Personal computing discussed

Moderators: renee, morphine, SecretSquirrel

 
Chrispy_
Maximum Gerbil
Topic Author
Posts: 4666
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Hawaii power limit settings

Sun Mar 29, 2015 5:28 am

I returned from a week of fresh powder in the Alps and found that this comment had wormed its way onto the frontpage, so whilst stuck at home doing the laundry I picked the older of my two 290X cards and decided to see if there was any kind of sweet spot or useful information to be gleaned from charting it properly.

I'm just going to leave this here:

Image

Numbers were obtained with one of these, ASIC quality 79.8% (which doesn't appear to be anything special). I used Furmark loading the GPU to 100% in a 1280x720 window, and each clockspeed data point was obtained from the GPU-Z 0.8.2 logfile using the average of at least one minute of sampling at 0.1s refresh rate (so 600+ samples per data point).

I also checked that setting the clock limit to just below that average speed in the results table gave me a stable core clock that didn't boost or throttle during Furmark. The two things that stand out for me after graphing the results are:
  • There's a definite curve to the performance drop-off, and at 1000MHz, there's an awful lot (22%) of power envelope that's unnecessary.
  • The largest perpendicular distance between the two lines when all axes are plotted from zero is in the 850-900MHz region, indicating that this is (my) Hawaii's most efficient clockspeed, and probably the clockspeed that the 290X should have launched at in a world without the 780 and Titan.
To those of you fed up with hot/noisy Hawaii cards, or just curious and willing to experiment, give this a whirl and post what you get.

Don't feel you have to plot a graph or sample every single powertune setting, but I guess it would be interesting to see at least two things:
  1. How little power do you need to run at, say, 900MHz?
  2. What can your card can do with a 200W TDP instead of the default 290W TDP (-31% power limit)?
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
anotherengineer
Gerbil Jedi
Posts: 1675
Joined: Fri Sep 25, 2009 1:53 pm
Location: Northern, ON Canada, Yes I know, Up in the sticks

Re: Hawaii power limit settings

Sun Mar 29, 2015 8:14 pm

Nice work.

My HD 6850 stock max core is 775MHz. Simple optimization, and it's too bad AMD didn't do a better job with voltage vs. freq vs. power consumption vs.performance optimization.

I'd guess 875MHz to 925MHz would have been somewhere in that sweet spot region. It's too bad we can't enter the VGA BIOS like the mobo bios and adjust it like we can with the CPU.
Life doesn't change after marriage, it changes after children!
 
CeeGee
Gerbil
Posts: 76
Joined: Sun Dec 22, 2013 11:16 am

Re: Hawaii power limit settings

Mon Mar 30, 2015 12:11 am

First of all thanks Chrispy_ for making this post. I agree that the default Powertune settings are too high and it's an interesting and important knob to turn if you're a hardware enthusiast or overclocker, however I can not get anywhere near your numbers in Furmark and I find the choice of Furmark a bit odd for this but oh well.

I have an MSI R9 290 that apparently has a 275w TDP. MSI Afterburner informs me this card has a +13mV boost over the default voltage from the factory.

* All testing done at 720p windowed with AA off.

If I leave everything at default and lower the Powertune setting by 27% to get about 200w TDP my core clock averages (over 60 seconds) 702mhz in the Furmark stress tester.

With my usual voltage setting of -37mV in MSI Afturburner and -27% in Powertune my core clock averages 816mhz in the Furmark stress tester.

My every day settings of -37mV and -15% Powertune (with a core clock set to 1040mhz) result in an average core clock of 879mhz in Furmark.

So in answer to your questions then...
1. I'd guess my card needs around 240w at my -37mV setting to run at 900mhz in Furmark and 253w with the factory overvolt.
2. My card can do 702mhz at 200w with it's factory overvolt and 816mhz with my preferred -37mV again at 200w.

My observations. Well I can't get anywhere near your numbers even if I lower my GPU volts I get 816mhz versus your 957mhz! I should also point out that at my preferred settings of -37mV, -15% Powertune and 1040mhz core clock I don't see throttling in any other software that I've tried. This includes 3DMark, Unigine Valley and games that I own. Here is a 3DMark Fire Strike score (9827) I got with these settings which as far as I can tell is right about where it should be.

* Edit to clarify settings.
Last edited by CeeGee on Mon Mar 30, 2015 6:39 pm, edited 1 time in total.
 
jihadjoe
Gerbil Elite
Posts: 834
Joined: Mon Dec 06, 2010 11:34 am

Re: Hawaii power limit settings

Mon Mar 30, 2015 1:06 am

@CeeGee
Can you post the ASIC quality of your card? Curious to see how that affects clocks/TDP.
 
CeeGee
Gerbil
Posts: 76
Joined: Sun Dec 22, 2013 11:16 am

Re: Hawaii power limit settings

Mon Mar 30, 2015 1:16 am

My cards ASIC quality is 78.6%.

It's not unreasonable to think that better chips go into the X cards but I still find the difference between Chrsipy_'s results and mine puzzling.
 
geekl33tgamer
Graphmaster Gerbil
Posts: 1022
Joined: Tue Aug 25, 2009 7:25 pm
Location: England
Contact:

Re: Hawaii power limit settings

Mon Mar 30, 2015 4:35 am

Mine are "hot clocked" out the gate by XFX to 1050Mhz Core / 5400Mhz Memory, and my first one fully loaded will hold that core speed rock steady (it won't when I turn CF back on). It has a small voltage bump of +27mV hardwired into it's BIOS, and is being apply right across the board for the results below.

Seeing as XFX in their infinite wisdom wont allow voltage changes, I just left it all as it was and lowered Powertune to -31% to get approximately 200W TDP. The core speed was rock steady at 972Mhz. To get it holding as close to 900Mhz as it would, I had to lower Powertune all the way down to -43% to get 898Mhz reported. If I used -42%, it would hold a steady clock of 904Mhz.

Looks like that permanent voltage bump is increasing those clocks a touch, and the TDP may not be quite right as a result. If anyone can take that into account and let us know compensated Powertune profiles to run at, I'll sort you a beer! :lol:

Edit: Never knew GPU-Z did this! The first card's ASIC quality is 82.1%, 2nd is 81.3%, 3rd is 81.9%.
Mega Beast - Intel i7 4790K | Gigabyte Z97X-UD3H-BK | 32GB DDR3 | SLI GTX 1070 | Samsung 850 Evo 1TB
Mini Beast - Intel C2E QX9770 | Gigabyte X48T-DQ6 | 16GB DDR3 | KFA2 LP GTX 750Ti | Seagate 2TB SSHD
 
vargis14
Gerbil Jedi
Posts: 1900
Joined: Fri Aug 20, 2010 6:03 pm
Location: philly suburbs

Re: Hawaii power limit settings

Mon Mar 30, 2015 7:14 am

I tried telling you Geekster when I was telling you to use yout 2 best cards...but it looks like all your cards are pretty darn close that pretty lucky. To have them all within less then 1% of each other is pretty rare.
2600k@4848mhz @1.4v CM Nepton40XL 16gb Ram 2x EVGA GTX770 4gb Classified cards in SLI@1280mhz Stock boost on a GAP67-UD4-B3, SBlaster Z powered by TX-850 PSU pushing a 34" LG 21/9 3440-1440 IPS panel. Pieced together 2.1 sound system
 
MEATLOAF2
Gerbil
Posts: 71
Joined: Tue Apr 16, 2013 11:40 pm

Re: Hawaii power limit settings

Mon Mar 30, 2015 7:58 am

The ASIC quality on my 290 is 70.8% :( that's too bad for me I suppose. I've gone ahead and dropped the power by 15% in CCC, not gonna do any testing, but if it's similar to the graph, I shouldn't notice any performance penalties, with a decent drop in power consumption.

If anyone is up for it, I'd like to see what effect this has on temperature.
 
geekl33tgamer
Graphmaster Gerbil
Posts: 1022
Joined: Tue Aug 25, 2009 7:25 pm
Location: England
Contact:

Re: Hawaii power limit settings

Mon Mar 30, 2015 8:12 am

Well, I'm not near my PC to go back over the captured data, but from memory I don't recall any temps higher than 77C. The cards usually shoot straight for the mid 90's, so yeah - that 100w less makes a very noticeable difference to the temps.

Edit: Remember this was run with a single card though. I usually have crossfire enabled and they run hotter with that on for obvious reasons. ;)
Mega Beast - Intel i7 4790K | Gigabyte Z97X-UD3H-BK | 32GB DDR3 | SLI GTX 1070 | Samsung 850 Evo 1TB
Mini Beast - Intel C2E QX9770 | Gigabyte X48T-DQ6 | 16GB DDR3 | KFA2 LP GTX 750Ti | Seagate 2TB SSHD
 
geekl33tgamer
Graphmaster Gerbil
Posts: 1022
Joined: Tue Aug 25, 2009 7:25 pm
Location: England
Contact:

Re: Hawaii power limit settings

Mon Mar 30, 2015 8:21 am

Ninja double post.

I'll test temps with crossfire when I get back home. My gut feeling is that with the bottoms of cards 3 and 2 flush against the tops of card 2 and 1, the temps wont be wildly different.
Mega Beast - Intel i7 4790K | Gigabyte Z97X-UD3H-BK | 32GB DDR3 | SLI GTX 1070 | Samsung 850 Evo 1TB
Mini Beast - Intel C2E QX9770 | Gigabyte X48T-DQ6 | 16GB DDR3 | KFA2 LP GTX 750Ti | Seagate 2TB SSHD
 
MEATLOAF2
Gerbil
Posts: 71
Joined: Tue Apr 16, 2013 11:40 pm

Re: Hawaii power limit settings

Mon Mar 30, 2015 9:10 am

I decided to do some testing, and had some odd results.

Using FurMark, on the default 720p window, my 290 (XFX DD) throttles at -15% power limit settings in CCC. However, if I turn on 8xMSAA it runs at the full 947MHz (0x, 2x, and 4xMSAA and it still throttles). This is regardless of temperature, and 100% GPU load being reported by GPUZ. I'll mess around with it some more and see what happens.

EDIT: Ran a few more tests, and logged the max core clock for each, using GPUZ. The only thing that changed was the MSAA amount in FurMark. The temps were below 70C for each run (I didn't run them for very long). Most of the time, the core clocks bounced around well below the listed max, except for 8xMSAA, where it mostly stayed at 1089MHz.

CCC Settings:
-15% Power Limit
+15% Core Clock

FurMark:
0xMSAA
Max Clock: 741MHz

2xMSAA
Max Clock: 799MHz

4xMSAA
Max Clock: 861MHz

8xMSAA
Max Clock: 1089MHz

EDIT 2: It's worth noting I'm using 15.3 drivers I suppose.
 
Topinio
Gerbil Jedi
Posts: 1758
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Hawaii power limit settings

Mon Mar 30, 2015 10:42 am

Nice post. I would test mine, but it is unstable without a +4% power setting since 14.8 (?). (XFX 290X, reference cooler and clocks FWIW, ASIC quality 75.2)
Desktop: E3-1270 v5, X11SAT-F, 32GB, RX 580, 500GB Crucial P1, 250GB Crucial MX500, 4TB 7E8, Xonar DGX, XL2730Z + L22e-20
HTPC: i5-2500K, DH67GD, 6GB, GT 1030, 250GB MX500, 1.5TB Barracuda, Xonar DX, G2420HDB
Laptop: MacBook6,1
 
Chrispy_
Maximum Gerbil
Topic Author
Posts: 4666
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Hawaii power limit settings

Mon Mar 30, 2015 11:50 am

Hmm, interesting, I didn't realise there was variance based on MSAA. I picked Furmark because it was a worst-case scenario. Running actual games I find the average GPU-Z reported clockspeed is much higher than Furmark clockspeeds, which is just fine by me ;)

I had no idea what the furmark defaults were (there's no button!) so I would have been running at 4x MSAA just because that's what it was set on at the time I did this.

Windowed 1280x720
4x MSAA
Dynamic Background ON
Dynamic Camera OFF
Post-FX OFF

I'm not home tonight, will have to check the lower 290X again to see what it gets without MSAA.
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
MEATLOAF2
Gerbil
Posts: 71
Joined: Tue Apr 16, 2013 11:40 pm

Re: Hawaii power limit settings

Mon Mar 30, 2015 12:03 pm

I'm interested to see the results, because at 0% power reduction, and no overclock, I was seeing some throttling, and it was fairly unsettling. I'm hoping it's just an issue with FurMark (I almost never use it so I don't know any of it's quirks), and not my card. :)

Perhaps running a game or a benchmark like Firestrike would be a better way to determine real world clock speeds.

For what it's worth, I ran a few games and MSAA or not, it ran at ~1089MHz with -15% power limit, so it might just be and issue with FurMark.
 
ultima_trev
Gerbil XP
Posts: 363
Joined: Sat Mar 27, 2010 11:14 am
Contact:

Re: Hawaii power limit settings

Mon Mar 30, 2015 12:25 pm

Wait, so you're saying an R9 290X can enjoy performance per watt comparable to the GTX 970 with a mere reduction to the Power Tune settings in CCC? Are you actually confessing Hawaii isn't a joke architecture like GTX 480 and FX 5800 before it??!!

Impossibru, I say!
Ryzen 7 1800X - Corsair H60i - GA AB350 Gaming - 32GB DDR4 2933 at 16,16,16,36 - GTX 1080 at 1924 / 5264 (undervolted) - 250GB WD Blue SSD - 2TB Toshiba 7200rpm HDD
 
CeeGee
Gerbil
Posts: 76
Joined: Sun Dec 22, 2013 11:16 am

Re: Hawaii power limit settings

Mon Mar 30, 2015 6:38 pm

Just to clarify all of my testing so far was done at 720p windowed with AA off.
 
CeeGee
Gerbil
Posts: 76
Joined: Sun Dec 22, 2013 11:16 am

Re: Hawaii power limit settings

Mon Mar 30, 2015 10:03 pm

I've had time to retest Furmark at 4xAA, results below along with my previous AA off results.

Furmark 720p windowed with AA off. Mhz is an average over 60s with 0.1s sample rate.
Voltage | TDP | Mhz
+13mV | 200w | 702Mhz
-37mV | 200w | 816Mhz

Furmark 720p windowed with 4xAA on. Mhz is an average over 60s with 0.1s sample rate.
Voltage | TDP | Mhz
+13mV | 200w | 739Mhz
-37mV | 200w | 892Mhz

As MEATLOAF2 noticed it seems like the exact settings used can have a dramatic effect on power draw. I'd be interested in seeing Chrispy_ and geekl33tgamer's results with AA off, I suspect they'll still be clocking higher than my setup though. :D

Furmark might be useful as a tool to explore Powertune behavior in extreme cases but I don't find it representative of any other software I own. I've been tinkering with voltage and powertune on my card for about a year now and I've settled on some settings (-37mV, -15% powertune and 1040mhz GPU clock) for my particular card that don't throttle at all in the games and software I use. If you like tinkering with this kind of thing then I think it's worth doing, it's certainly possible to improve on the default settings the cards come with.
 
ultima_trev
Gerbil XP
Posts: 363
Joined: Sat Mar 27, 2010 11:14 am
Contact:

Re: Hawaii power limit settings

Tue Mar 31, 2015 7:47 am

Has anyone tried gaming with these decreased power tune settings? For example, could an R9 290X sustain 1 GHz core at -10% or -20% (and perhaps undervolted) in, say, Crysis 3 or Far Cry 4? I think this could potentially become front page article worthy, proving GCN isn't inherently power hungry, just poorly tuned from AMD/board partners.
Ryzen 7 1800X - Corsair H60i - GA AB350 Gaming - 32GB DDR4 2933 at 16,16,16,36 - GTX 1080 at 1924 / 5264 (undervolted) - 250GB WD Blue SSD - 2TB Toshiba 7200rpm HDD
 
MEATLOAF2
Gerbil
Posts: 71
Joined: Tue Apr 16, 2013 11:40 pm

Re: Hawaii power limit settings

Tue Mar 31, 2015 9:18 am

ultima_trev wrote:
Has anyone tried gaming with these decreased power tune settings?


I haven't done any performance testing in games, but I did just run the 3dMark Firestrike benchmark in 5% intervals, starting from -30% up to 0% power limit. I'll post the results here in a while, although I'm not great at managing the data and such, so I'll probably just list the scores, and link to the results. Unfortunately I didn't log power usage, or clock speeds. Also, I only ran one benchmark per power limit setting, so there could be a good amount of variance.

One odd thing I noticed is the physics score, it went down as I increased the power limit, but the overall (and graphics) score went up.
 
MEATLOAF2
Gerbil
Posts: 71
Joined: Tue Apr 16, 2013 11:40 pm

Re: Hawaii power limit settings

Tue Mar 31, 2015 9:51 am

Well, here are some 3dMark Firestrike scores. I'll start by listing a few specs I guess.

CPU: i5 3570K (4.2Ghz Turbo)
GPU: R9 290 (XFX DD 947Mhz, 1089Mhz OC using CCC)
GPU Drivers: 15.3 beta
OS: Windows 7 64bit

I'll just jump right into it I guess, I'll post the overall, graphics and physics score for each run, and also link to the result for anyone interested in the finer details, like the fps for each part of the test etc. All tests were done with a 15% OC on the GPU core clock, and the memory clock was left at the default 1250Mhz, this is because that's pretty much what I run it at 24/7. Later I will perhaps run it at the default clocks, and do 3 runs per power setting, maybe..

The only changing factor in all of the tests was the power limit settings in CCC, everything else was left untouched, and the test was using the default Firestrike settings (not extreme or ultra). Here we go:

-30% power limit: http://www.3dmark.com/3dm/6455007
3DMark Score: 7749
Graphics Score: 9233
Physics Score: 7784

-25% power limit: http://www.3dmark.com/3dm/6455056
3DMark Score: 8222
Graphics Score: 9931
Physics Score: 7548

-20% power limit: http://www.3dmark.com/3dm/6455115
3DMark Score: 8533
Graphics Score: 10556
Physics Score: 7041

-15% power limit: http://www.3dmark.com/3dm/6455168
3DMark Score: 8843
Graphics Score: 11134
Physics Score: 6698

-10% power limit: http://www.3dmark.com/3dm/6455240
3DMark Score: 8844
Graphics Score: 11309
Physics Score: 6342

-5% power limit: http://www.3dmark.com/3dm/6455276
3DMark Score: 8844
Graphics Score: 11373
Physics Score: 6272

-0% power limit: http://www.3dmark.com/3dm/6455328
3DMark Score: 8892
Graphics Score: 11365
Physics Score: 6484

Later I'll try to run some games that have built in benchmarks, for consistency. Probably the Tomb Raider benchmark or something.
 
ultima_trev
Gerbil XP
Posts: 363
Joined: Sat Mar 27, 2010 11:14 am
Contact:

Re: Hawaii power limit settings

Tue Mar 31, 2015 10:05 am

Impressive results, seems to be most efficient -10% to -20%, very slight decrease in performance if it scales anything like Furmark.

Edit: Very impressive results actually, that's just a 290 (I read 290X at first) and it breaks 10k at -20%??????????! Beastly. My 7850 barely breaks 5k overclocked, where its TDP should be roughly 145 versus 230 for the R9 290 at -20% according to Chrispy's graph. So in that regards it's roughly 100% more performance for a mere 50% increase to TDP... That isn't bad at all, especially considering it has double the VRAM in addition to the bigger die.
Ryzen 7 1800X - Corsair H60i - GA AB350 Gaming - 32GB DDR4 2933 at 16,16,16,36 - GTX 1080 at 1924 / 5264 (undervolted) - 250GB WD Blue SSD - 2TB Toshiba 7200rpm HDD
 
MEATLOAF2
Gerbil
Posts: 71
Joined: Tue Apr 16, 2013 11:40 pm

Re: Hawaii power limit settings

Tue Mar 31, 2015 10:44 am

My 290 is kind of heavily overclocked though, at almost 1100MHz, I've tried to push it further but it's at about it's upper limit, at least as far as I can get it using CCC, I don't really feel like upping the voltage in MSI afterburner or anything, since it performs well enough as is.

I actually had a 7850 before the 290, it's not a bad card at all, I was happy with it. Here is what I think is my best score with the 7850 if you are interested:

http://www.3dmark.com/fs/1740931

Most of the parts that matter in the machine were the same when I ran that benchmark, and it was probably heavily overclocked.
 
sweatshopking
Graphmaster Gerbil
Posts: 1463
Joined: Fri Aug 15, 2008 10:37 am

Re: Hawaii power limit settings

Tue Mar 31, 2015 10:51 am

my asic quality is 88%, but i haven't had a ton of success dropping my gpu speeds. i have an msi 290 gaming 4g, but i get freezes, crashes and hangs if i drop by like 20% in powertuner. I don't have a lot of experience with asic quality or power adjustments to GPU's, so i'm not really sure what i'm doing besides just moving the slider in CCC
 
MEATLOAF2
Gerbil
Posts: 71
Joined: Tue Apr 16, 2013 11:40 pm

Re: Hawaii power limit settings

Tue Mar 31, 2015 11:14 am

sweatshopking wrote:
my asic quality is 88%, but i haven't had a ton of success dropping my gpu speeds. i have an msi 290 gaming 4g, but i get freezes, crashes and hangs if i drop by like 20% in powertuner. I don't have a lot of experience with asic quality or power adjustments to GPU's, so i'm not really sure what i'm doing besides just moving the slider in CCC


That's odd, all I'm doing is adjusting the "power limit settings" (not "GPU clock settings") in CCC, in AMD overdrive, and, to verify, I ran another bench, at -50% power. Besides horrible performance, it was perfectly stable, even with a 1089Mhz core, 1275Mhz memory overclock. At least I wasn't getting any graphics corruption like when I overclock it too far: http://www.3dmark.com/3dm/6456764

My ASIC quality is 70.8%. I'll leave it at -50% power limit for a while to see if I notice any problems or see any crashing during normal use.
 
Chrispy_
Maximum Gerbil
Topic Author
Posts: 4666
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Hawaii power limit settings

Tue Mar 31, 2015 1:12 pm

AFAIK it shouldn't crash with a -50% power limit.

Let's face it, it runs at idle at 300MHz using about 25W or something!
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
anotherengineer
Gerbil Jedi
Posts: 1675
Joined: Fri Sep 25, 2009 1:53 pm
Location: Northern, ON Canada, Yes I know, Up in the sticks

Re: Hawaii power limit settings

Tue Mar 31, 2015 1:50 pm

sweatshopking wrote:
my asic quality is 88%, but i haven't had a ton of success dropping my gpu speeds. i have an msi 290 gaming 4g, but i get freezes, crashes and hangs if i drop by like 20% in powertuner. I don't have a lot of experience with asic quality or power adjustments to GPU's, so i'm not really sure what i'm doing besides just moving the slider in CCC


Mail it to me, I'll try it in my PC :)

Also I believe MSI has a dip switch to select which GPU BIOS to access. I believe one GPU BIOS is for conventional mobo BIOS and the other GPU BIOS is for UEFI mobo BIOS.

I don't own an MSI card, so you would have to check and confirm yourself.
Life doesn't change after marriage, it changes after children!
 
sweatshopking
Graphmaster Gerbil
Posts: 1463
Joined: Fri Aug 15, 2008 10:37 am

Re: Hawaii power limit settings

Tue Mar 31, 2015 1:57 pm

maybe my 88% asic isn't a good quality one? that seems low, but the ones in the thread seem shockingly lower... i was thinking 88% was horrible.
 
Chrispy_
Maximum Gerbil
Topic Author
Posts: 4666
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Hawaii power limit settings

Tue Mar 31, 2015 2:19 pm

IME ASICs over 85% are as rare as asics under 65%

I'm not entirely sure what the correlation between the ASIC quality value and over/underclocking is though. I think it's a good general indicator but I've seen low-quality cards clock to the moon and I've seen high-quality cards fail to clock more than 10%
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
anotherengineer
Gerbil Jedi
Posts: 1675
Joined: Fri Sep 25, 2009 1:53 pm
Location: Northern, ON Canada, Yes I know, Up in the sticks

Re: Hawaii power limit settings

Tue Mar 31, 2015 2:36 pm

sweatshopking wrote:
maybe my 88% asic isn't a good quality one? that seems low, but the ones in the thread seem shockingly lower... i was thinking 88% was horrible.



check this maybe?

http://event.msi.com/2014/vga/Hybrid_BIOS/
Life doesn't change after marriage, it changes after children!
 
Topinio
Gerbil Jedi
Posts: 1758
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Hawaii power limit settings

Tue Mar 31, 2015 3:23 pm

Chrispy_ wrote:
AFAIK it shouldn't crash with a -50% power limit.

Let's face it, it runs at idle at 300MHz using about 25W or something!

Maybe Hawaii needs the juice to be stable? My 290X certainy produces crashes, freezes and hangs at default power settings and I need +4% power then the box is rock solid stable. This has been the case since the summer, previous drivers did not give me this problem (and yes, it took me a while to work out).

Maybe this is a factor in the power limit being set so high: enough cards would -- for whatever weird reason -- be flakey at lower TDP that AMD needs to spec them power hungry to shift volume of workable cards?
Desktop: E3-1270 v5, X11SAT-F, 32GB, RX 580, 500GB Crucial P1, 250GB Crucial MX500, 4TB 7E8, Xonar DGX, XL2730Z + L22e-20
HTPC: i5-2500K, DH67GD, 6GB, GT 1030, 250GB MX500, 1.5TB Barracuda, Xonar DX, G2420HDB
Laptop: MacBook6,1

Who is online

Users browsing this forum: No registered users and 1 guest
GZIP: On