NVIDIA GeForce GTX 980 Ti

Very interesting! Was considering water cooling mine. If I'm reading it correctly, it looks like there's less value in doing so than I'd thought. Mine has 60.4%
 
Last edited:
Huum, well it iss allways the same, low asic quality accept more voltage ( as they leak more electricity ) but need better cooling for it, ( Dice or LN2 ).. ( same for AMD gpus or any chips ).. the problem in the middle is for the watercooled or chilled water one ( as you are in a middle cooling position ( not subzero ) so your overclocking will depend badly..

This said, ASIC quality is not a warranty at all.. you can have 90% + Asic who will hit a wall on Air faster than a 75% one and got 75% Asic who will oc as bad as a 85% one under LN2... but i expect they have bin well their chips on the " week of fabrication " for collect as normalized sample as they can "-- well i hope it...

You know whatever it is for air, water or Ln2, for cpu or gpu's overclockers still need find the golden sample, and this mean trying as many of parts as you can...

Personally, im a bit dubitative about this method from Evga .. specially when price differ from a Asic quality to another.. ( Seriously i see it a bit like a marketing plots, but i can be wrong )

By the past we was not look at the Asic, but at the week of fabrication...
 
Last edited:
Considering I can run it at +200c, +300m at stock voltages I'm not really that fussed about the low ASIC %. Just makes me wonder 'what if?' :)

These cards are monsters.
 
Not sure how far you got into the thread, but KingP|n does respond later to what the OP mentioned. Not really sure where marketing comes into play here, but considering some of the other reviewed 980 Ti's and their OC's (especially the Review: Zotac GeForce GTX 980 Ti AMP! Extreme review in post #84 above) I'd say it's part marketing to get hardcore OC'ers on board for the higher ASIC models.

TBH you can still put yours under water and get an excellent, stable OC. I've had about 4 780 Classy's, all air cooled with ASIC ranging from 62% - 91%. The only difference on air between low and high ASIC was more voltage tweaks was required on low ASIC cards to achieve the same clock, and higher ASIC could ultimately achieve higher clocks.
 
What is "ASIC Quality" as a percentage value even measuring? How is it set? At test and packaging time? Using what criteria/formula? Seems like an utterly bonkers way to normalise the bin curve for frequency and voltage. What am I missing?
 
Not sure how far you got into the thread, but KingP|n does respond later to what the OP mentioned. Not really sure where marketing comes into play here, but considering some of the other reviewed 980 Ti's and their OC's (especially the Review: Zotac GeForce GTX 980 Ti AMP! Extreme review in post #84 above) I'd say it's part marketing to get hardcore OC'ers on board for the higher ASIC models.

TBH you can still put yours under water and get an excellent, stable OC. I've had about 4 780 Classy's, all air cooled with ASIC ranging from 62% - 91%. The only difference on air between low and high ASIC was more voltage tweaks was required on low ASIC cards to achieve the same clock, and higher ASIC could ultimately achieve higher clocks.

No subzero overclockers want a high ASIC ( serious overclockers who use subzero cooling or extremely good cooling setup ) ... High ASIC are for aircooling ( stock one ) , moderate watercooling ( as i have said, watercooling is a complicate case anyway for overclocking since some years ) .... Its harder to find Low Asic than High ASIC.. .. and so, the fact they charge more for high ASIC make me think they want to attire the basic consumers who want a good stock cooling overclocking by just pushing some sliders on the driver, not serious overclockers who will mount on it ..

But again, the ASIC will not make that your gpu is a good overclocker or a bad one ... its a part of it, but really not the final one ... It could be "ONE" indication .. not the final one ...

hence why my " remark " about marketing..... I dont say it is a bad things for peoples who want purcahse this gpu, thinking my moving the sliders it will goes to xx or yyy mhz.. just the way it is presented is a bit special at my sense.

This dont speak to the overclockers, but to the general consumers.
 
Last edited:
I would have agreed with you before the KingPin 980 Ti Classy came out, but according to KingP|n now the opposite is true with 980's. Since he currently has the highest rank (as if today) for all three FireStrike categories using a KingPin Classified 980 Ti there may be more truth to what he says. I'm curious to see what the ASIC values are for this card but I'm sure we will hear more about the new ASIC controversy as other scores are posted. Classifieds have always been in short supply so it really doesn't point entirely to a "marketing" ploy because Evga would have much more to lose as a company than a small increase in profit.
 
Last edited:
What is "ASIC Quality" as a percentage value even measuring? How is it set? At test and packaging time? Using what criteria/formula? Seems like an utterly bonkers way to normalise the bin curve for frequency and voltage. What am I missing?
It's a little bit like a 3DMark score in that regard...
 
I'm under the impression that ASIC quality does play a bigger part with Maxwell than previously. It seems that for air and watercooling high asic quality correlates pretty well with higher overclocking headroom and that these chips don't really respond all that well to higher voltages unless extreme cooling is used. Previous nVidia chips responded very well to voltages even with air and watercooling and ASIC quality wasn't the factor that determined max overclocks on those cooling methods. Even with Kepler, ASIC quality made a difference with power consumption, high ASIC chips had lower voltages to reach the same speeds in the normal speed range. I had two 780Tis with a significant difference in ASIC quality and the higher consumed around 10% less watts at stock clocks.
 
Additional K|ngP|n comments:
July 21, 2015 8:00 pm
Well, the ASIC stuff is interesting right. I tested around 15pcs on air cooling.
From lowest bin to the highest bin. Almost all KP980Ti seems to be clocking at around 1530-1590mhz at default voltage (around 1.16v). Higher asic cards will be the higher clocking ones, lower asic cards are at the lower end of the range.
There are definitely exceptions, not all gpus go like that, but most do.
ALL KPE 980ti are binned gpus :), even the lowest asic level base card is still capable to hit 1520-1550 usually or more.
That's the air part of asic. Its pretty consistent, but there can be rogue gpus.
The LN2 part of of ASIC value is that higher asic cards clock higher on LN2, have more leakage which will helps with cold problems, and have better memory controllers(highest memory clocks).
ASIC value really applies to LN2 the most. On air, it can help with maxwell because you cannot add additional voltage at ambient cooling. Highest clocks at default voltage almost always on higher asic gpus.
This is what makes higher asic better on Maxwell than previous (Kepler could scale voltage on air, so lower asic was better). Lower ASIC % in the past (pre-maxwell) was the best for raising voltage on gpu at air, because there is less leakage on lower asic gpus. Less leakage = less heat/less noise and also the ability to increase voltage.

http://kingpincooling.com/forum/showpost.php?p=30615&postcount=101

Another explanation from K|ngP|n at OC'ers.


How does ASIC relate to Maxwell Air overclocking? With maxwell gpus the above definitions of ASIC do not apply Well you guys know maxwell 980,980ti, titanx have ambient cooling voltage limit on what v's you can give it on air/water. That's just the way it is. Its been proven over and over on every single manuf brand 980,titanx, 980ti. kp980 owners as well, no different. These gpus don't like voltage on air over 1.22-1.23v usually max. Just green garbage all over the screen with more, no better clocks. Best clocks usually achieved with stock voltage or maybe slightly. So given what we know about ASIC quality and the voltage scaling capability of 980,980ti,t-x on air/water(NONE), it indicates the best gpus on 980Ti, will be the ones that can overclock the highest on default voltage or near default voltage. Ever noticed why almost every single review of 980ti (any brand) is around 1500mhz? The reviewer never can never add much voltage for better OC result.
....
How does ASIC relate to LN2 overclocking and more specifically kp980ti?
As explained above, higher asic = higher leakage. Leakage is actually a good thing and can be contained on LN2 cooling. Cards with more leakage will run a bit hotter, usually extending cold limits on gpus and getting more core MHZ on LN2. Every overclocker wants every last mhz right? Higher ASIC GPUS also have better memory controllers and typically can overclock the memory very high on LN2. Lower ASIC gpus usually are not so good at driving the memory on Ln2 and the overclocker will lose a lot of MHZ when going cold.
http://www.overclock.net/t/1565834/...-980ti-kinpin-cards-by-asic/140#post_24201543

Seems efficiency may play a part with Maxwell's overclocking ...
 
Last edited:
We can be cynical and dismiss that blob of text as pre-sales tactics...
How accurate is GPU-z reading of ASIC numbers?
And how do they bin such chips? Does Nvidia have grade A/B/C GM200 wafers and sell those to the highest bidders?
What we know so far...Nvidia just separates those that made the grade to full GM200 and sell those as TX...Nvidia will disable weak CUDA cores and sell those chips as 2nd tier cards...the 980Ti...what about chips that can't make it to 980Ti status?
ASIC is an interesting question for now and future of GPU clocking...
 
When using phrases as "most value" for a $850 graphic card I can't help but roll my eyes.

Still it shows Maxwell v2 does scale exceptionally well, which in itself is pretty impressive when we are still stuck at 28nm. Personally I am not upgrading until 16nm cards hit the market though.
 
Glancing around online, 1500MHz on air doesn't seem to be a serious stretch for a lot of the 980Ti's out there. A BIOS modification is probably the only real prerequisite.
 
I bought a GTX 970 a few months ago so while I'm fairly amazed at how much performance nVidia is squeezing out of the 28nm process, I can't imagine justifying the purchase of this new card. Especially so as I have a nice 1080p HDTV/monitor that I'm holding off on upgrading.

But progress is progress and I'm encouraged to see these 980 Ti cards, and AMD's Fury cards. The bar is now set pretty high for the next generation of cards on the newer process.
 
Back
Top