  
Hard DriveOS
250gb Samsung Evo 850 Windows 10 & Ubuntu 15.10 
  hide details  
Reply
post #52 of 58
Thread Starter 
Quote:
Originally Posted by transhour View Post
fx-4100 didn't do half bad in these test, in most, better than the fx-8150, but that apu that was in there, was doing quite well as well, maybe there is something to die shrinking the thuban to 32nm or coming out with a octo-core phenom II.
that should be the first clue, that something ain't right with the software etc. My suspicion is that the data for various threads is being spread all over the cpu caches and its making the cpu waste lots of cycles moving, colating etc the data to and from system memory/l3 l2 etc etc etc.

So changes to the scheduler design of windows and linux could pay huge performance dividends.

but that remains to be seen still.

We shall have to wait for a bulldozer optimized scheduler to see what comes of this long term.

die shrunk phenom2 is here, fm1 socket athalons are available right now. per your request.
post #53 of 58
Quote:
Originally Posted by Demonkev666 View Post
I was thinking the same thing about the L1
i mean everyone keep saying oh it only reaches the x6but yet the X6 has384Kbytes of L1 cache vs bulldozers 256Kbytes.

meh the L3 cache is most likely be only 15-20 watts of the power draw and it really only increase preformance by about 1-2%:1000mhz/1mbs
Isn't that one of the first Piledriver improvement rumors to trickle in, increasing the L1?
Everyday PC
(13 items)
 
  
CPUMotherboardGraphicsRAM
i5-3470 (+4 Turbo bins) Biostar T77XE3 Radeon 7950 16GB DDR3 1600 
Hard DriveOSMonitorKeyboard
Samsung F3 1TB Windows 7 Pro x64 Samsung 24" TN MS Natural Elite 4000 
PowerCaseMouse
Antec HPG 900W Lian-li K57W Logitech G500 
  hide details  
Reply
Everyday PC
(13 items)
 
  
CPUMotherboardGraphicsRAM
i5-3470 (+4 Turbo bins) Biostar T77XE3 Radeon 7950 16GB DDR3 1600 
Hard DriveOSMonitorKeyboard
Samsung F3 1TB Windows 7 Pro x64 Samsung 24" TN MS Natural Elite 4000 
PowerCaseMouse
Antec HPG 900W Lian-li K57W Logitech G500 
  hide details  
Reply
post #54 of 58
Quote:
Originally Posted by SCollins View Post
that should be the first clue, that something ain't right with the software etc. My suspicion is that the data for various threads is being spread all over the cpu caches and its making the cpu waste lots of cycles moving, colating etc the data to and from system memory/l3 l2 etc etc etc.

So changes to the scheduler design of windows and linux could pay huge performance dividends.

but that remains to be seen still.

We shall have to wait for a bulldozer optimized scheduler to see what comes of this long term.

die shrunk phenom2 is here, fm1 socket athalons are available right now. per your request.
yeah that is what i was thinking. cause it comes very close to the 2500k performance in these test, for almost a $100 less.

i know the fusion/apu series is a phenom II architecture die shrunk, there was a petition i seen the other day here or somewhere else, people are wanting thuban to be die shrunk as well. with the results of the apu i saw on these test, it would be interesting to see a die shrunk thuban.

i will eagerly await a kernel scheduler that is tweaked for BD, cause these results so far have been the most interesting.
Bazinga Punk
(12 items)
 
ooh shiny!
(6 items)
 
 
CPUMotherboardGraphicsRAM
Intel Xeon 3440 AsRock P55 extreme Evga 8800 GT 512 MB Gskill Ripjaws 
Hard DriveCoolingOSMonitor
Western Digital Blue Antec Khuler 620 Ubuntu 11.10 Asus vw264H 
KeyboardPowerCaseMouse
GIGABYTE KM7600 CORSAIR TX 650 Cooler Master 590 GIGABYTE GM-M6800 
CPUMotherboardGraphicsRAM
Intel Core I5 6500 Gigabyte z170xp-SLI Nvidia 970gtx Corsair 16gb ddr4 2666mhz  
Hard DriveOS
250gb Samsung Evo 850 Windows 10 & Ubuntu 15.10 
  hide details  
Reply
Bazinga Punk
(12 items)
 
ooh shiny!
(6 items)
 
 
CPUMotherboardGraphicsRAM
Intel Xeon 3440 AsRock P55 extreme Evga 8800 GT 512 MB Gskill Ripjaws 
Hard DriveCoolingOSMonitor
Western Digital Blue Antec Khuler 620 Ubuntu 11.10 Asus vw264H 
KeyboardPowerCaseMouse
GIGABYTE KM7600 CORSAIR TX 650 Cooler Master 590 GIGABYTE GM-M6800 
CPUMotherboardGraphicsRAM
Intel Core I5 6500 Gigabyte z170xp-SLI Nvidia 970gtx Corsair 16gb ddr4 2666mhz  
Hard DriveOS
250gb Samsung Evo 850 Windows 10 & Ubuntu 15.10 
  hide details  
Reply
post #55 of 58
Quote:
Originally Posted by transhour View Post
yeah that is what i was thinking. cause it comes very close to the 2500k performance in these test, for almost a $100 less.

i know the fusion/apu series is a phenom II architecture die shrunk, there was a petition i seen the other day here or somewhere else, people are wanting thuban to be die shrunk as well. with the results of the apu i saw on these test, it would be interesting to see a die shrunk thuban.

i will eagerly await a kernel scheduler that is tweaked for BD, cause these results so far have been the most interesting.
Im pretty sure i read something about amd ending the phenom line.
Main Rig
(12 items)
 
  
CPUMotherboardGraphicsRAM
R7 1800x ASUS Prime 370 Pro RX 480 Corsair Dominator 
Hard DriveOptical DriveOSMonitor
Samsung 850 Pro  None Windows 10 Pro HP w19e: 1440x900 | Sony KDL40S-5100 1920x1080 
KeyboardPowerCaseMouse
CMStorm Quick Fire. CM 1050w  Cooler Master Cosmos II Roccat Kone XTD (Optical) 
  hide details  
Reply
Main Rig
(12 items)
 
  
CPUMotherboardGraphicsRAM
R7 1800x ASUS Prime 370 Pro RX 480 Corsair Dominator 
Hard DriveOptical DriveOSMonitor
Samsung 850 Pro  None Windows 10 Pro HP w19e: 1440x900 | Sony KDL40S-5100 1920x1080 
KeyboardPowerCaseMouse
CMStorm Quick Fire. CM 1050w  Cooler Master Cosmos II Roccat Kone XTD (Optical) 
  hide details  
Reply
post #56 of 58
Quote:
Originally Posted by Vesku View Post
Isn't that one of the first Piledriver improvement rumors to trickle in, increasing the L1?
that was the TLB entries going form 32 to 64
    
CPUMotherboardGraphicsRAM
amd Phenom II x6 1090T gigabye UD7 990FX 5870 G.skill flare 2 x 4gbs 2000mhz  
Hard DriveCoolingOSMonitor
westerdigital cooler master eisberg 240L Vista 64 bit spceptre 1920 x 1200 
KeyboardPowerCaseMouse
muli-media ftw lol 1200 watt silverstone none another cheap one $20 
Mouse PadOther
none ATi 650 pro theater  
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
amd Phenom II x6 1090T gigabye UD7 990FX 5870 G.skill flare 2 x 4gbs 2000mhz  
Hard DriveCoolingOSMonitor
westerdigital cooler master eisberg 240L Vista 64 bit spceptre 1920 x 1200 
KeyboardPowerCaseMouse
muli-media ftw lol 1200 watt silverstone none another cheap one $20 
Mouse PadOther
none ATi 650 pro theater  
  hide details  
Reply
post #57 of 58
OK, so if I read those Linux based benches right, the 8150 wins about half the time, the 2500K wins about a third of the time with two other chips splitting the change...some of which had obviously fixed framerates and would've once again thrown the win to the more powerful chips.

It's interesting what happens when not testing against chips optimized to run software built on their proprietary compiler and vice versa...I think this is a more 'Even Steven' comparo.
post #58 of 58
Thread Starter 
First round of testing with the Kernel patch

http://openbenchmarking.org/result/1...AR-BULLDOZER41
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD CPUs
  • Linux bulldozer Benchmarks Trickling In "update with more benchmarks 10-20-2011"
Overclock.net › Forums › AMD › AMD CPUs › Linux bulldozer Benchmarks Trickling In "update with more benchmarks 10-20-2011"