[Various] 1060 Reviews

[Krizby87]

Lets keep things as relevant as possible shall we? We aren't talking about the GTX 1070 or 1080, but 480 against 1060. So, keeping things as relevant as possible using you're logic people can set the power target to its maximum on a 1060 GPU and wow, look at that, it'll consume more power where it has the TDP room to maintain higher clocks, just like the AMD 480 GPU will if you remove its power target. To that end have you ever considered why the 480 uses more power? It simply has more going on under its hood (read: more future proof architecture), but we'll get on to that in a moment.

Again keeping things as relevant as possible you saying the 480 is "weak" because it has more bandwidth is utter nonsense. Sorry but you clearly gone and pulled that out your arse. The 480 not only has more "physical" bandwidth than a 1060 because you know, its just on a wider, more potent bus, it also makes use of delta colour compression just like a 1060 does to further enhance its actual bandwidth. Further, each Polaris geometry engine has a index cache for what AMD calls "small instanced geometry." The cache supposedly reduces the need to move data around, reducing internal bandwidth requirements and improving throughput. Pure and simple, this gives the 480 a advantage. In case you want to dispute this (I somehow get the feeling you will...) Proof. Check the bandwidth test, the 480 beats out the GTX 980 (sorry, no actual 480 vs 1060 I could find, but supposedly the 1060 is more or less 980 performance so close as possible match) with random textures and is within spitting distance with black textures. As a minor side note some people may also like to look at the polygon throughput test in that link, and the ALU tests, all of which show very potent numbers. Particularly the ALU tests where the 480 beats the 980 in every single one of them by very comfortable margins (you can also extrapolate these results as wins or on par with potential 1060 results). No I'm not conveniently ignoring the texture fillrate test either, just chalking it up to a architectural bottleneck, or hell, even a driver issue as results are very similar to that of a 380X which has less texture units.

Finally, back to that more going on under the 480s hood, just give THIS a read.

So bottom line, no, the 480 is not a weak GPU and is the more future proof with stronger DX12 and Vulkan showings than the 1060. To be clear, I'm not saying the 1060 is a bad card, its just not the smart choice with its lack of Async Compute performance, very uninspiring DX12 performance, and somewhat poor Vulkan performance. GTX 1060 is for "now", RX 480 is for now and the future, ie; as I already stated, those who don't plan to upgrade their GPU for 3+ years.



Trying to compare apples to apples much? Was the Athlon64 a worse CPU because people couldn't get bananas overclocks out of it like people could do with the Pentium 4? Hell no, you could get a monster OC on a P4 and it still sucked. To a lesser extent the same is true of the 480 it (currently) doesn't clock as high as a 1060 but it doesn't need to it basically gets the same amount done @ 1.3GHz as the GTX 1060 does @ 1.6 - 1.7GHz. Needing more speed to do the same amount of work does not mean a better GPU, it means the absolute opposite, which to spell it out means the 1060 has crap IPC performance. This should not be any surprise to anyone nvidia themselves have said they took away some IPC in a bid to maximise clock frequency.

I know there is no point discussing with you but what the heck i will do just that anyways for the heck of it lol.

All this talk about future proofing, IPC performance, tdp really don't give you any more credibility but rather make you look mentally challenged.

Fact: GTX 1060 is faster than RX480 at stock clock, in which the gtx 1060 can overclock an additional 10-15% (in pure fps term) while the Rx480 can oc like 5-10% (in pure fps term) under air. And pls dont talk about IPC performance like it is the only saving grace AMD has, in fact AMD has many saving grace and IPC is not one of them lol.

About future proofing: I believe Gtx 1060 and Rx480 stand on equal footing here. Dx12 performance is still game specific rather than AMD favored, much like dx11; Vulkan has just been released on Doom, maybe wait some more game with vulkan to come out before making a judgement here. And if you think people will buy rx480 base solely on Doom's performance, look at Overwatch performance benches (gtx 1060 is 15% faster than rx480) and the players base; in fact I believe the deciding factor for mainstream buyers are looking at are MMORPG and MMO shooter performance where Nvidia is still leading by a landslide, asking budget buyers to spend 60usd on games to play for a week is a bit stretching it don't you think ?

TDP: Gtx 1060 win here no discussion needed

[qcmadness]

I'm still curious how much manufacturing process is playing a role in the power consumption. If you just do the raw math, the 480 is 16% larger but has 29% more transistors. On paper, it should be on the more efficient transistor technology. But - transistor size does not necessarily directly correlate with gate leakage. Even if the gates are physically smaller, they could be more leaky, and thus part of why the 480 consumes so much more power than expected.

Again, I think the whole discussion of which card is better is moot. The 1060 is faster on average, and consumes less power. In terms of better engineering, it is the better card. I just don't think it's more valuable - especially not compared to the 4GB 480 which is like 7% slower than the 3rd party cards but 40% cheaper.

It seems that GCN is coming to an end with less efficient architecture and too-heavy emphasis on computing performance.

[AliG]

It seems that GCN is coming to an end with less efficient architecture and too-heavy emphasis on computing performance.

Where did you get any evidence of them moving away from GCN though? I agree it's an unbalanced architecture, and frankly could use twice the ROPs. But that doesn't necessarily mean they're moving away from GCN.

[iboomalot]

look at Overwatch performance benches (gtx 1060 is 15% faster than rx480

I play Overwatch and my aging 7970-GHZ Iam getting 65-75 fps during gameplay at 2560x1600 on EPIC settings.

I was shocked I could get those frame rates on any modern game but then again Overwatch with its lag 20 tick count is dumbed down some anyways.

rx 480 AIB or the Gtx 1060 looks like a winner winner chicken for dinner. Might replace my 7970 with a WC 1060 just wished the 1070 wasn't so overpriced ATM.

[AliG]

I wouldn't bother watercooling a 1060. For the cost of the watercooling components you could just upgrade to a 3rd party 1070.

Rule of thumb is never watercool anything that isn't a high end part. Odds are your return on investment will be better to just buy better hardware, and then upgrade cooling in the future (if even necessary).

[Metroid]

Yeah no point watercooling anything lower than a gtx 1070, usually the cost-benefit is to watercool only 1080 and higher.

[dasa]

Yeah no point watercooling anything lower than a gtx 1070, usually the cost-benefit is to watercool only 1080 and higher.

for anyone with a universal waterclock it costs nothing to watercool a new gpu provided it has a heat plate for the vrm separate to the main hsf
im still using a mcw60 ever since x1900xtx which is about to go from a 290 to a 1070
the 7970 cost ~$10 for some copper to make a shim
the 290 cost ~$10 for a vrm hs

[iboomalot]

for anyone with a universal waterclock it costs nothing to watercool a new gpu provided it has a heat plate for the vrm separate to the main hsf
im still using a mcw60 ever since x1900xtx which is about to go from a 290 to a 1070
the 7970 cost ~$10 for some copper to make a shim
the 290 cost ~$10 for a vrm hs

I WC due to fan noise I hate fan noise.

[AliG]

I WC due to fan noise I hate fan noise.

Giant heatsink + slow revolving fan = cheaper and just as quiet.

Plenty of people have done the math: watercooling is not financially worth it unless if you are looking to take a high end product to the extreme performance level. The only other reasons to go water are if you have a personal interest in liquid systems, or if you want bragging rights.

[iboomalot]

Giant heatsink + slow revolving fan = cheaper and just as quiet.

Plenty of people have done the math: watercooling is not financially worth it unless if you are looking to take a high end product to the extreme performance level. The only other reasons to go water are if you have a personal interest in liquid systems, or if you want bragging rights.

Don't care about bragging rights I leave that to the LN2 guys, I have a MO-RA3 Radiator with 4 Silverstone 180mm currently running at 600-700 rpm. From 2 feet away you can't gear them with temps between 38c-42c during gaming.

I removed the shrouds off my Sabertooth MB due to the stupid fans. What air cooling solution would keep these cards under 60c and be silent under gaming ??

[AliG]

Don't care about bragging rights I leave that to the LN2 guys, I have a MO-RA3 Radiator with 4 Silverstone 180mm currently running at 600-700 rpm. From 2 feet away you can't gear them with temps between 38c-42c during gaming.

I removed the shrouds off my Sabertooth MB due to the stupid fans. What air cooling solution would keep these cards under 60c and be silent under gaming ??

Gelid and Arctic Cooling both make huge GPU coolers that are known for being effective and very quiet. You won't get true silence even with watercooling (unless if somehow you find a way to make an effective system that is both fanless and pumpless).

[Ket]

I know there is no point discussing with you but what the heck i will do just that anyways for the heck of it lol.

All this talk about future proofing, IPC performance, tdp really don't give you any more credibility but rather make you look mentally challenged.

Fact: GTX 1060 is faster than RX480 at stock clock, in which the gtx 1060 can overclock an additional 10-15% (in pure fps term) while the Rx480 can oc like 5-10% (in pure fps term) under air. And pls dont talk about IPC performance like it is the only saving grace AMD has, in fact AMD has many saving grace and IPC is not one of them lol.

About future proofing: I believe Gtx 1060 and Rx480 stand on equal footing here. Dx12 performance is still game specific rather than AMD favored, much like dx11; Vulkan has just been released on Doom, maybe wait some more game with vulkan to come out before making a judgement here. And if you think people will buy rx480 base solely on Doom's performance, look at Overwatch performance benches (gtx 1060 is 15% faster than rx480) and the players base; in fact I believe the deciding factor for mainstream buyers are looking at are MMORPG and MMO shooter performance where Nvidia is still leading by a landslide, asking budget buyers to spend 60usd on games to play for a week is a bit stretching it don't you think ?

TDP: Gtx 1060 win here no discussion needed

You know, the simple fact the first thing you have to do is attempt a insult just says one thing very clearly, how to put this in polite a way as possible.. you're knowledge and\or understanding is very limited of what I put in that post. To say the least.

So, lets start dissecting what you said shall we?

1. Nobody is arguing the 1060 isn't faster than the 480 at stock clocks, what was being pointed out is the lack of Instructions Per Clock compared to Polaris, you clearly aren't understanding that. Simple truth; lower clocks to do more or less the same amount of work due to stronger IPC = better architecture, its more of a "brain". Simply put, this means you just do not need a chip that overclocks as much for the same performance or better. For a good example of this, go check some Athlon64 performance results against Pentium4. They should be simple enough to follow and give you a rudimentary understanding of IPC.

2. GTX 1060 is good in DX11, and mediocre in DX12 particularly titles that make use of Async Compute and actually loses performance in Vulkan, not because of some imagined bias or being unoptimised but because its less suited to nvidia hardware. In the case of the former the GTX 1060 performs badly because of a limited hardware scheduler, thats a architectural shortfall. Here is a simple chart that demonstrates the more limited scheduler in nvidia hardware compared to the one in AMD hardware click. That right there shows the better scheduler in AMD hardware at work, a midrange 200-250 card showing better scheduler performance than any Pascal or Maxwell II based card. Any title that makes use of AC, which it appears more and more games are, the 480 is the smart choice it will outperform and last longer than the 1060 in such titles because its better equipped to handle workloads requested of it in titles like that.

3. Overwatch is more of a exception than the norm, if you want to use that as a example theres certainly nothing wrong with me using Doom Vulkan results as a example, which you may need reminding of that nvidia has actively worked hardcore with Vulkan and Doom, its well documented in fact in numerous articles if you just do a bit of googling. The results you see now are pretty much just how it is for nvidia the results will not change that much. More pointedly though, why would you care? I picked Doom to show the 480s performance using Vulkan compared to the 1060 but if you are a nvidia user as I am, you just use OpenGL for similar performance. The point in the Doom example was to show any engine that may exclusively use Vulkan as a API (which could well be many given AMD tech is in MS and Sony consoles) you are going to be far better off with a AMD card.

4. Not that a price of a game has any relevance as people will buy whatever game they think they will enjoy; according to the steam store database as of about a month ago Doom seems to be selling for around 36 bucks, thats a bit cheaper than Overwatch AFAIK. You were aware of that, right champ?

Finally, I've explained and bought charts to the table for everything I've bought up, not to mention I've outright said the 1060 isn't awful but it doesn't represent value for money to the mainstream gamer given it has far more shortcomings (which I've gone in to extensive detail about) than the 480 does. I've backed everything I've said up, so you calling me, what was it? Right "mentally retarded", is also saying that every one of the authors of those independent articles I've linked to are also "mentally retarded", people who have clearly demonstrated superior knowledge to your own.

It seems that GCN is coming to an end with less efficient architecture and too-heavy emphasis on computing performance.

Where did you read that? I fairly often say GCN would be nice if it had extra ROPs, but its a strong architecture apart from that.

[Ket]

Dude, thats absurd comparison, 1060 is faster AT STOCK AND IT OVERCLOCKS MUCH BETTER, also athlons WERE overclocking VERY GOOD, i got mine 3200+ up to 2880Mhz,FX F3`s were doing like 3200mhz , RX 480 BARELY overclocks, that means that the performance delta (in most games) INCREASES which was very different from A64 vs P4NW.We get that nvidia did lose some "IPC" but overall, it worked for them.
Also remember that RX does less work (in most games, besides pretty much 3-4 of them) per watt, and P4 (especially 90nm) were powerhogs.
Comparison is just bad bad bad.
Also, how in the hell its comparing apples and bananas, these are same generation cards, made for the same segment sold at similar price.

This is clearly splitting hairs but as you want to, fine I will too.

1. Already said, multiple times, not arguing the 1060 isn't faster at stock or that it OCs better, just that its core speed proportionate to how much faster it is over the 480 shows how bad the IPC is of Pascal and that if you have better IPC, you don't need huge clock frequency to begin with to do more or less the same amount of work.

2. I didn't say the A64s didn't OC well did I? I said that the P4s clocked better than the A64s but despite this the A64s were still better. The point was, and it is obvious, you even say it yourself yet didn't realise thats exactly what I said, is that if you have a architecture thats more of a "brain" you don't need as high of a clock to match or exceed a architecture thats less of a brain but clocked faster.

3. You trying to do apples to apples comparisons with nvidia and AMD architecture is just rather blanket, they are both very different designs and need to be treated as such, hence why I go in to so much detail for any hardware.

[AliG]

@all ya'll bickering:
Just keep it civil, we don't need the mods shutting this thread down

[Ket]

I'm trying to, name calling is just juvenile and tends to just make me roll my eyes and ignore it I'm more mature than to let something like that bother me.

[vario]

This is clearly splitting hairs but as you want to, fine I will too.

1. Already said, multiple times, not arguing the 1060 isn't faster at stock or that it OCs better, just that its core speed proportionate to how much faster it is over the 480 shows how bad the IPC is of Pascal and that if you have better IPC, you don't need huge clock frequency to begin with to do more or less the same amount of work.

2. I didn't say the A64s didn't OC well did I? I said that the P4s clocked better than the A64s but despite this the A64s were still better. The point was, and it is obvious, you even say it yourself yet didn't realise thats exactly what I said, is that if you have a architecture thats more of a "brain" you don't need as high of a clock to match or exceed a architecture thats less of a brain but clocked faster.

3. You trying to do apples to apples comparisons with nvidia and AMD architecture is just rather blanket, they are both very different designs and need to be treated as such, hence why I go in to so much detail for any hardware.

1).Yes, pascal IPC is worse in most cases comparing to previous generation.However, if it allows it to clock substantially higher while consuming less power ,then its just the better aproach and or execution and the point of worse IPC is moot.
And yes, if you have high IPC arhitecture, you dont need as high as clocks, you arent preaching anything that anyone doesnt know already .Problem is, polaris is not only achieving much lower clocks, it also overclocks much much worse percentage wise (not pure clocks wise).In effect, it DOES NOT do the same amount of work while being more complicated (more transistors, higher memory bandwith ,higher pcb complexity which results in more power draw).So what bringing this up really do ? Yes , polaris has higher IPC, but result is still the slower more complicated card, it costs AMD more to build them, and they will get less for them.

2.In general , yes, A64 do more work which results in not needing as high clocks, but comparison really breaks here, because polaris is more akin to thoroughbred-A, which had higher IPC but was running HOT, barely overclocked and did not meet clock targets and needed manufacturing hardware fix.

3.Thats absurd, we cant compare products because they were did by different people in different ways ?! Thats the whole concept!We as customers look at the finished product, we dont really care how it does the things it does :P AMD could use black magic for all i care.

Going back to bad low IPC designs, IF first BD clocked 5ghz at 90W and oced to 6ghz, nobody would care that it has lower ipc than the intels core.

[iboomalot]

The GTX 1060 is slightly faster but does take more MHZ to do so. Testing a Rx480 should be done at 1266mhz and 1060 at 1506mhz(majority of stock speeds on newegg) for comparison purposes.

OCing IMO should be done at 1400-1500 on RX and 1900-2000 on the 1060 for comparison purposes. I am not a fan of NV price gouging but NV did a great job and should be faster overall during these tests.

[Krizby87]

The GTX 1060 is slightly faster but does take more MHZ to do so. Testing a Rx480 should be done at 1266mhz and 1060 at 1506mhz(majority of stock speeds on newegg) for comparison purposes.

OCing IMO should be done at 1400-1500 on RX and 1900-2000 on the 1060 for comparison purposes. I am not a fan of NV price gouging but NV did a great job and should be faster overall during these tests.

Pitching best rx480 sample against worst gtx 1060 sample sound like fair to you ? Can you list some review where rx480 reach above 1400mhz with custom cooler ?

Comparing IPC or mhz without regards to actual performance is like comparing 5gear AT car with the same top speed as a 6gear AT car and you go like "if this car had 6gear it would be much faster" lol. Good luck adding an additional gear or oc the rx480 beyond 10% without hardmodding lol.

IMHO best value card for the money is rx480 4gb (as agreed by many review site) for 1080p gaming. MMO players however should look for gtx 1060. Some posts before mention his 7970 getting 75fps, yes you can play Overwatch but not competitively i presume.

[vario]

The GTX 1060 is slightly faster but does take more MHZ to do so. Testing a Rx480 should be done at 1266mhz and 1060 at 1506mhz(majority of stock speeds on newegg) for comparison purposes.

OCing IMO should be done at 1400-1500 on RX and 1900-2000 on the 1060 for comparison purposes. I am not a fan of NV price gouging but NV did a great job and should be faster overall during these tests.

Stock speed of a GTX 1060 is 1506Mhz BASE clock and 1708Mhz boost clock, and RX 480 has a BOOST clock of 1260Mhz and a base of 1120Mhz.So right there you should see where the problem with your understanding lies.From what i gather while gaming both cards spent most of their time at boost.
And pretty much all the reviews were done on stock non oced samples so what youre looking at are the scores you want.
As for OC, most reviewers were not able to pass 1350 on RX480`s even using custom coolers, so doing a normal on air comparison on 1500mhz is a pipe dream.

[iboomalot]

Stock speed of a GTX 1060 is 1506Mhz BASE clock and 1708Mhz boost clock, and RX 480 has a BOOST clock of 1260Mhz and a base of 1120Mhz.So right there you should see where the problem with your understanding lies.From what i gather while gaming both cards spent most of their time at boost.
And pretty much all the reviews were done on stock non oced samples so what youre looking at are the scores you want.
As for OC, most reviewers were not able to pass 1350 on RX480`s even using custom coolers, so doing a normal on air comparison on 1500mhz is a pipe dream.

My Bad, you are correct should be tested at the clocks you mentioned and I said the gtx 1060 came out ahead. Overwatch has a 20 tick count I could run at 30 fps and be competitive on those laggy servers.

Which reviewer on updated drivers with an aftermarket aircooled solution couldn't pass 1350? Can you link that please.

[vario]

https://www.techpowerup.com/reviews/...RIX_OC/26.html

my bad, 1355Mhz.Was writing from memory.

http://www.gamersnexus.net/guides/24...r-overclocking

This one is a watercooling result. 1390Mhz.
Overall im pretty sure there are going to be better chips out there, but im not sure anyone will be binng cards...

[dasa]

Giant heatsink + slow revolving fan = cheaper and just as quiet.

Plenty of people have done the math: watercooling is not financially worth it unless if you are looking to take a high end product to the extreme performance level. The only other reasons to go water are if you have a personal interest in liquid systems, or if you want bragging rights.

if i hadnt swapped parts for the fun of moving to a matx system water would have worked out cheaper vs buying expensive hsf for cpu\gpu most upgrades
water is good for those that want low noise combined with high performance on cpu & gpu's and dont mind the extra maintenance that comes with it

cpu block storm g5 12 years old gone through 6 cpu
gpu block mcw60 10 years old used on 6 gpu (grabbed another ek block for cf & sli)
dc pump mcp355 to replace ehime 1250 pond pump 6 years old
ThermoChill PA120.4 to replace heatercore 6 years old
barbs\clamps were only a few $ screw bling compressions

with massive hsf like nh-d15 available these days il admit the gains from watercooling the cpu are minimal temps may be lower but not enough to make a noticeable difference to clock speeds
its still a huge help with low noise in cf\sli on high end overclocked systems
not so much with a single gtx 1060

[SinOfLiberty]

water cooling is useless.

[iboomalot]

water cooling is useless.

1. temps are lower which reduces failure rates. GPU running 75c when gaming vs 40c
2. Nearly silent during gaming or silent depending on setup
3. slightly better overclocking
4. Done correctly you can re-use the blocks on different CPU and GPUs and once purchased your main expense is done.

I have had the same case and rad for 6 years and Pumps & CPU block for 3+ years.

Lets compare Hybrid cooling solution by arctic is 35dba my 180mm fans at my rpm is 18dba even with all 4 running my DBA is only 24dba

1 sone = 40 dB which is loud in comparison

So water cooling isn't useless

[SinOfLiberty]

only for bragging.

980Ti 70% fan is barely loud.

Plus you must change water blocks every few months.