AMD FX-8350 Cracks 7.443 GHz

[btarunr]

AMD's next-generation FX series processors may be just around the corner, but OC feats around its flagship model, the FX-8350, are already beginning to take shape. Chinese overclocker Ma Yuchuan (马宇川) achieved 7.443 GHz with the FX-8350, except with just one of its four modules enabled (two cores). The clock speed of 7443 MHz was set with a base clock of 244.3 MHz, and multiplier of 30.5X. A core voltage of 1.968V was used, and the chip was cooled with liquid nitrogen. 8 GB of dual-channel DDR3-1600 MHz memory by Patriot was used, and seating it all was Gigabyte 990FXA-UD7 motherboard. Find the CPUID validation here. The feat still trails Andre Yang's 8.71 GHz on FX-8150. That's not to say that the FX-8350 isn't as good a chip with overclocking, not enough enthusiasts have their hands on it, yet.



Source: Expreview

[BlackOmega]

Not bad. Was he able to run super-pi or anything?

[cdawall]

Good to see they are already clocking up. Wonder what we will get out of them with all modules enabled and stable on air/water. Hopefully in the 5.5ghz+ range easy.

[Disruptor4]

Originally Posted by cdawall

Good to see they are already clocking up. Wonder what we will get out of them with all modules enabled and stable on air/water. Hopefully in the 5.5ghz+ range easy.

Good feat, but I think I'd like to see right now is how they compare clock for clock with BD. What kinda IPC changes have they done? Voltage changes? etc etc

[Delta6326]

Wonder whats faster this or a stock FX-8350 with all of its cores, for real.

[TheGuruStud]

Originally Posted by Disruptor4

Good feat, but I think I'd like to see right now is how they compare clock for clock with BD. What kinda IPC changes have they done? Voltage changes? etc etc

Really only about to about 7.5% max IPC increase. Maximum of 15% increase from the higher clocks.

[tacosRcool]

Does it fare better than the older AMD chips in terms of performance?

[de.das.dude]

arent all these chinese announcements usually fake?

[Hustler]

Originally Posted by TheGuruStud

Really only about to about 7.5% max IPC increase.

Lol..that still makes them slower clock for clock than a Phenom II.

[eidairaman1]

Originally Posted by Hustler

Lol..that still makes them slower clock for clock than a Phenom II.

this was just a minor improvement, AMD appears to be more focused on steamroller

now lets compare a 8350 to a 8150 at same clock speed then a 8320 to a 8150, then a 8320 to 8120

[nt300]

Originally Posted by Hustler

Lol..that still makes them slower clock for clock than a Phenom II.

They are both from different gens. Clock for clock means nothing, but compared to the Bulldozer yes you can make that comparison.

[eidairaman1]

Originally Posted by nt300

They are both from different gens. Clock for clock means nothing, but compared to the Bulldozer yes you can make that comparison.

architecture is different.

Good design poor execution.

btw didnt the Bulldozer only reach 8 GHz on a single core and not a module where as this hit 7+ with a module (contains 2 cores), I mean wouldnt that technically be faster than that Bulldozer then

[Fourstaff]

Well done to him for getting good results, but I suspect this will not matter too much just like how the Pentium 4's managed insane clockspeeds but no meaningful way forward.

[eidairaman1]

Originally Posted by Fourstaff

Well done to him for getting good results, but I suspect this will not matter too much just like how the Pentium 4's managed insane clockspeeds but no meaningful way forward.

hey you recall BD breaking a OC record, wasnt that just with a single core and not a module since a module contains 2 cores sharing certain resources internally

[NeoXF]

BD hit 8,4xxGHz sometime in September, but as you guys said, I think only on 1 core/1 module. Either way, I'm sure we'll keep hearing about higher and higher records on PD as well.

[Fourstaff]

Originally Posted by eidairaman1

hey you recall BD breaking a OC record, wasnt that just with a single core and not a module since a module contains 2 cores sharing certain resources internally

Yes I vaguely remember BD snatching the OC crown, but I am not convinced until I see tangible results in commercial applications (read: them start selling Bulldozers clocked at 5Ghz etc.)

[eidairaman1]

Originally Posted by NeoXF

BD hit 8,4xxGHz sometime in September, but as you guys said, I think only on 1 core/1 module. Either way, I'm sure we'll keep hearing about higher and higher records on PD as well.

ok a FX 8 has 2 cores per module and 4 modules per CPU.

FX 6 has 3 modules

FX 4 has 2 modules

[w3b]

Originally Posted by de.das.dude

arent all these chinese announcements usually fake?

Article features a CPUID validation so perhaps that may not be the case in this instance

Originally Posted by btarunr

Find the CPUID validation here.

Agent 86: "Missed it by that much."

[No_Asylum]

These tests are stupid and pointless. NO ONE uses their PC this way. Show us how high it will overclock with the best air or water cooling available. Not this ridiculous Liquid Nitrogen crap that has zero real world implications.

[AphexDreamer]

Originally Posted by No_Asylum

These tests are stupid and pointless. NO ONE uses their PC this way. Show us how high it will overclock with the best air or water cooling available. Not this ridiculous Liquid Nitrogen crap that has zero real world implications.

Its just done to see the limits. Like jet cars breaking the sound barrier or F1 racing. No one is going to use their car like that.

[cdawall]

Originally Posted by NeoXF

BD hit 8,4xxGHz sometime in September, but as you guys said, I think only on 1 core/1 module. Either way, I'm sure we'll keep hearing about higher and higher records on PD as well.

8.7ghz in Nov of last year with one entire module running

Originally Posted by No_Asylum

These tests are stupid and pointless. NO ONE uses their PC this way. Show us how high it will overclock with the best air or water cooling available. Not this ridiculous Liquid Nitrogen crap that has zero real world implications.

Please don't troll. No one cares what you think about CPUz and max clocks.

[trickson]

Originally Posted by No_Asylum

These tests are stupid and pointless. NO ONE uses their PC this way. Show us how high it will overclock with the best air or water cooling available. Not this ridiculous Liquid Nitrogen crap that has zero real world implications.

I agree, But this is kind of good in a way. At least they will be able to see how far they can push the silicon and thus may even net us higher speeds for future stuff. Maybe AMD will have us stock 5GHz Chips in another years time, that will clock to 6.5GHz on Air! Now this would be very impressive!

[nt300]

Originally Posted by No_Asylum

These tests are stupid and pointless. NO ONE uses their PC this way. Show us how high it will overclock with the best air or water cooling available. Not this ridiculous Liquid Nitrogen crap that has zero real world implications.

Will all due respect, what is wrong with running liquid 24/7? I use liquid nitrogen and it great.

High clocks are great, it shows you how much that silicon can be pushed.

[anubis44]

Originally Posted by No_Asylum

These tests are stupid and pointless. NO ONE uses their PC this way. Show us how high it will overclock with the best air or water cooling available. Not this ridiculous Liquid Nitrogen crap that has zero real world implications.

Why do I have such a nagging feeling that if these were results for an Intel processor, you'd be shouting about how it proves the superiority of Intel's overclocking ability, and how that would apply even for air or liquid coolers, too...

As others have said, this kind of overclocking is done to show the upper limits of the process technology and to push the boundaries to the extreme. Of course nobody is going to run their system on liquid nitrogen on a regular basis - nobody here is saying anything of the kind. But these results do give us some glimpse into the overclockability of this CPU, even using more conventional cooling methods. I just bought a closed loop water cooler for the first time, and I'm very excited to see what the new Vishera chips can do on water now.

[eidairaman1]

Originally Posted by nt300

Will all due respect, what is wrong with running liquid 24/7? I use liquid nitrogen and it great.

High clocks are great, it shows you how much that silicon can be pushed.

Originally Posted by trickson

I agree, But this is kind of good in a way. At least they will be able to see how far they can push the silicon and thus may even net us higher speeds for future stuff. Maybe AMD will have us stock 5GHz Chips in another years time, that will clock to 6.5GHz on Air! Now this would be very impressive!

Originally Posted by anubis44

Why do I have such a nagging feeling that if these were results for an Intel processor, you'd be shouting about how it proves the superiority of Intel's overclocking ability, and how that would apply even for air or liquid coolers, too...

As others have said, this kind of overclocking is done to show the upper limits of the process technology and to push the boundaries to the extreme. Of course nobody is going to run their system on liquid nitrogen on a regular basis - nobody here is saying anything of the kind. But these results do give us some glimpse into the overclockability of this CPU, even using more conventional cooling methods. I just bought a closed loop water cooler for the first time, and I'm very excited to see what the new Vishera chips can do on water now.

Ya the troll below probably would praise Intel on such a high overclock, probably did it with the 65nm P4 Cedarmill model at the time too. I dont need to say anything more as everyone quoted above has said it for me.

Originally Posted by No_Asylum

These tests are stupid and pointless. NO ONE uses their PC this way. Show us how high it will overclock with the best air or water cooling available. Not this ridiculous Liquid Nitrogen crap that has zero real world implications.

go troll elsewhere