Overclock.net › Forums › AMD › AMD CPUs › Bulldozer Questions from an Intel User.
New Posts  All Forums:Forum Nav:

Bulldozer Questions from an Intel User. - Page 10

post #91 of 162
I think it would be quite a feat for their SMT design to equal the per core performance of Nehalem.

Not saying it won't happen, just saying its not their focus... If they can pull of Nehalem per core performance, enhance overclocking to at least 4.5Ghz, and provide a 8 core BE for $316 I will be very impressed.
    
CPUMotherboardGraphicsGraphics
Intel Core i5 2500K P8P67 PRO NVIDIA GeForce GTX 470 NVIDIA GeForce GTX 470 
GraphicsRAMRAMRAM
NVIDIA GeForce 9800 GT G-Skill A-Data G-Skill 
RAMHard DriveOptical DriveOS
A-Data Crucial M4 64GB + 1TB F3 Spinpoint $155 LS/DL DVD RW $?? Windows 8 64-bit "Epic Registry" Edition 
MonitorPowerCase
ASUS 21.5 1920x1080 2ms $135 CORSAIR HX850 $120 Mother Earth $free 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel Core i5 2500K P8P67 PRO NVIDIA GeForce GTX 470 NVIDIA GeForce GTX 470 
GraphicsRAMRAMRAM
NVIDIA GeForce 9800 GT G-Skill A-Data G-Skill 
RAMHard DriveOptical DriveOS
A-Data Crucial M4 64GB + 1TB F3 Spinpoint $155 LS/DL DVD RW $?? Windows 8 64-bit "Epic Registry" Edition 
MonitorPowerCase
ASUS 21.5 1920x1080 2ms $135 CORSAIR HX850 $120 Mother Earth $free 
  hide details  
Reply
post #92 of 162
Quote:
Originally Posted by Behemoth777 View Post
hahaha you can't be serious. In starcraft 2, I went from 30fps to 60fps just by switching to this 2500k from an unlocked x4 phenom II. Yes, the 2500k was clocked quite a bit higher, but that doesn't mean that your supposed to rule it out just because my phenom II couldn't clock as high. A performance increase is a performance increase.
SC2 is very CPU dependent as most RTS games are.
Edited by love9sick - 5/25/11 at 8:18pm
    
CPUMotherboardGraphicsRAM
I5 2500k @4.2 P67A-GD80(B3 revision) MSI R6970 MSI 16GB Kingston HyperX 1600mhz 
Hard DriveOSMonitorKeyboard
Seagate 1 Terrible bite Windows 7 64bit Pro 27" inch Acer LED 1920x1080 Logitech G15 
PowerCaseMouseMouse Pad
Corsair TX850watt Corsair T600 Logitech MX desk 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
I5 2500k @4.2 P67A-GD80(B3 revision) MSI R6970 MSI 16GB Kingston HyperX 1600mhz 
Hard DriveOSMonitorKeyboard
Seagate 1 Terrible bite Windows 7 64bit Pro 27" inch Acer LED 1920x1080 Logitech G15 
PowerCaseMouseMouse Pad
Corsair TX850watt Corsair T600 Logitech MX desk 
  hide details  
Reply
post #93 of 162
Quote:
Originally Posted by BallaTheFeared View Post
None of my facts are made up, they come directly from the horses mouth.

I will admit I'm playing connect the dots, since they haven't given out a lot of information. However with the information made available, how CMT works, the loss of performance that comes with it, the current pricing graph we have, AMDs current position as far as performance per core goes and JFs general demeanor when it comes to "per core" performance tells me I'm probably not too far off with my assumptions.


What else are we going to do? Its a forum for computer parts, just don't take anything personally and nobody will get affected over another's opinion.
PLEASE STOP QUOTING ME.

How many times have I said that single threaded performance will be higher on Opteron? Many times. 50% more throughput with only 33% more cores. How can it be lower?

If you want to make a statement about your opinion you are free to do so, but when you drop my name in it you are only doing it to lend credibility to your theories.

And you keep making statements that I am saying things that I am not.

Enough.
post #94 of 162
Quote:
Originally Posted by JF-AMD View Post
PLEASE STOP QUOTING ME.

How many times have I said that single threaded performance will be higher on Opteron? Many times. 50% more throughput with only 33% more cores. How can it be lower?

If you want to make a statement about your opinion you are free to do so, but when you drop my name in it you are only doing it to lend credibility to your theories.

And you keep making statements that I am saying things that I am not.

Enough.
No disrespect or anything but, Can't you release some sort of information on the AMD website stateing what the cores single / multi threaded performance is all about? so people can't quote you on anything anymore?

It would make verything around here alot easier and people woulden't have to make these pointless threads.... soo.... if someone quotes you again with the wrong information, you can quote them back to AMD weblink which will provide the correct performance information and the result is..... no more discredit, and people woulden't have to gain fame anymore for doing so.
The Ancient
(14 items)
 
FZ-G1
(8 items)
 
 
CPUMotherboardGraphicsRAM
Phenom X6 1055T 3.9GHz NB 3.0GHz ASUS Crosshair V  GTX 770 8GB G.Skill TridentX 1652MHz CL7-7-8-20-22 1T 
Hard DriveOptical DriveCoolingOS
OCZ 60GB Vertex 2E - 3x HDD's Samsung Blu-Ray Drive Custom W/C Loop Windows 7 64bit 
PowerCase
Superflower Leadex 650W Coolermaster Storm Scout Mod 
CPUGraphicsRAMHard Drive
Core i5-3437U Intel HD4000 8GB 256GB SSD 
OSMonitorKeyboardCase
Windows 8 Pro 10.1 WUXGA Touchscreen Worlds Durable Tablet 
  hide details  
Reply
The Ancient
(14 items)
 
FZ-G1
(8 items)
 
 
CPUMotherboardGraphicsRAM
Phenom X6 1055T 3.9GHz NB 3.0GHz ASUS Crosshair V  GTX 770 8GB G.Skill TridentX 1652MHz CL7-7-8-20-22 1T 
Hard DriveOptical DriveCoolingOS
OCZ 60GB Vertex 2E - 3x HDD's Samsung Blu-Ray Drive Custom W/C Loop Windows 7 64bit 
PowerCase
Superflower Leadex 650W Coolermaster Storm Scout Mod 
CPUGraphicsRAMHard Drive
Core i5-3437U Intel HD4000 8GB 256GB SSD 
OSMonitorKeyboardCase
Windows 8 Pro 10.1 WUXGA Touchscreen Worlds Durable Tablet 
  hide details  
Reply
post #95 of 162
I have made these statements over and over again. On these forums and on my blogs.

If Balla would just actually quote something that I said properly, using the quote feature and using my actual words I would be fine.

What is bothering me is that Balla keeps saying "JF said..." and then goes on to say something that I have not said.

My name is being invoked to add credibility to statements - statements that I am not making.
post #96 of 162
Quote:
Originally Posted by JF-AMD View Post
I have made these statements over and over again. On these forums and on my blogs.

If Balla would just actually quote something that I said properly, using the quote feature and using my actual words I would be fine.

What is bothering me is that Balla keeps saying "JF said..." and then goes on to say something that I have not said.

My name is being invoked to add credibility to statements - statements that I am not making.
Fair enough, I see your point, but at the end of the day, everyone is just little hyped up about the upcoming new FX processors, I don't blame them, becuase the FX chips looks like a big jump for AMD, so people on here start quoteing misinformation stuff whilst referring other people into theres quote, which they shoulden't unless they quoted the post that you made within the text also.

I'm probably not making much sence
The Ancient
(14 items)
 
FZ-G1
(8 items)
 
 
CPUMotherboardGraphicsRAM
Phenom X6 1055T 3.9GHz NB 3.0GHz ASUS Crosshair V  GTX 770 8GB G.Skill TridentX 1652MHz CL7-7-8-20-22 1T 
Hard DriveOptical DriveCoolingOS
OCZ 60GB Vertex 2E - 3x HDD's Samsung Blu-Ray Drive Custom W/C Loop Windows 7 64bit 
PowerCase
Superflower Leadex 650W Coolermaster Storm Scout Mod 
CPUGraphicsRAMHard Drive
Core i5-3437U Intel HD4000 8GB 256GB SSD 
OSMonitorKeyboardCase
Windows 8 Pro 10.1 WUXGA Touchscreen Worlds Durable Tablet 
  hide details  
Reply
The Ancient
(14 items)
 
FZ-G1
(8 items)
 
 
CPUMotherboardGraphicsRAM
Phenom X6 1055T 3.9GHz NB 3.0GHz ASUS Crosshair V  GTX 770 8GB G.Skill TridentX 1652MHz CL7-7-8-20-22 1T 
Hard DriveOptical DriveCoolingOS
OCZ 60GB Vertex 2E - 3x HDD's Samsung Blu-Ray Drive Custom W/C Loop Windows 7 64bit 
PowerCase
Superflower Leadex 650W Coolermaster Storm Scout Mod 
CPUGraphicsRAMHard Drive
Core i5-3437U Intel HD4000 8GB 256GB SSD 
OSMonitorKeyboardCase
Windows 8 Pro 10.1 WUXGA Touchscreen Worlds Durable Tablet 
  hide details  
Reply
post #97 of 162
You are making total sense. People should not paraphrase what other people are saying. Just use the quote button and do it right.
post #98 of 162
Quote:
Originally Posted by BallaTheFeared View Post
JF made the 80% comment, I just used it.
http://semiaccurate.com/forums/showp...postcount=1107


Quote:
Originally Posted by Behemoth777 View Post
Hopefully this "big improvement" actually translates into more fps.
*fingers crossed*
Edited by pietro sk - 5/26/11 at 8:04am
    
CPUMotherboardGraphicsRAM
Opteron 6386SE G34r1 H8SGL Radeon R9 380X 32GB ECC_DDR3 1333 4ch 
Hard DriveHard DriveCoolingOS
Seagate 600 Pro Enterprise SSD WD Green+Black some water Win8.1_64 
KeyboardPowerCaseMouse
K120 TX850 Tai-chi M560 
AudioAudio
Xmos_u8_DAC; (diy Pluto 2A3/SE estat amp) STAX cookies 3pcs 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Opteron 6386SE G34r1 H8SGL Radeon R9 380X 32GB ECC_DDR3 1333 4ch 
Hard DriveHard DriveCoolingOS
Seagate 600 Pro Enterprise SSD WD Green+Black some water Win8.1_64 
KeyboardPowerCaseMouse
K120 TX850 Tai-chi M560 
AudioAudio
Xmos_u8_DAC; (diy Pluto 2A3/SE estat amp) STAX cookies 3pcs 
  hide details  
Reply
post #99 of 162
Quote:
Originally Posted by JF-AMD View Post
PLEASE STOP QUOTING ME.

How many times have I said that single threaded performance will be higher on Opteron? Many times. 50% more throughput with only 33% more cores. How can it be lower?

If you want to make a statement about your opinion you are free to do so, but when you drop my name in it you are only doing it to lend credibility to your theories.

And you keep making statements that I am saying things that I am not.

Enough.
Quote:
The 80% number keeps getting thrown around. Nobody understands it. The number that we have said is that 2 bulldozer cores in a module would be 180% of the throughput for a single bulldozer core. 90% + 90% = 180%. So the "overhead", so to speak, of the architecture is ~10% per core.

We have an 8-core die that in simple terms is 4 modules, or 180+180+180+180 or 720.

If we did not share resources we would have 6 cores at 100% in the same die size or 6x100=600.

So more total throughput from having more cores, by ~20%.
He just said it.

A 4 core bulldozer chip will operate at 90% + 90%, where as if it had full cores it would have operated at 100% + 100% Which is as far as I'm concerned a ~20% performance hit per module.

Know any dual threaded games JF? ~20% performance loss.

You say it has higher single thread performance over the Opteron, how much clock speed advantage does it require to do it? Or can't you release that information? You can say its faster, but you can't say how many mhz vs mhz is required to be faster? Sure thing marketing guy.

Thanks for the quote, I'm sure I'll be needing that later. I'll make sure I use the direct quote of bulldozer sacrificing per core performance for more cores.
Edited by BallaTheFeared - 5/26/11 at 9:21am
    
CPUMotherboardGraphicsGraphics
Intel Core i5 2500K P8P67 PRO NVIDIA GeForce GTX 470 NVIDIA GeForce GTX 470 
GraphicsRAMRAMRAM
NVIDIA GeForce 9800 GT G-Skill A-Data G-Skill 
RAMHard DriveOptical DriveOS
A-Data Crucial M4 64GB + 1TB F3 Spinpoint $155 LS/DL DVD RW $?? Windows 8 64-bit "Epic Registry" Edition 
MonitorPowerCase
ASUS 21.5 1920x1080 2ms $135 CORSAIR HX850 $120 Mother Earth $free 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel Core i5 2500K P8P67 PRO NVIDIA GeForce GTX 470 NVIDIA GeForce GTX 470 
GraphicsRAMRAMRAM
NVIDIA GeForce 9800 GT G-Skill A-Data G-Skill 
RAMHard DriveOptical DriveOS
A-Data Crucial M4 64GB + 1TB F3 Spinpoint $155 LS/DL DVD RW $?? Windows 8 64-bit "Epic Registry" Edition 
MonitorPowerCase
ASUS 21.5 1920x1080 2ms $135 CORSAIR HX850 $120 Mother Earth $free 
  hide details  
Reply
post #100 of 162
i think per core performance would not be lowered unless the other core in the module is used at the same time.
So probably amd is going to map the cores like tihs:
core 1 first core in first module
core 2 first core in second module
core 3 first core in third module
core 4 first core in fourth module
core 5 second core in first module
core 6 second core in second module
core 7 second core in third module
core 8 second core in fourth module
Like this there wouldn't be any performance decrease for per-core performance unless more than 4 cores are used, and in that case it would be ok to do, because the workload is highly threaded.

edit: That would mean that if we add threads, we would get this performance (100% is one core performance)
1thread 100%
2threads 200%
3threads 300%
4threads 400%
5threads 480%
6threads 560%
7threads 640%
8threads 720%
Or, if all the workloads are not 100% cpu intensive, it could add threads to a core in the least occupied module, and core, thereby reducing the resource sharing.
Edited by creisti86 - 5/26/11 at 10:05am
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
phenom 2 black x4 960T @3.819GHz GA-M78LMT-US2H Sapphire HD 7850 8 GB A-Data @ 1333 
Hard DriveOptical DriveCoolingCooling
SSD ocz vertex 2 60GB, 1.5TB Seagate Barracuda 11 none cpu: TRUE push-pull scythe s-flex 800RPM (quiet) gpu: Accelero s1 modified + 120mm fan 
OSMonitorKeyboardPower
Win 8 Samsung UED325000 32" led-lcd-tv 1920x1080 Logitech 2.4Ghz Corsair CX 400 
CaseMouse
Antec SLK3000 soundproofed Logitech 2.4Ghz 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
phenom 2 black x4 960T @3.819GHz GA-M78LMT-US2H Sapphire HD 7850 8 GB A-Data @ 1333 
Hard DriveOptical DriveCoolingCooling
SSD ocz vertex 2 60GB, 1.5TB Seagate Barracuda 11 none cpu: TRUE push-pull scythe s-flex 800RPM (quiet) gpu: Accelero s1 modified + 120mm fan 
OSMonitorKeyboardPower
Win 8 Samsung UED325000 32" led-lcd-tv 1920x1080 Logitech 2.4Ghz Corsair CX 400 
CaseMouse
Antec SLK3000 soundproofed Logitech 2.4Ghz 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD CPUs
Overclock.net › Forums › AMD › AMD CPUs › Bulldozer Questions from an Intel User.