Richland & Kabini rumours

[ShintaiDK]

I see the kübler-Ross model got another rerun.

Its all on the table, yet some still deny it or behaves very aggressively.

 

[Abwx]

The only thing that is agressive is your post that is irrelevant
in respect of a technical discussion.

One who has not the necessary technical knowledge will forcibly
try to displace the discussion in personnal attacks , that s the syndrom.

 

[ShintaiDK]

The only thing that is agressive is your post that is irrelevant
in respect of a technical discussion.

One who has not the necessary technical knowledge will forcibly
try to displace the discussion in personnal attacks , that s the syndrom.

You should start read your own posts then, and see where that puts you.

 

[Phynaz]

But people don't want to hear it, and they won't stand for anyone else who might talk about it. Don't waste your breath.

You are correct once again, and your advice is solid. Time for me to bail out of this discussion.

The best we can do with our knowledge at this point is keeping an uninformed novice from making a bad purchase decision.

 

[Abwx]

You should start read your own posts then, and see where that puts you.

My recent posts are filled with technical explanations about switching
mode PSUs properties that are not known by the general public,
while yours , as the two last ones, are personnal attacks.

Stick to the discussion , or if you dont know how PSUs work just ask
but do not post if it is just to disrupt the subject because you dont
understand my technical explanations.

 

[Charles Kozierok]

Google "power factor".....

Google "power factor correction".

 

[Ajay]

Is it possible to make a new thread for this stuff next time?

Possible, yes. Likely, not so much - this is SOP @ AT. It does get tiresome for sure.

 

[Abwx]

Google "power factor correction".

If you had read my posts you would have seen that i mentionned
active PF correction , but neverless , without PFC the PSU can have
as low as 0.7 PF and PFC wont increase it far above 0.95 , and that is
with a minority of the PSUs.

 

[Charles Kozierok]

If you had read my posts you would have seen that i mentionned
active PF correction , but neverless , without PFC the PSU can have
as low as 0.7 PF and PFC wont increase it far above 0.95 , and that is
with a minority of the PSUs.

Sorry if I missed something, it's a big thread.

Active PFC is becoming pretty standard and I think eventually all PSUs except for the really cheap crap will have it. I just looked at a PC Power and Cooling unit that was spec'ed at 0.98. Antec has it on several of their lines.

I suppose we'd need to know what PSU was involved in these tests to figure out whether or not PF is an issue.

 

[Khato]

I suppose we'd need to know what PSU was involved in these tests to figure out whether or not PF is an issue.

Or know if for some incomprehensible reason they were measuring apparent power instead of actual. Because even the basic Kill A Watt model can give you either measurement - press the button once and you get actual delivered power in watts while if you press it again you get the apparent power in VA. I'd be somewhat surprised if there was a model on the market that only reports apparent power given that such isn't of any concern to consumers, because ya know... your utility bill is based on real power in watts, not apparent power. Unless the utility company is one of those that charges a penalty for customers who have too low of a power factor that is.

 

[tweakboy]

Disneyland is coming OMG!,, go big RED... lol

 

[Charles Kozierok]

Either way, this is pretty hard to argue with. I trust motherboard engineers to be able to measure power consumption.

 

[Khato]

Either way, this is pretty hard to argue with. I trust motherboard engineers to be able to measure power consumption.

Pretty much. It's amusing to watch the many straws that are grasped upon to try and explain why a processor exceeding its specified thermal design power isn't the processor's fault.

 

[Idontcare]

So again, 190W is not only the CPU power usage difference in the measurements from the wall.

I have never disagreed with that analysis, I have disagreed with the conclusions drawn from it.

If the conclusions you and Abwx draw regarding power consumption data, TDP, and the 125W FX-8350 were correct then the entire situation with MSI would be moot because no one's processors would be throttling for having hit >125W power usage.

You guys are fixating on a red herring here, trying to divide and conquer every anecdotal observation that exists, one by one, as being something other than what the sum total of all those observations are pointing to.

And then you've got the MSI situation which is about as cut-and-dry as it gets. If there was no smoke then there would be no fire, MSI says there is a fire and they corroborate our observations of smoke by even linking directly to them.

If processors were not exceeding 125W on the MSI mobos then the power-consumption circuitry would not be getting triggered in the first place, and yet we have multiple reports of members seeing fictitious "255°C" temperature readings on their MSI board, exactly as MSI engineers said would happen if the board detected the CPU using excessive power.

Why are FX-8350's throttling on the MSI boards when running standard encoding apps if those cpus are not exceeding 125W per the engineers observations? Why are we even having this debate, the authorities on the matter have already spoken. What is left to discuss?

 

[Idontcare]

IDC, could you run x264 HD5.0.1 and tell us your energy readings with Voltages on AUTO ??

No problem, here are the results:

Run 1: Pass 1 of 2
50-60% CPU utilization, 4.1GHz, 1.4125VID, 1.360V actual, 208-212W, 27C/67.1F ambient
Run 1: Pass 2 of 2
89-100% CPU utilization, 4.0GHz, 1.3875VID, 1.322V actual, 235-244W, 31C/68.0F ambient

For comparison I ran the same on the 2600K:

Run 1: Pass 1 of 2
3.8GHz, 50-60% CPU, 72C, 148-154W, 1.3411V
Run 1: Pass 2 of 2
3.8GHz, 67-100% CPU, 81C, 166-174W, 1.3461V

The turbo never turned off for the 2600K, not by choice as the BIOS was set to auto on all the power features.

The 2600K was running with stock Intel HSF, hence the temps, whereas the 8350 was running with an NH-D14.

 

[Piroko]

Because even the basic Kill A Watt model can give you either measurement - press the button once and you get actual delivered power in watts while if you press it again you get the apparent power in VA.

Kill-a-Watts can still report false values if the sine wave is distorted, just saying.

If processors were not exceeding 125W on the MSI mobos then the power-consumption circuitry would not be getting triggered in the first place, and yet we have multiple reports of members seeing fictitious "255°C" temperature readings on their MSI board, exactly as MSI engineers said would happen if the board detected the CPU using excessive power.

Newegg has 65 AM3+ Motherboards listed, why is the MSI one the only one which has trouble running the 8350? Do you have proof that the circuitry throttles at or above 125W? That only the Processor itself is powered by those VRMs? That the board or bios itself isn't faulty and setting more Voltage than needed?

We've seen boards before where a simple bios update shaved off or added 20 Watts to the system power consumption and we've seen plenty of boards before where the manufacturer screwed up in one way or another.

 

[mrmt]

Newegg has 65 AM3+ Motherboards listed, why is the MSI one the only one which has trouble running the 8350? Do you have proof that the circuitry throttles at or above 125W? That only the Processor itself is powered by those VRMs? That the board or bios itself isn't faulty and setting more Voltage than needed?

Count two, as some AsRock boards appears to have the same problem:

http://forums.overclockers.com.au/showthread.php?p=15100731

And so far no plausible explanation on why the 8350 is the only 125W AM3+ processor that throttles on those boards.

 

[bgt]

http://forums.overclockers.com.au/showthread.php?p=15100731

Good to know. Will have a look later. Saw some throttling also when using OCCT. I noticed the VRM heatsink uses thermal pads. Maybe thermal glue is better(like I did on the small ITX board from ASROCK). Have to do a temp. test on the VRM heatsink first.

update: burned my fingers on the heatsink when doing prime95. Temp was 89C after 2 minutes. Heatsink is far too small. It throttles back to 1400mhz.

 

[NTMBK]

No problem, here are the results:


For comparison I ran the same on the 2600K:


The turbo never turned off for the 2600K, not by choice as the BIOS was set to auto on all the power features.

The 2600K was running with stock Intel HSF, hence the temps, whereas the 8350 was running with an NH-D14.

I hate to suggest this to a mod, but could the (interesting and valid) discussion on 8350 power draw go into a separate topic? This one's not talked about Kabini and Richland much lately

 

[bgt]

I hate to suggest this to a mod, but could the (interesting and valid) discussion on 8350 power draw go into a separate topic? This one's not talked about Kabini and Richland much lately

good id:thumbsup:

 

[Piroko]

Count two, as some AsRock boards appears to have the same problem:

http://forums.overclockers.com.au/showthread.php?p=15100731

And so far no plausible explanation on why the 8350 is the only 125W AM3+ processor that throttles on those boards.

This board also apparently needs more voltage for the same frequency as other boards. Some reported that they couldn't overclock at all on the Asrock board, saw BSODs or funky behaviour at stock speeds or had to apply ridiculous voltages to get to 4.4 Ghz while other boards can hit 4.6 Ghz without much trouble.

So yea, that board can either be weighted against AMD (8350 out of spec) or against MSI (shoddy board). Or perhaps against both.

 

[AtenRa]

Five Kabini APUs planned for June 2013

 

[Piroko]

Hmm those names do not match with http://forums.anandtech.com/showpost.php?p=34621830&postcount=1357

 

[Idontcare]

If processors were not exceeding 125W on the MSI mobos then the power-consumption circuitry would not be getting triggered in the first place, and yet we have multiple reports of members seeing fictitious "255°C" temperature readings on their MSI board, exactly as MSI engineers said would happen if the board detected the CPU using excessive power.

Newegg has 65 AM3+ Motherboards listed, why is the MSI one the only one which has trouble running the 8350? Do you have proof that the circuitry throttles at or above 125W? That only the Processor itself is powered by those VRMs? That the board or bios itself isn't faulty and setting more Voltage than needed?

We've seen boards before where a simple bios update shaved off or added 20 Watts to the system power consumption and we've seen plenty of boards before where the manufacturer screwed up in one way or another.

Proof? Go ask MSI. It is not my job to prove anything here.

If you take MSI to be liars or incompetent then I'm confident you can track them down and deftly vanquish the injustice. No need to have me be the middle-man in that showdown.

 

[pablo87]

Kill-a-Watts can still report false values if the sine wave is distorted, just saying.
Newegg has 65 AM3+ Motherboards listed, why is the MSI one the only one which has trouble running the 8350? Do you have proof that the circuitry throttles at or above 125W? That only the Processor itself is powered by those VRMs? That the board or bios itself isn't faulty and setting more Voltage than needed?

We've seen boards before where a simple bios update shaved off or added 20 Watts to the system power consumption and we've seen plenty of boards before where the manufacturer screwed up in one way or another.

I agree with you - unless the problem is across multiple mb vendors and models, you have to investigate msi's design. My experience with them has not been good quality wise.