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

Page 15 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

ShintaiDK

Lifer
Apr 22, 2012
20,378
146
106
As i wrote , 150W despite and external memory controller.

Anyway the chip wont absorb 220W , it will probably
at most consume 40% more than the 8350 assuming
they did cranck up the volts significantly...

Looks like the chips will be 1.3825V at iddle and 1.46
when turbo kicks , sorry to disapoint you...

Just as the FX8350 is a 125W chip....right?
 

Abwx

Lifer
Apr 2, 2011
12,034
4,995
136
Just as the FX8350 is a 125W chip....right?

As much as HW that has 153W peak TDP , isnt it....

You re unrelentlessly coming again with a single data
measured at the power wall , that is the worst way
to estimate a CPU TDP , and that is at odd with all
sites measurements including Anand.

If you cant stand a discussion without always
creating straws please do not answer me again.
 

wlee15

Senior member
Jan 7, 2009
313
31
91
lol their slides are always good for a laugh.

I know my i5 will already beat it's Cinebench score while using 1/3 the power :thumbsup:

I'm not sure that you could match it even if you oc your 4670 up to 4.8 GHz. Cinebench is one those programs that benefits greatly from hyperthreading and CMT, and a 4.8 GHz Piledriver should get about 8.3.
 

Abwx

Lifer
Apr 2, 2011
12,034
4,995
136
I'm not sure that you could match it even if you oc your 4670 up to 4.8 GHz. Cinebench is one those programs that benefits greatly from hyperthreading and CMT, and a 4.8 GHz Piledriver should get about 8.3.

At 73W CPU power comsumption according to his estimations ,
let see if he manage the number , and that s assuming
that the FX will draw 220W when cinebenching , wich i can
tell you for sure that it wont be the case , he will have to
get 60W or so to match his claim of 1/3 the power draw.
 

AtenRa

Lifer
Feb 2, 2009
14,003
3,362
136
220 watts for the stock cpu? lol my whole system only uses a little more power than that even in Crysis 3 which pushes my oced 2500k and gtx660ti.

Once again, TDP is not power consumption and obviously your CPU doesn’t operates at 100% when playing Crysis 3.[FONT=&quot]
[/FONT]
 

Zucker2k

Golden Member
Feb 15, 2006
1,810
1,159
136
Once again, TDP is not power consumption and obviously your CPU doesn’t operates at 100% when playing Crysis 3.[FONT=&quot]
[/FONT]
Yes, but the truth is these are going to be consuming more than the rated 220w TDP, if previous trends hold. :D
 

Abwx

Lifer
Apr 2, 2011
12,034
4,995
136
Yes, but the truth is these are going to be consuming more than the rated 220w TDP, if previous trends hold. :D

Typical useless post that talks of trends just for the purpose
of trashing a firm , no number , no site link , nothing than
blatant false information witrh the hope that it will be trusted
by some gullible by stander...

24 posts in 7 years.?..
It makes me wonder about those silent members
suddenly popping up for trashing purposes..
 

d3m

Junior Member
Jun 5, 2013
23
0
66
I am more interested in some benchmarks rather than reading all of your bitching about how bad this chip is.

My GPU consumes more than this chip.
 

Abwx

Lifer
Apr 2, 2011
12,034
4,995
136
Knowing your historical record of failed predictions, this confirms that it will be a success.

The most striking is to see people overclocking 2500Ks
to death wich will bring TDPs at high levels , yet even
with 50% overcolck it wont match a 8350 at stock
on multithreaded scenarii , now there is others that start
from 4670k with the same claim , ignoring thzat it wont*even
reach a 2500K because of poor overclocking.

http://techreport.com/review/23750/amd-fx-8350-processor-reviewed/13
 

toyota

Lifer
Apr 15, 2001
12,957
1
0
Once again, TDP is not power consumption and obviously your CPU doesn’t operates at 100% when playing Crysis 3.[FONT=&quot]
[/FONT]
lol its most certainly related is it not? and actually it does hit 90-100% in that game at times. thats the reason I mentioned Crysis 3 because it taxes both my cpu and gpu. and yes I know its not the same as running IBT but the point is that a 220 watt rated cpu out of the box is ridiculous.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
146
106
I am more interested in some benchmarks rather than reading all of your bitching about how bad this chip is.

My GPU consumes more than this chip.

Its rather easy to predict its performance. Its not a new uarch, just clocked up.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
146
106
As much as HW that has 153W peak TDP , isnt it....

You re unrelentlessly coming again with a single data
measured at the power wall , that is the worst way
to estimate a CPU TDP , and that is at odd with all
sites measurements including Anand.

If you cant stand a discussion without always
creating straws please do not answer me again.

Yet the FX triggers VRM throttle under heavy load, unlike Haswell.

It seems these 220W chips will be integrator only and not hit the channel. Guess thats how much faith it was given. A novelty product for PR only.
 

Zucker2k

Golden Member
Feb 15, 2006
1,810
1,159
136
Typical useless post that talks of trends just for the purpose
of trashing a firm , no number , no site link , nothing than
blatant false information witrh the hope that it will be trusted
by some gullible by stander...

24 posts in 7 years.?..
It makes me wonder about those silent members
suddenly popping up for trashing purposes..
I was being fair (read kind) when I referenced trends. Besides, outside of real numbers, we only have AMD's own 220watts to go by, and the (wait for it) trend is, AMD's high end chips have always consumed more than rated TDP. Also the same piece of silicon from which these chips are derived as evidenced in the flagship 8350 consumes way above 220w when overclocked to 4.7Ghz. This is what I'm basing my argument on. What is yours? What is your justification for thinking the 220w rating is just an arbitrary figure? I'm waiting to read what you have to say. Bring your post count/history to bear on the answer, please. :D

As for my post count, what has that got to do with anything? You've been here only 2 years with 2,324 posts! How dare I have 24 posts in 7 years? Sorry dude, do you need a hug?
 

AtenRa

Lifer
Feb 2, 2009
14,003
3,362
136
lol its most certainly related is it not? and actually it does hit 90-100% in that game at times. thats the reason I mentioned Crysis 3 because it taxes both my cpu and gpu. and yes I know its not the same as running IBT but the point is that a 220 watt rated cpu out of the box is ridiculous.

Your CPU at 4.4GHz will have more than 120W TDP and i dont know if your GPU is OCed or not but even at default the total will be 270W TDP (GTX660Ti = 150).

What is your Power consumption at the wall playing Crysis 3 ??
 

Abwx

Lifer
Apr 2, 2011
12,034
4,995
136
I was being fair (read kind) when I referenced trends. Besides, outside of real numbers, we only have AMD's own 220watts to go by, and the (wait for it) trend is, AMD's high end chips have always consumed more than rated TDP. Also the same piece of silicon from which these chips are derived as evidenced in the flagship 8350 consumes way above 220w when overclocked to 4.7Ghz. This is what I'm basing my argument on. What is yours? What is your justification for thinking the 220w rating is just an arbitrary figure? I'm waiting to read what you have to say. Bring your post count/history to bear on the answer, please. :D

As for my post count, what has that got to do with anything? You've been here only 2 years with 2,324 posts! How dare I have 24 posts in 7 years? Sorry dude, do you need a hug?

I posted above that theses chips will work at their base
frequencies with voltages slightly lower than the 8350
so the 20% frequency bump will forcibly increase TDP
by the same amount , while higher turbo frequency will
be reached with no more than 4.5% higher voltage
than 8350 , hence 10% more TDP on top of the 20%
mentionned , that s overall 35% over existing FXs ,
i estimate it at 40% at most assuming base voltage
is the same as FX , i.e , not reduced.
 

toyota

Lifer
Apr 15, 2001
12,957
1
0
http://www.anandtech.com/show/6910/origin-genesis-review-triple-titan-terror/6
For Pre-build High Performance Gaming System, 220W TDP for the CPU is not that much.
it is to many when a cpu using half the power is just as fast. of course it matters very little to overall power consumption if you are going to throw multiple high end gpus in the mix. for common people using a single gpu its a massive difference though. plus its easy as hell and quiet to keep a 100 watt cpu nice and cool.
 

Abwx

Lifer
Apr 2, 2011
12,034
4,995
136
Yet the FX triggers VRM throttle under heavy load, unlike Haswell.

It seems these 220W chips will be integrator only and not hit the channel. Guess thats how much faith it was given. A novelty product for PR only.

Apparently OEM only chips are not a problem when
it s something like the 4950HQ , i didnt read that
it was a novelty , but true that it s .....
 

Zucker2k

Golden Member
Feb 15, 2006
1,810
1,159
136
I posted above that theses chips will work at their base
frequencies with voltages slightly lower than the 8350
so the 20% frequency bump will forcibly increase TDP
by the same amount , while higher turbo frequency will
be reached with no more than 4.5% higher voltage
than 8350 , hence 10% more TDP on top of the 20%
mentionned , that s overall 35% over existing FXs ,
i estimate it at 40% at most assuming base voltage
is the same as FX , i.e , not reduced.
Your numbers are all over the place! So much for 2,326 posts!

Base frequencies = 4.4/4.7Ghz
Power Consumption is not linear on the frequency scale
Stop pulling numbers from your a*se, please

Fact: TDP = 220 watts!!!!!!!!!!!!
 

Abwx

Lifer
Apr 2, 2011
12,034
4,995
136
Your numbers are all over the place! So much for 2,326 posts!

Base frequencies = 4.4/4.7Ghz
Power Consumption is not linear on the frequency scale
Stop pulling numbers from your a*se, please

Fact: TDP = 220 watts!!!!!!!!!!!!

Lol , you are totaly cluless and just making fool of yourself ,
any second rate EE student know that TDP increase linearly
with frequency assuming constant voltage , it will increase
at a square law rate with voltage assuming constant frequency.

Keep on your insult register , that s the only thing
you re good at , obviously.
 

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
Yes, but the truth is these are going to be consuming more than the rated 220w TDP, if previous trends hold. :D

No, that isn't "the truth". My FX-8120 systems don't use anything close to even 120w during normal usage. The high power usage results obtained at Anandtech and other sites are from running programs that I never use, and I'd be willing to bet most home owners never use either. For gaming, browsing, office usage, even multitasking all of these at the same time, power usage doesn't even reached TDP. Only things like folding or encoding with all 8 cores cause the CPU to reach peak usage. How often do you encode? I never do, and I bet most other people don't encode more than an hour or two a week.

100W * 8760 hours (24 hours a day 365 day) = 876000 watt hours

876 kWh * $0.10 per kWh = $87.60 for 1 year continuous operation at the TDP difference (so assumes you use the exact workload that would closely approximate how AMD determines TDP)


See above: real world usage doesn't even reach TDP. Also, 24/7 usage? You don't sleep? For a realistic usage pattern, divide your numbers by 10. Because most people don't use a computer for more than 8 hours a day, and even while using a computer the majority of time the processor is idle waiting for data from the network connection, or hard disk, or memory. The CPU isn't usually the slowest part of the computer except for very specific workloads.

Now if you are the .001% of users who encode videos professionally 24/7, then your argument holds water. But trying to claim that such a usage pattern is normal or average is a bit flawed.
 
Last edited: