AMD unleashes first ever commercial “5GHz” CPU, the FX-9590

[ViRGE]

Enough arguing about cores.
-ViRGE

 

[galego]

This is a simulation of the performance of the FX-9590 for crysis 3

That is a previous CPU comparison AMD vs Intel

Here a comparison Centurion vs 3930K and 3960X on x264 5.0

This test seems to be intel optimized, and they are using W7 plus the manual hotfixes, when the automatic updates or the improved scheduler in W8 are a bit better, but already shows that the Centurion competes with both 3930K and 3970X. Look the Haswell score as well.

 

[sushiwarrior]

With turbo enabled. Do you want me to disable it?

Sounds like you don't have CPU load line calibration set to Ultra High/Extreme, also try disabling C1E and Cool n Quiet. With those enabled the chip will revert to stock clocks under load.

 

[loccothan]

i will have one of these ;-) my cooler combo is able to take min. 350W so 220W Tops AMD is not a big deal to me and my solution ! Great job AMD.... SteamRoller will be on 1100FX chipset and DDR4 Quad 16 cores 8 modules lol im waiting for that too for 230$

 

[guskline]

My 8350 stock (mem. @ 1866) takes 2m57s.
Sounds as if both times you had it OCed.

How are you OCing it? Through BIOS or some program on Windows?

No all of my OCing is through the BIOS. I'll rerun the 8350 stock. I also have a 8320 so, time permitting, I'll run it.

 

[guskline]

Sounds like you don't have CPU load line calibration set to Ultra High/Extreme, also try disabling C1E and Cool n Quiet. With those enabled the chip will revert to stock clocks under load.

Thanks. Will try that.

 

[jhu]

With turbo enabled. Do you want me to disable it?

No that's fine. Just need to know what speed the processor is at with those results.

 

[AtenRa]

@guskline

Try those settings when OC and plz rerun PovRay with 8-cores at 4.7GHz and see if you get a different outcome than beffore.

EPU Power Saving Mode [Disabled]
DIGI+ VRM
CPU Load-line Calibration [Extreme]
CPU/NB Load Line Calibration [Extreme]
CPU Power Phase Control [Extreme]
CPU Power Duty Control [C.Probe Thermal]

Also, turbo mode(or CPB) should be disable.

 

[toyota]

This is a simulation of the performance of the FX-9590 for crysis 3

That is a previous CPU comparison AMD vs Intel

Here a comparison Centurion vs 3930K and 3960X on x264 5.0

This test seems to be intel optimized, and they are using W7 plus the manual hotfixes, when the automatic updates or the improved scheduler in W8 are a bit better, but already shows that the Centurion competes with both 3930K and 3970X. Look the Haswell score as well.

and do we need to post the slides from other Crysis 3 reviews that show the 8350 being slower?

 

[BallaTheFeared]

He doesn't actually read anything people post, same thing as in the PS4 thread.

 

[jhu]

and do we need to post the slides from other Crysis 3 reviews that show the 8350 being slower?

That would be informative. Please post them.

 

[thepath]

That would be informative. Please post them.

Here check this
http://www.overclock.net/t/1364211/pclab-more-crysis-3-cpu-benchmarks

Also, check this The Lost Island DLC benchmark for crysis 3. Even 2nd gen i5 wins
http://www.overclock.net/content/type/61/id/1501649/width/500/height/1000/flags/LL

 

[Sweepr]

Here check this
http://www.overclock.net/t/1364211/pclab-more-crysis-3-cpu-benchmarks

Also, check this The Lost Island DLC benchmark for crysis 3. Even 2nd gen i5 wins
http://www.overclock.net/content/type/61/id/1501649/width/500/height/1000/flags/LL

The reason he always posts this Welcome to the Jungle benchmark is that its the only level the FX8350 manages to match Intel's quads (average FPS).

Also, check this:

1-2+ average FPS dont always mean best gaming experience.

 

[toyota]

That would be informative. Please post them.

http://www.techspot.com/review/642-crysis-3-performance/page6.html

here the 4 core 2500k matches the 8350 even though the game is using 8 threads.
http://gamegpu.ru/action-/-fps-/-tps/crysis-3-the-lost-island-test-gpu.html

 

[galego]

here the 4 core 2500k matches the 8350 even though the game is using 8 threads.
http://gamegpu.ru/action-/-fps-/-tps/crysis-3-the-lost-island-test-gpu.html

Misusing 8 threads is a more adequate description. Look at the core load percentages. That explain why the 8 threads 2600k (3.4GHz) got almost the same FPS than the 4 threads 2500k (3.3GHz), the difference being entirely explained by the higher clock.

 

[galego]

It is unlikely that people will buy the enthusiast line 9000 to run software that exploits only a half of the chip.

I already remarked that the 9000 series main goal will be future highly multi-threaded games. The first time I used crysis 3 Welcome to the Jungle benchmark as example of how next games will be. Several of you missed my point entirely.

The second time I used it for comparison with the FX-9590 simulation.

 

[toyota]

Misusing 8 threads is a more adequate description. Look at the core load percentages. That explain why the 8 threads 2600k (3.4GHz) got almost the same FPS than the 4 threads 2500k (3.3GHz), the difference being entirely explained by the higher clock.

welcome to the real world. if you owned a pc you would know thats how things work most of the time. instead you search day and night to find a few niche results that back up your pathological obsession with anything AMD.

 

[monstercameron]

welcome to the real world. if you owned a pc you would know thats how things work most of the time. instead you search day and night to find a few niche results that back up your pathological obsession with anything AMD.

and your personal attack is any better? argue the fact/fictions and move on. In any case, we all know that the amd solution would perform better with better multi-threaded software.

 

[galego]

welcome to the real world. if you owned a pc you would know thats how things work most of the time. instead you search day and night to find a few niche results that back up your pathological obsession with anything AMD.

Besides being used as excuse for another ad hominem, your post lacks any relevance to my points.

 

[guskline]

@guskline

Try those settings when OC and plz rerun PovRay with 8-cores at 4.7GHz and see if you get a different outcome than beffore.

EPU Power Saving Mode [Disabled]
DIGI+ VRM
CPU Load-line Calibration [Extreme]
CPU/NB Load Line Calibration [Extreme]
CPU Power Phase Control [Extreme]
CPU Power Duty Control [C.Probe Thermal]

Also, turbo mode(or CPB) should be disable.

Same outcome. 2m 31 s. BTW, I already had those settings.

 

[toyota]

Besides being used as excuse for another ad hominem, your post lacks any relevance to my points.

you have no pc and know nothing about real world performance on anything pc related. again all you do is scour the web trying to only find anything to back up your obsession with anything AMD.

 

[AtenRa]

Same outcome. 2m 31 s. BTW, I already had those settings.

Very strange, is there a CPB option in the BIOS ??



What project do you bench in PovRay ??

 

[guskline]

Very strange, is there a CPB option in the BIOS ??



What project do you bench in PovRay ??

POVray 3.7 RC7. I go to the render all cores option for the benchmark and run it.

What do you get on your 8350 at 4.7Ghz?

 

[ShintaiDK]

That is a previous CPU comparison AMD vs Intel

You should read the updated test:
http://www.pcgameshardware.de/Crysis-3-PC-235317/Tests/Crysis-3-CPU-Test-1068140/

Even the 3770 is now over 30% faster in your favourite bench than that FX CPU you keep promoting. Even a stock 3770, not to mention 4770, would easily beat a 5Ghz FX.

 

[Enigmoid]

You should read the updated test:
http://www.pcgameshardware.de/Crysis-3-PC-235317/Tests/Crysis-3-CPU-Test-1068140/

Even the 3770 is now over 30% faster in your favourite bench than that FX CPU you keep promoting. Even a stock 3770, not to mention 4770, would easily beat a 5Ghz FX.

Wow, look at that HT gain.

57 -> 75 fps

3770k ->75 fps
3570k ->55 fps

36% increase because of hyperthreading.