Overclock.net › Forums › AMD › AMD CPUs › The Zambezi CPU Performance Estimate...
New Posts  All Forums:Forum Nav:

The Zambezi CPU Performance Estimate... - Page 26

post #251 of 354
Quote:
Originally Posted by 2010rig View Post
You really want to talk about the best processors?

Let's look at the last few years and AMD's chips on desktops.
Considering the enthusiast area for Intel is pretty much rebranded Server chips...If AMD did the same thing with G34, in any scenario that uses over 4 cores Intel would be pretty much owned.

Quote:
Originally Posted by BallaTheFeared View Post
Nobody is jelly, and almost nobody here cares about server chips either.
Ironic, Intel just made a cheaper version of their server chipset and chips for Core i7 on 1366...And personally, if we could OC server platforms, I'd be all over that as the CPUs would own. (Dual 16 core BD @ 4Ghz anyone?)

Quote:
Originally Posted by pietro sk View Post
few words on server oftopic -

IBM presented their new cpu - it has awesome specs.
It runs 24/7 @5.2GHz (no turbo) produced at 45nm PDSOI
and of course it has HUGE badass caches (even L4 ..)

Bad thing is - it costs arm and leg, and is not x86 compatible

Therefore i´m eyerolling on SB overclocks... (32nm)

(Thuban, i am disapoint )
Linux for the win, it has PowerPC ports.

Quote:
Originally Posted by _Chimera View Post
Ok so, AMD compares the performance of AMD X6 to an Intel X4 (they're both equal in the image), then compare the Intel X4 to an AMD X8 and they're so proud of the x1.5 performance? OMG are you sure this image is from AMD? and not Intel? Or just a fake by some Intel lover?

I know Nº of cores isn't everything but...seriously...
You're forgetting the price and the fact Intel has HT, the HT makes it roughly equal to a 6 core without HT, just like how a 6 core chip with HT would be roughly equal to a 9 core chip without HT.
Assuming similar IPC and clocks.

Besides, it's not the clock speed, core count, etc that matters, its the maximum performance you can get for your $$$, AMD owns in this as the cheapest Intel CPU worth getting for half of us (2500K) costs AU$279, I could get a 1055T or 965 BE for far cheaper than that and still get 90% of the performance.

Quote:
Originally Posted by 2010rig View Post
AMD cried to the courts about Intel not optimizing their code for AMD chips, and got paid over $1 Billion for it! Why should Intel be forced to optimize their code to be compliant with their competing processors? You are VERY right in that regard.

Based on your logic, instead of crying about it, perhaps AMD should write their own code libraries, instead of using Intel's!
A compiler should check for the instruction sets, etc, so if a chip says "I have SSE3!" it should automatically let it use that, instead Intel goes off the VendorID so their chips are optimized, but AMD and VIAs aren't.

You obviously are either trolling and a fanboy or you know very little about coding.
My system
(23 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 3570k @ 4.5Ghz ASRock Z77 Pro3 Powercolor Radeon HD7950 3GB @ 1150/1350 4x4GB G.Skill Ares 2000Mhz CL9 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 250GB Western Digital Black 1TB WD1002FAEX Seagate Barracuda 3TB ST3000DM001 Samsung Spinpoint EcoGreen 2TB 
Optical DriveCoolingCoolingCooling
Pioneer DVR-220LBKS Noctua NH-D14 Scythe Gentle Typhoon 1850rpm Corsair AF140 Quiet Edition 
CoolingOSMonitorMonitor
Arcitc Cooling Acclero Twin Turbo II Arch Linux x86-64, amdgpu BenQ G2220HD BenQ G2020HD 
KeyboardPowerCaseMouse
Ducky Shine III Year of the Snake, Cherry Blue Silverstone Strider Plus 600w CoolerMaster CM690 II Black and White SteelSeries Sensei Professional 
Mouse PadAudioOther
Artisan Hien Mid Japan Black Large ASUS Xonar DX NZXT Sentry Mesh 30w Fan Controller 
  hide details  
Reply
My system
(23 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 3570k @ 4.5Ghz ASRock Z77 Pro3 Powercolor Radeon HD7950 3GB @ 1150/1350 4x4GB G.Skill Ares 2000Mhz CL9 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 250GB Western Digital Black 1TB WD1002FAEX Seagate Barracuda 3TB ST3000DM001 Samsung Spinpoint EcoGreen 2TB 
Optical DriveCoolingCoolingCooling
Pioneer DVR-220LBKS Noctua NH-D14 Scythe Gentle Typhoon 1850rpm Corsair AF140 Quiet Edition 
CoolingOSMonitorMonitor
Arcitc Cooling Acclero Twin Turbo II Arch Linux x86-64, amdgpu BenQ G2220HD BenQ G2020HD 
KeyboardPowerCaseMouse
Ducky Shine III Year of the Snake, Cherry Blue Silverstone Strider Plus 600w CoolerMaster CM690 II Black and White SteelSeries Sensei Professional 
Mouse PadAudioOther
Artisan Hien Mid Japan Black Large ASUS Xonar DX NZXT Sentry Mesh 30w Fan Controller 
  hide details  
Reply
post #252 of 354
Quote:
Originally Posted by scyy View Post
Ok, ignore anandtech then. Now explain the dozens of other reviews that showed the exact same results. As I already said complaining about a site being biased is such a cop out it's not even funny. I can guarantee if anandtech showed bulldozer beating SB you would completely change your tune but because it doesn't fit into your current preconceptions about performance you call it biased. Just like for years AMD people would complain about cinebench being biased. Than the 1090t came out and beat the i7's and suddenly it was a totally legit bench, then sandybridge came out and its biased again. It's so ridiculous and you don't even realize it.
Its true that all revievers shows bigger performance on SB, than other cpus. i didnt said else
But : What other sites shows, is NOT EXACT SAME scores like that at anandtech..


I dont : -visit -read anandtech.
Because they did and will do - not ideal/weird comparisons between HW. (not only few times)

Therefore its wrong if somebody builds his beliefs about performance
only on anandtech - thats why i criticized it. (2010rig)


---------------
About cinebench
cinebench the old version r10 32b cripples score on amd cpus. Thats been told many times.

New version 11,5 seems "fixed". Anyway , i personally ignore it - i dont see valid reason,
why standard desktop user should bother with render benchmark.
The weirdest part is, when reviewers test small netbooks /laptops with this render test..
    
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 #253 of 354
I guess I'm coming from left field on this when I say I only need a chip fast enough for my gpu(s).

I don't really care how many cores the CPU has, all that I care about is my gpu usage. If its dropping when I have 4 cores in a game coded for 2, then getting 4 or 8 more of those same cores will do nothing for me.

Phenom II wasn't able to do it for me, hopefully Bulldozer can.
    
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 #254 of 354
Quote:
Originally Posted by pietro sk View Post
Wroooooong, you misunderstood
Intel was "invited" to court

1- not because : intel did not optimized their compiler for AMD cpus (thats not their business) ...

2-BUT BECAUSE : intel compiler takes the worst code path if AMD (or other non Intel) cpu is detected
==> to maximize harm / performance penalty on these processors..
Shady as sh_t, ugly Intel ..

... and of course ,also because of nasty coercing and bribing IT industry (OEMs)..

----------------

You misunderstood again, AMD does not use intel libs - but SW vendors do..
John will tell - AMD supports many opensource and microsoft compilers, AMD has not own compiler.

But at the end, software producer picks the tools (compiler)..
And that piece of software shouldn´t generate buggy code ..

-------------------

About nvidia part : i dont care if nvidia goes bankrupt or not - its not my interest
1 my idea was - with proper SW , ati HW may outperform CUDA..
2 opensource vs proprietary
The point of using Intel libs is that AMD should provide their own libraries to SW vendors, so they can use those INSTEAD of using Intel's. I bet AMD would ensure that those libraries are optimized for Intel's processors right?

If Intel's libraries provide "buggy" code as you put it, why doesn't AMD step up to the plate and write better libraries?

It's ugly when Intel does something that doesn't benefit the competition, but it's ok for AMD to do it - oh the irony in your statements, that's the point I was making. You call me a fanboy and whatever, but you're the biggest AMD fanatic here. You claim benchmarks are rigged by ex-Intel employees, and you are shown otherwise by others, but you're still in denial about it.

Face it AMD is 2nd best right NOW, out of 2 options, quit being in denial about it.

Your idea with proper SW which AMD doesn't write or own, you're speculating that it MAY outperform CUDA. You ignored my statements on how far behind AMD is with OpenCL API, and thus it's not even an option.
Quote:
"While AMD will tell you that they're all for open standards and push OpenCL, the sad truth is that the company representatives will remain shut when you ask them about the real status of their OpenCL API"

"I mean at this point AMD is so far behind in development tools they are not even worth pursuing right now."
Until AMD provides something worth pursuing, all you can do is SPECULATE. While Nvidia provided a stable platform to use, which Adobe CHOSE to use to build the Mercury Playback Engine with.
Edited by 2010rig - 2/18/11 at 6:13pm
2010rig
(14 items)
 
Galaxy S3
(8 items)
 
 
CPUMotherboardGraphicsRAM
X5660 @ 4.5  ASUS P6X58D-E 980TI? 12GB OCZ Platinum - 7-7-7-21 
Hard DriveCoolingOSMonitor
1 80GB SSD x25m - 3TB F3 + F4 NH-D14 Windows 7 Ultimate LG 47LH55 
KeyboardPowerCaseMouse
Natural Wireless Keyboard Corsair 750HX CM 690 II Advanced MX 518 
CPUGraphicsRAMHard Drive
Snapdragon S4 Dual core 1500mhz Adreno 225 Samsung 2GB 16GB Onboard Flash 
OSMonitorPowerCase
Android 4.4.2 - CM11 4.8" AMOLED 1280x720 2100 mAh battery Otterbox Defender 
  hide details  
Reply
2010rig
(14 items)
 
Galaxy S3
(8 items)
 
 
CPUMotherboardGraphicsRAM
X5660 @ 4.5  ASUS P6X58D-E 980TI? 12GB OCZ Platinum - 7-7-7-21 
Hard DriveCoolingOSMonitor
1 80GB SSD x25m - 3TB F3 + F4 NH-D14 Windows 7 Ultimate LG 47LH55 
KeyboardPowerCaseMouse
Natural Wireless Keyboard Corsair 750HX CM 690 II Advanced MX 518 
CPUGraphicsRAMHard Drive
Snapdragon S4 Dual core 1500mhz Adreno 225 Samsung 2GB 16GB Onboard Flash 
OSMonitorPowerCase
Android 4.4.2 - CM11 4.8" AMOLED 1280x720 2100 mAh battery Otterbox Defender 
  hide details  
Reply
post #255 of 354
Quote:
Originally Posted by Brutuz View Post
Linux for the win, it has PowerPC ports.

You're forgetting the price and the fact Intel has HT, the HT makes it roughly equal to a 6 core without HT, just like how a 6 core chip with HT would be roughly equal to a 9 core chip without HT.
Assuming similar IPC and clocks.

A compiler should check for the instruction sets, etc, so if a chip says "I have SSE3!" it should automatically let it use that, instead Intel goes off the VendorID so their chips are optimized, but AMD and VIAs aren't.

You obviously are either trolling and a fanboy or you know very little about coding.
thumbs up !

that IBM MCM moster has 192MB L4 - scary one..
    
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 #256 of 354
Quote:
Originally Posted by 2010rig View Post
Thanks for clearing that up, but actually, this is what you said, which is exactly what I posted:
...
No, you are absolutely correct in that context. Without sifting back through that thread, I can't really say the context, but what I THINK that I was saying was that "the slide did not come from AMD" was a reference to AMD giving the slide and AMD leaking, not necessarily that AMD did not create the slide.

I have replied to dozens of questions on dozens of sites regarding that slide.

All I can say for sure is that this was not part of some manipulated leak campaign. If it is an AMD slide it was never meant to be shared.

But again, I can't say whether it is real or not, only that I had never seen it. As a client slide, I only see a fraction of what they create.
post #257 of 354
Quote:
Originally Posted by pietro sk View Post
Wroooooong, you misunderstood
Intel was "invited" to court

1- not because : intel did not optimized their compiler for AMD cpus (thats not their business) ...

2-BUT BECAUSE : intel compiler takes the worst code path if AMD (or other non Intel) cpu is detected
==> to maximize harm / performance penalty on these processors..
Shady as sh_t, ugly Intel ..

... and of course ,also because of nasty coercing and bribing IT industry (OEMs)..

----------------

You misunderstood again, AMD does not use intel libs - but SW vendors do..
John will tell - AMD supports many opensource and microsoft compilers, AMD has not own compiler.

But at the end, software producer picks the tools (compiler)..
And that piece of software shouldn´t generate buggy code ..

-------------------

About nvidia part : i dont care if nvidia goes bankrupt or not - its not my interest
1 my idea was - with proper SW , ati HW may outperform CUDA..
2 opensource vs proprietary

Software developers pick compilers upon many factors. One of the most significant is cost vs featureset. Hmm free vs $
Gunslinger
(13 items)
 
  
CPUMotherboardGraphicsRAM
1090 MSI 890FXA GD70 5870 Corsair 
Hard DriveOSMonitorPower
C300 Windows I-INC Thermaltake 850 
Case
HAF 932 
  hide details  
Reply
Gunslinger
(13 items)
 
  
CPUMotherboardGraphicsRAM
1090 MSI 890FXA GD70 5870 Corsair 
Hard DriveOSMonitorPower
C300 Windows I-INC Thermaltake 850 
Case
HAF 932 
  hide details  
Reply
post #258 of 354
Quote:
Originally Posted by BallaTheFeared View Post
AVX has nothing to do with 3D Mark Vantage.
Over reliance upon one benchmark is not prudent as shown previous.
Gunslinger
(13 items)
 
  
CPUMotherboardGraphicsRAM
1090 MSI 890FXA GD70 5870 Corsair 
Hard DriveOSMonitorPower
C300 Windows I-INC Thermaltake 850 
Case
HAF 932 
  hide details  
Reply
Gunslinger
(13 items)
 
  
CPUMotherboardGraphicsRAM
1090 MSI 890FXA GD70 5870 Corsair 
Hard DriveOSMonitorPower
C300 Windows I-INC Thermaltake 850 
Case
HAF 932 
  hide details  
Reply
post #259 of 354
Quote:
Originally Posted by pietro sk View Post
Its true that all revievers shows bigger performance on SB, than other cpus. i didnt said else
But : What other sites shows, is NOT EXACT SAME scores like that at anandtech..


I dont : -visit -read anandtech.
Because they did and will do - not ideal/weird comparisons between HW. (not only few times)

Therefore its wrong if somebody builds his beliefs about performance
only on anandtech - thats why i criticized it. (2010rig)


---------------
About cinebench
cinebench the old version r10 32b cripples score on amd cpus. Thats been told many times.

New version 11,5 seems "fixed". Anyway , i personally ignore it - i dont see valid reason,
why standard desktop user should bother with render benchmark.
The weirdest part is, when reviewers test small netbooks /laptops with this render test..
Pietro it appears some of the poster here do not like facts and do not listen when present different benchmarks. Oh well, show the ineffectivenes of the education system
Edited by Kmon - 2/18/11 at 6:22pm
Gunslinger
(13 items)
 
  
CPUMotherboardGraphicsRAM
1090 MSI 890FXA GD70 5870 Corsair 
Hard DriveOSMonitorPower
C300 Windows I-INC Thermaltake 850 
Case
HAF 932 
  hide details  
Reply
Gunslinger
(13 items)
 
  
CPUMotherboardGraphicsRAM
1090 MSI 890FXA GD70 5870 Corsair 
Hard DriveOSMonitorPower
C300 Windows I-INC Thermaltake 850 
Case
HAF 932 
  hide details  
Reply
post #260 of 354
Sounds good JF-AMD,

I wasn't trying to put words in your mouth or anything. The context of my reply was to let that guy know that the slide did not come from AMD, as this is what you said in that other thread, in response to the Turkish slide.



AFAIK - We don't have any reliable BD data to go by, and as you can see by this thread, it leads to a lot of speculation, and in many instances name calling, etc.
Edited by 2010rig - 2/18/11 at 6:36pm
2010rig
(14 items)
 
Galaxy S3
(8 items)
 
 
CPUMotherboardGraphicsRAM
X5660 @ 4.5  ASUS P6X58D-E 980TI? 12GB OCZ Platinum - 7-7-7-21 
Hard DriveCoolingOSMonitor
1 80GB SSD x25m - 3TB F3 + F4 NH-D14 Windows 7 Ultimate LG 47LH55 
KeyboardPowerCaseMouse
Natural Wireless Keyboard Corsair 750HX CM 690 II Advanced MX 518 
CPUGraphicsRAMHard Drive
Snapdragon S4 Dual core 1500mhz Adreno 225 Samsung 2GB 16GB Onboard Flash 
OSMonitorPowerCase
Android 4.4.2 - CM11 4.8" AMOLED 1280x720 2100 mAh battery Otterbox Defender 
  hide details  
Reply
2010rig
(14 items)
 
Galaxy S3
(8 items)
 
 
CPUMotherboardGraphicsRAM
X5660 @ 4.5  ASUS P6X58D-E 980TI? 12GB OCZ Platinum - 7-7-7-21 
Hard DriveCoolingOSMonitor
1 80GB SSD x25m - 3TB F3 + F4 NH-D14 Windows 7 Ultimate LG 47LH55 
KeyboardPowerCaseMouse
Natural Wireless Keyboard Corsair 750HX CM 690 II Advanced MX 518 
CPUGraphicsRAMHard Drive
Snapdragon S4 Dual core 1500mhz Adreno 225 Samsung 2GB 16GB Onboard Flash 
OSMonitorPowerCase
Android 4.4.2 - CM11 4.8" AMOLED 1280x720 2100 mAh battery Otterbox Defender 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD CPUs
Overclock.net › Forums › AMD › AMD CPUs › The Zambezi CPU Performance Estimate...