Nvidia is at it again: 3060 8GB released, up to 35% slower than its 12GB counterpart

[Blademaster91]

On 12/9/2022 at 12:24 AM, HenrySalayne said:

That's not a good point. We have seen XX70 series dice getting put into XX60 cards, but Nvidia tuned the performance to be slightly better or on par with the XX60 series die.

 

Even if they do, all reviews dating back to the 3060 (12 GB) launch simply call the card 3060. Nvidia not only knew this, their homepage also doesn't differentiate between the 8 GB model and the 12 GB model.

The final verdict can only be one thing: Nvidia is deceiving their costumers.

True, there is also the revision of 30 series cards with LHR, going by the die number alone isn't a very good way to indicate a difference.

And yeah the spec page isn't up to date either, so either Nvidia rushed the 3060 8GB to launch and didn't update the page, or they're trying to hide the difference. But Nvidia also didn't announce the 3060 8GB, or tell reviewers and offer any review samples, Nvidia isn't stupid which makes this product seem like they're intentionally trying to mislead consumers with two different cards with the same product naming.

[leadeater]

18 minutes ago, Blademaster91 said:

Nvidia isn't stupid which makes this product seem like they're intentionally trying to mislead consumers with two different cards with the same product naming.

I think it's more likely they are just being lazy or some form of that since it's a region specific variant. These odd ball cards are almost always only available in the Asia Pacific areas, probably because logistically it's easier as that is where they get made. May or may not be in response to an over supply or under supply issue, some actually legitimate business reason made them do this. It's not like Nvidia create products like this just for fun, everything has a cost so not doing it and sticking with current models would be cheaper without factors that would change that.

 

To be honest I wouldn't be surprised if they did none to minimal testing and found the performance difference to be small based on what they did and authorized this. I doubt there is any reference Nvidia designs for it and it's on AIBs to use existing design knowledge and rules. Basically it would not shock me at all if Nvidia had no idea the performance difference on average across a large sample of games was this large. 

 

Giving it an official model designation like RTX 3050 Ti would likely necessitate a global release and availability as well as creating all the supporting marketing and product information for it. Something Nvidia seems they did not want to do.

 

Not that it excuses any of this.

[HenrySalayne]

1 hour ago, leadeater said:

I think it's more likely they are just being lazy or some form of that

I honestly doubt that. The lazy thing would have been a different name. Nvidia knows that all their publications and all other publications just said 3060 and there was no mention of "12 GB" anywhere. Sorting through this mess to differentiate the two SKUs was always a hassle they could have prevented.

1 hour ago, leadeater said:

To be honest I wouldn't be surprised if they did none to minimal testing and found the performance difference to be small based on what they did and authorized this.

Just a quick reminder that the memory width dropped by a third. I think that would have come up in testing, since all benchmarks done on this card show it pretty clearly (not always 30% though, but that's expected).

1 hour ago, leadeater said:

Basically it would not shock me at all if Nvidia had no idea the performance difference on average across a large sample of games was this large. 

They design a new SKU, they cut a third of the memory bus and they had no idea? That's impossible.

Deliberate is a strong word but in this case it applies beyond reasonable doubt.

[leadeater]

44 minutes ago, HenrySalayne said:

Just a quick reminder that the memory width dropped by a third. I think that would have come up in testing, since all benchmarks done on this card show it pretty clearly (not always 30% though, but that's expected).

I know how much less it is but that really does depend on what was tested and also if it was tested. Also the review/article may well just have chosen to show all the worst case examples on purpose to illustrate the problem, I know I would have done that.

 

Quote

So let's take a look at that, we'll go over a few games and then show you a 12 game average. For collecting these results we used fresh data with both 12GB and 8GB RTX 3060 cards using the latest display drivers. Other GPUs are included for comparison purposes.

Quote

That's enough with the individual game results, you can easily get the idea. On average you can expect 17% better performance with the 12GB model at 1080p, but as we saw that margin can be as high as 35%...

Most of the shown examples were above the 17% so I kind of have to assume the worst case examples were shown mostly.

 

I honestly don't think Nvidia did anymore more than projected performance difference and not actually product testing since I don't think they built a single one of these and are AIB only.

 

44 minutes ago, HenrySalayne said:

They design a new SKU, they cut a third of the memory bus and they had no idea? That's impossible.

They knew it would perform worse the difference is knowing by how much. They may have projected 10%-15% and decided "eh that should be fine".

 

Like I said doesn't make it ok but I can definitely see how Nvidia did next to no product testing and just signed off on it to move more product and resolve what might be a temporary GDDR6 shortage. There actually has to be a reason to do this, simply being jerks actually isn't one. There actually has to be a larger business driver than that.

 

AIBs have to buy both the die and GDDR from Nvidia, they don't source the GDDR directly themselves so any supply issues with GDDR would be dealt with directly by Nvidia be that over or under supply. Or it's an oversupply of GPU dies.

[Fasterthannothing]

Wow you guys are still going at it? 9 pages of arguing if users are too stupid to read numbers on a box or not. Get a life people it's not that big of a deal.

[Shzzit]

This any different then 1060 3gb vs 6gb? I mean both company’s have been doing this for a long time. 

[HenrySalayne]

6 hours ago, leadeater said:

I honestly don't think Nvidia did anymore more than projected performance difference and not actually product testing since I don't think they built a single one of these and are AIB only.

I'm pretty sure it's in Nvidia's power to turn 12 GB models into 8 GB models with a flick of a switch.

They other way around would be more complicated though...

[leadeater]

2 hours ago, HenrySalayne said:

I'm pretty sure it's in Nvidia's power to turn 12 GB models into 8 GB models with a flick of a switch.

They other way around would be more complicated though...

Well yea they have to authorize it, or more likely in this case require AIBs to make it but that doesn't mean they actually made any engineering samples and ran it through testing. 

 

4 hours ago, Fasterthannothing said:

Wow you guys are still going at it? 9 pages of arguing if users are too stupid to read numbers on a box or not. Get a life people it's not that big of a deal.

It's only as big of a deal as people want to make bad arguments and pretend it's not an issue. It's only a small issue, it only get bigger when certain people want to argue against all the evidence in front of them and ignore everything that is said and shown to them.

 

It's a bad name and should have been named better. Since far as I can see literally everyone agrees, any arguments otherwise is entirely pointless. Making it necessary to point out that foolishness is the problem, more so than a dumb graphics card name that could have easily been anything else.

[leadeater]

4 hours ago, Shzzit said:

This any different then 1060 3gb vs 6gb? I mean both company’s have been doing this for a long time. 

Did you know there was a GTX 1060 5GB that performed worse than the GTX 1030 3GB? Did you know my opinion of the 3GB vs 6GB that time is the same as this time. Repeated bad doesn't change to good because it's repeated.

 

But bigger number is better. But VRAM capacity is adequate differentiator always. But people can easily read 

 

Bad arguments are bad. People just need to learn when they just don't have a good one.

[mr moose]

19 hours ago, leadeater said:

Ok now that is a good step in the right direction. Now tell me how they would know to go buy it?

They would know for the same reason they did know and and we know, Nvidia released it and they went and got one.  Just because there was no fan fare doesn't mean no reviewer is ever going to know it exists.  

19 hours ago, leadeater said:

That is not my only opinion nor my sole comments about this. If you were reading you'd know this.

Having multiple opinions doesn't mean any single one of them can't be just yours as opposed to being factual or relevant to everyone.

 

19 hours ago, leadeater said:

And it's not a good one either. Like too bad if they even tried the most likely outcome would have been finding information about the 12GB card.

I think a product being uniquely identified no matter where you look (makers website, online store, box on shelf or at any review site) is probably the single most important requirement of transparency for consumers.

19 hours ago, leadeater said:

As bad as those are those do not spread the misinformation that vram capacity is directly related to performance and anything that would normalize this I will object to.

Are you saying the card doesn't have 8G of ram?  It does not claim on either box or listing how good the ram is, none of them do, just like hdd and ssd size doesn't indicate performance.   If any consumer doesn't look at reviews to see if the 8G version performs adequately for the price or in comparison to the 12G version then it doesn't matter how good or bad the name is, that consumer is going to buy a random product.   If we are talking specifically about those who are uneducated, then they are far more likely to look at reviews between them or ask on forums what the difference is. 

19 hours ago, leadeater said:

Die on the hill of "but the spec sheet", it's not sufficient information to any buyer as to what they are getting. If they can make the effort for the original product, the 12GB, they can make the effort for the 8GB. They didn't, I am and will criticize them for it while also saying naming it in that why is bad regardless if they had bothered to.

I don;t know why you think I intend on dying on any hill, it's a pretty simple concept,  you would have to be a blind moron to not know there is a difference between the two cards, Nvidia don't hide the difference if you go look at the specs and reviewers are showing consumers what that difference is. 

 

19 hours ago, leadeater said:

I as an educated buyer that knows memory bus matters has no way of knowing what the difference between a 128bit and 192bit variant would be. I could only rely on information from Nvidia or independent reviews. Had I been buying a month ago and looking at these two I would have zero information necessary from both those options. Going back days while it was possible to get it from Techspot/Hardware Unboxed I would have to be careful not to end up looking at the original RTX 3060. So much of this is relying on the buyer that I find that a problem.

If you were buying a month ago and could not find a review would even buy it?  Would you be trying to find out what it's all about?   Of course you would, because as you have just illustrated you noticed from the get go that it isn't the 12G version and something is different.  That should tell you all you need to know.  they are the same price and one has more ram,  even if the 8G version was cheaper the first question anyone (dumb or educated) is going to ask is why is it cheaper and will I notice the difference? is that saving worth it?   Until you can answer those then it's just an unknown.  Besides all that, 8G does not insinuate better performance, if anything having it the direct option beside the 12G version would suggest it is worse.

 

The entire premise and argument I am making is that the card is uniquely identified at all levels of advertising, review and sales.  If a consumer cannot get the information they need to make a wise choice either by lack of care for research or because no reviews exist then they have a choice to buy it or wait or buy something they know the performance of.  In every situation the consumer has the choice and is not misled. 

[mr moose]

17 hours ago, HenrySalayne said:

They design a new SKU, they cut a third of the memory bus and they had no idea? That's impossible.

Deliberate is a strong word but in this case it applies beyond reasonable doubt.

I think there is no doubt they knew exactly what they were releasing,  they know exactly how it performs and deliberate is a reasonable word.    

 

 

[CTR640]

On 12/10/2022 at 12:17 PM, mr moose said:

 

I like how everyone is just skipping past the specs page that tells you everything you need to know:

 

 

It clearly shows there are 2 versions of the non TI, a 12G and an 8G, click view full specs and voila:

 

 

They are not hiding anything and separate the two cards specs clear enough.

 

Now if I go to buy either one this is how they are listed in nearly every store:

 

Nvidia and every card maker use the RAM size as a postfix to differentiate between models.  It has always been this way which is why the 1030 issue was unforgivable.

 

 

Dutch version doesn't have the different 3060:

 

Any other Dutch member can verify that.

 

It's from the link you posted and nVidia directed me autmatically to the Dutch version.

[HenrySalayne]

11 hours ago, mr moose said:

If any consumer doesn't look at reviews to see if the 8G version performs adequately for the price or in comparison to the 12G version then it doesn't matter how good or bad the name is, that consumer is going to buy a random product.

I have shown earlier that even if a costumer watches reviews, most older reviews are not even mentioning the RAM size.This lures even consumers who try to educate themselves into a trap. And what you are doing here is just victim shaming.

 

[leadeater]

20 hours ago, mr moose said:

If you were buying a month ago and could not find a review would even buy it?  Would you be trying to find out what it's all about?   Of course you would, because as you have just illustrated you noticed from the get go that it isn't the 12G version and something is different.  That should tell you all you need to know.  they are the same price and one has more ram,  even if the 8G version was cheaper the first question anyone (dumb or educated) is going to ask is why is it cheaper and will I notice the difference? is that saving worth it?   Until you can answer those then it's just an unknown.  Besides all that, 8G does not insinuate better performance, if anything having it the direct option beside the 12G version would suggest it is worse.

The 8GB version is actually available for cheaper and the way it is named in the complete absence of information would lead a lot of people to assume since it's "An RTX 3060" that it is the same fundamental product just with less VRAM which it isn't.

 

And to remind you complete absence of information had been the actual situation for a decent period of time. Yet even now when that information can be found it is too easy to end up looking at information for the 12GB card by mistake.

 

20 hours ago, mr moose said:

Besides all that, 8G does not insinuate better performance, if anything having it the direct option beside the 12G version would suggest it is worse.

No it suggest 8GB VRAM versus 12GB VRAM.

 

Promoting and accepting anything otherwise is literally a primary objection of mine or have you not noticed? To use it to signify a performance difference or a product difference that is befitting of a different model name as Nvidia would normally do is wrong always, Every single time.

[leadeater]

21 hours ago, mr moose said:

They would know for the same reason they did know and and we know, Nvidia released it and they went and got one.  Just because there was no fan fare doesn't mean no reviewer is ever going to know it exists.

And how and where exactly did Nvidia tell anyone? Do something for me, Google GTX 1060 5GB review. Surely you agree enough time has passed for reviews of that to be out by now?

 

Do not assume reviewers will know and reviews will exist, that is a bad and unsafe assumption to make.

 

21 hours ago, mr moose said:

I think a product being uniquely identified no matter where you look (makers website, online store, box on shelf or at any review site) is probably the single most important requirement of transparency for consumers.

No the single most important thing is that the name difference actually portrays and prompts people in to thinking the right thing. Just because one says 8GB and the other says 12GB while both being called RTX 3060 does not mean this. You say it does, I say it does not.

 

I think there is a non zero group of people that would make this mistake, I think there is a non zero group of people who actually tried to look for the difference would end up looking at the wrong information due to the name conflict. Since there is certainly a non zero group of people the use of a different name not containing RTX 3060 followed by VRAM capacity or AIB product naming then VRAM capacity (as is standard so not out of the normal) would be the safer option.

 

My proposal is safer than the current name and the naming you are defending as adequate while making the argument that not everyone can be saved from themselves while disregarding those that could have been. Since neither of us will ever be able to quantify any of this I will maintain my position is the safer one.

 

21 hours ago, mr moose said:

Are you saying the card doesn't have 8G of ram?  It does not claim on either box or listing how good the ram is, none of them do, just like hdd and ssd size doesn't indicate performance.

RTX 3060 is the signifier of product configuration and performance. So you are very wrong here, the boxes all literally do say this unlike your incorrect statement right now. Why are you even saying this? Literally the fact the box does not adequately portray that there is such a large difference in the underlying product and it's performance is literally, literally, what I am pointing out and complaining about. 

 

If you cannot figure out that a RTX 3060 and an RTX 3070 signifies these types of differences then you have a seriously large problem. Looking at a review is not required to understand this.

 

RTX 3060 vs RTX 3070 does so adequately

RTX 3060 8GB vs RTX 3060 12GB does not adequately 

RTX 3050 Ti 8GB vs RTX 3060 12GB does so adequately 

RTX 3050 Ti vs RTX 3060 does so adequately 

 

3 out of 4 informs any buyer that there is a fundamental difference between the products that goes beyond minor differences. This even the least educated person already understands

 

Just like for an SSD I know a Samsung 970 Pro 1TB is a higher performing and higher endurance product than a Samsung 970 EVO 2TB. Just as you now said capacity doesn't indicate performance for HDDs and SSDs so why would it for GPU VRAM? You seem to both know this and ignore it at the same time.

 

You do realize you just counter argued your own point nicely with that you said right?

 

21 hours ago, mr moose said:

The entire premise and argument I am making is that the card is uniquely identified at all levels of advertising, review and sales.

And your reasoning, logic and arguments are full of holes and history proving you wrong. The names being unique isn't even in question. Nobody ever has said otherwise. RTX 3060 8GB and RTX 3060 12GB are unique yet do not sufficiently differentiate how different these products are and in what way. As has been pointed out to you countless times now.

[mr moose]

15 hours ago, CTR640 said:

Dutch version doesn't have the different 3060:

 

Any other Dutch member can verify that.

 

It's from the link you posted and nVidia directed me autmatically to the Dutch version.

Maybe the 8G versions is not up for sale ther or maybe whgoever is in charge of that part is lazy? I don't know.

 

13 hours ago, HenrySalayne said:

I have shown earlier that even if a costumer watches reviews, most older reviews are not even mentioning the RAM size.This lures even consumers who try to educate themselves into a trap. And what you are doing here is just victim shaming.

 

Which review didn't specify the ram size?  

3 hours ago, leadeater said:

The 8GB version is actually available for cheaper and the way it is named in the complete absence of information would lead a lot of people to assume since it's "An RTX 3060" that it is the same fundamental product just with less VRAM which it isn't.

So you think people can't differentiate between a 12 and an 8?    Only in a few heads does the black and white absolutes conclude that only the first 4 numbers are important and everything else can be ignored.

 

3 hours ago, leadeater said:

And to remind you complete absence of information had been the actual situation for a decent period of time. Yet even now when that information can be found it is too easy to end up looking at information for the 12GB card by mistake.

So?  if you buy a product that is not reviewed and you don't know how it performs who's fault is that?    Thinking that 3060 by itself should tell you everything you need to know to make an informed decision is just fantasy. 

3 hours ago, leadeater said:

No it suggest 8GB VRAM versus 12GB VRAM.

Huh? so we have gone from 8G is not different to 12 G to 8g is a suggestion? Come on.

3 hours ago, leadeater said:

Promoting and accepting anything otherwise is literally a primary objection of mine or have you not noticed? To use it to signify a performance difference or a product difference that is befitting of a different model name as Nvidia would normally do is wrong always, Every single time.

That is still just your personal opinion.  I got shitty with nvidia for making two 2G 1030's that were fundamentally different for this exact reason, but here they have not done that. There has always been a fundamental attribute that makes it clear the products are different, most of the time RAM is the difference.   IF you read all of the articles aregarding this very topic you can clearly see the very reason they believe there is a problem is due to RAM size offered.    Hell the OP article even they identify it purely by ram size:

 

From the article:

Quote

RTX 3060 8GB: Way slower, no cheaper, avoid, says first online review

Quote

downspecced 8GB version of the Nvidia GeForce RTX 3060 (opens in new tab) has been tested downunder by Hardware Unboxed (opens in new tab) and found horribly wanting despite offering no cost savings, for now.

Quote

The typical performance deficit versus the original 12GB model

 

Quote

The cheapest 8GB RTX 3060 from Newegg is

 

 

Every single time they want ot identify which card they are talking about hey use the Ram, they don;t say, the version with the lower memory bus bandwidth, or the card that is downspec'd.  Why? because Nvidia, card makerts and consumers all know that the ram size is the differentiator.  You know, I know, everyone knows it. Stop pretending it's not it doesn't do exactly that. 

 

3 hours ago, leadeater said:

And how and where exactly did Nvidia tell anyone? Do something for me, Google GTX 1060 5GB review. Surely you agree enough time has passed for reviews of that to be out by now?

I don't know, but they did such a good job at not telling anyone that a few reviews are out and we all know about it.

 

3 hours ago, leadeater said:

Do not assume reviewers will know and reviews will exist, that is a bad and unsafe assumption to make.

they do know and have done,  the only reason we found out about it was because exactly that happened.

 

3 hours ago, leadeater said:

No the single most important thing is that the name difference actually portrays and prompts people in to thinking the right thing. Just because one says 8GB and the other says 12GB while both being called RTX 3060 does not mean this. You say it does, I say it does not.

That seems to be the only thing you have to underpin your entire argument.  Ram size means something other than ram size is not exactly a great foundation for claiming misleading or deceptive marketing.

 

3 hours ago, leadeater said:

I think there is a non zero group of people that would make this mistake, I think there is a non zero group of people who actually tried to look for the difference would end up looking at the wrong information due to the name conflict. Since there is certainly a non zero group of people the use of a different name not containing RTX 3060 followed by VRAM capacity or AIB product naming then VRAM capacity (as is standard so not out of the normal) would be the safer option.

 

There is always a none zero group of people who will make all manner of mistakes,  but as I said earlier, if someone is that stupid that they will not look up reviews or find out what the difference is between 8 and 12 for this model then no naming convention will help them.

 

3 hours ago, leadeater said:

My proposal is safer than the current name and the naming you are defending as adequate while making the argument that not everyone can be saved from themselves while disregarding those that could have been. Since neither of us will ever be able to quantify any of this I will maintain my position is the safer one.

I see no evidence that the current prefix/GPU number/post fix and ram size naming scheme (when honest) actually misleads consumers.   You can have two identical GPU's with identical ram bandwidth, power connectors the whole shebang bar for ram size and you will still need reviews to know if the ram size limitation is a problem for your chosen game/setup.  The fact that other specs change with ram size should not be a problem because it has already been established that the product is not the same.    And if you think you are good enough to know how much ram you need for a given GPU and game without looking up reviews then good for you, 99% of consumers don't know or only think they know.

 

3 hours ago, leadeater said:

RTX 3060 is the signifier of product configuration and performance. So you are very wrong here, the boxes all literally do say this unlike your incorrect statement right now. Why are you even saying this? Literally the fact the box does not adequately portray that there is such a large difference in the underlying product and it's performance is literally, literally, what I am pointing out and complaining about. 

Your argument boils down to "the box doesn't tell me how the card will perform" so it must be misleading.   Nothing on the box including the GPU name will tell you how it will perform, you need reviews for that.  What the information on the box does do is identify exactly which product this is so when you look up reviews you can see how it performs. This is the one thing you fail to address.  There is a review out, it's not hard to find the review, we all know about it so you can ask on this forum and someone will tell you straight away not to buy it,  hell you can even go to Nvidia and get the actual specs on the card.   These are all facts you seem to think aren't as important as making the name tell you more than it actually can.

 

3 hours ago, leadeater said:

If you cannot figure out that a RTX 3060 and an RTX 3070 signifies these types of differences then you have a seriously large problem. Looking at a review is not required to understand this.

All you need to know and all that really tells you is they are different.  You go to a review site to find out which one is better.    Just because the 3070 typically performs better does not mean that every 3070 will or that there might be a 3060 that performs better for the price, or that even something from AMD might be a better choice again. 

 

Hell, you wouldn't even know which one was better without reviews.

 

3 hours ago, leadeater said:

RTX 3060 vs RTX 3070 does so adequately

RTX 3060 8GB vs RTX 3060 12GB does not adequately 

RTX 3050 Ti 8GB vs RTX 3060 12GB does so adequately 

RTX 3050 Ti vs RTX 3060 does so adequately 

 

3 out of 4 informs any buyer that there is a fundamental difference between the products that goes beyond minor differences. This even the least educated person already understands

 

Just like for an SSD I know a Samsung 970 Pro 1TB is a higher performing and higher endurance product than a Samsung 970 EVO 2TB.

 

Only because you have seen reviews,  and only because you can identify each product in said reviews. 

3 hours ago, leadeater said:

Just as you now said capacity doesn't indicate performance for HDDs and SSDs so why would it for GPU VRAM? You seem to both know this and ignore it at the same time.

If you re read my posts you will see that I am not arguing RAM size is an indicator of performance, at best I am acknowledging that the average uneducated consumer will see a higher ram amount and think that means better.  My argument is that traditionally ram size is unique to each model/version (so much so that it is a defacto part of the name).  When Nvidia released two cards that are the same prefix, same GPU and same ram size but one is spec'd lower than I am totally on your side.  But my point in this case is that you can identify these products very clearly because of the ram size, therefore when you look at the review you will know which is better. 

 

 

3 hours ago, leadeater said:

You do realize you just counter argued your own point nicely with that you said right?

I really don't see how.

3 hours ago, leadeater said:

And your reasoning, logic and arguments are full of holes and history proving you wrong. The names being unique isn't even in question. Nobody ever has said otherwise. RTX 3060 8GB and RTX 3060 12GB are unique yet do not sufficiently differentiate how different these products are and in what way. As has been pointed out to you countless times now.

No, of course they don't tell you how it performs, it's just a bunch of numbers.   Reviews tell you how a product performs is specifically different in every attribute, names and specs identify the product, it really is that simple.  You cannot argue that the name needs to illustrate a degree of performance or the detail of the product because they just can't, there are way to many variations for that.  All the name has to do is be unique to the product so it cannot be confused with another product. this is why I said earlier what they call it doesn't matter, so long as it is uniquely different for each version/product.

[MageTank]

3 hours ago, mr moose said:

Which review didn't specify the ram size? 

Almost all of them specify it in the "specs" pages of the review. The problem is, almost everyone looking at reviews is going to skip the written essay in favor of the visual graphs. None of the visual graphs included the VRAM size for the 3060 so people googling "RTX 3060 review" are likely going to find those older reviews from the usual reputable sources and assume that the 3060 referenced in the review matches the one they are buying, when that may not be the case.

 

This means every review outlet is going to have to start specifying VRAM capacity in their GPU names on their charts going forward in anticipation of Nvidia's shenanigans or risk potentially misleading their audience unintentionally. 

 

Or, Nvidia can have my idea for free and bring back the xxx5 names to their GPUs. Seriously, there was nothing stopping them from calling this an RTX 3055, lol.

[leadeater]

17 hours ago, mr moose said:

So you think people can't differentiate between a 12 and an 8?    Only in a few heads does the black and white absolutes conclude that only the first 4 numbers are important and everything else can be ignored.

If you are asking this question then you have listened to nothing and are not even trying to understand what the problem might be.

 

17 hours ago, mr moose said:

So?  if you buy a product that is not reviewed and you don't know how it performs who's fault is that?    Thinking that 3060 by itself should tell you everything you need to know to make an informed decision is just fantasy. 

Bad logic, these products should at least have something from the original manufacture. Nvidia should acknowledge with actual information what the product and performance difference might be. A spec sheet listing is not sufficient.

 

RTX 3060 by itself tells you it's an RTX 3060, 12GB and 8GB on the end tells you what the VRAM capacity is. Thinking that people are inherently going to think they will preform so differently is idiotic and goes against all understanding of Nvidia graphic card naming.

 

17 hours ago, mr moose said:

There is always a none zero group of people who will make all manner of mistakes,  but as I said earlier, if someone is that stupid that they will not look up reviews or find out what the difference is between 8 and 12 for this model then no naming convention will help them.

So you are just going to dismiss how easy it is to find the wrong information, that information about it from Nvidia does not exist and the time span where it did not exist from everyone.

 

Here is a suggestion, instead of passing this off and calling people idiots you apply a little bit of actual thinking to the issue. Making such a mistake is simple and does not make such a person stupid.

 

Don't ask me what I think makes anyone saying what you are to be...

 

17 hours ago, mr moose said:

The fact that other specs change with ram size should not be a problem because it has already been established that the product is not the same.

And it's been established and proven that those changes typically result in an actual model change i.e. RTX 3070 Ti. A change in ram size does not mean a change in memory bus, it can change not it will change.

 

17 hours ago, mr moose said:

I see no evidence that the current prefix/GPU number/post fix and ram size naming scheme (when honest) actually misleads consumers.

Right so you ignored evidence proving this to be true. No I will not post it again. You have lost your right due to your own actions for me to bother giving you the effort to do it again.

 

It's been proven it does, end of story. Historic evidence is there, go find it.

 

You not seeing it is a you problem not proof it's not a thing. Yet again I cannot fix a you problem.

 

17 hours ago, mr moose said:

But my point in this case is that you can identify these products very clearly because of the ram size, therefore when you look at the review you will know which is better

Does not address the issue of existing review not including 12GB in the graphs. Does not address the issue of Nvidia not clearly supply equivalent information that exists for the 12GB original model. Does not address the issue of how easy it is to end up looking at the wrong one, large part due to the two former things.

 

You argument is insufficient and is not better than any of proposed name others have given.

 

Notice how I do not agree with your point. Here's a hint, I likely will never agree with your point and you're supporting arguments you've tried to put forward are not good. If they were good I would have change my opinion, even slightly. Nothing you have said has changed anything in any way for me. Re-explaining again will not change the situation, I do understand what you have been saying but me understanding is not agreeing.

 

17 hours ago, mr moose said:

Your argument boils down to "the box doesn't tell me how the card will perform" so it must be misleading. 

No it doesn't. Why are you consistently proving so hard you read nothing? Maybe clue in to me repeating this to you multiple times now that you do actually not understand a single argument given to you because seemingly you don't want to.

 

If this is the case then simply exist the conversation.

 

17 hours ago, mr moose said:

Only because you have seen reviews,  and only because you can identify each product in said reviews. 

No I can because that's how Nvidia has always named products. I do not need to read a review to know that an RTX 3060 is fundamentally different to an RTX 3070 because Nvidia tells me so and gives me the information directly about how they are different to an actually sufficient standard.

 

This is not a argument for a replacement of reviews. This is an argument about how the naming convention is sufficient to indicate it's fundamental product difference and how an RTX 3060 is not an RTX 3070.

 

I can read exactly zero reviews and understand there is a large difference between those two products. Without a review of the RTX 3060 8GB I could not do the same, it is currently impossible to be informed in any about how different it might be to the RTX 3060 12GB.

 

No 128bit vs 192bit is not sufficient to inform me of that.

 

17 hours ago, mr moose said:

Hell, you wouldn't even know which one was better without reviews.

Yes I can, that were not true then Nvidia would be guilty of false advertising.

 

I should consult reviews but I do not need to consult them to know which is better. I can get other important information from reviews but that has nothing to do with this and as much as you want to try and bring that in to the conversation it is not relevant to the issue between the RTX 3060 8GB and RTX 3060 12GB naming.

 

17 hours ago, mr moose said:

You cannot argue that the name needs to illustrate a degree of performance or the detail of the product because they just can't, there are way to many variations for that.

They literally already do, right now and in the past. What on earth are you saying?

 

I legitimately think you've lost the plot trying to defend your point that you are now saying things you know not to be the case that you would at any other time realize.

[leadeater]

16 hours ago, mr moose said:

My argument is that traditionally ram size is unique to each model/version (so much so that it is a defacto part of the name).

Then you'd be wrong because the defacto stance about how people talk about Nvidia graphics card is without the VRAM.

 

You're fooling yourself by looking at URL titles on online stores. Those have zero baring on how real people in the real world and in reviews talk about the GPUs, write them down and speak audibly about them.

 

You would be lying if you said you hear people regularly say out loud the VRAM amount for any typical graphics card and you'd be lying if you said that VRAM capacity is normally included in review graphs.

[mr moose]

19 hours ago, MageTank said:

Almost all of them specify it in the "specs" pages of the review. The problem is, almost everyone looking at reviews is going to skip the written essay in favor of the visual graphs. None of the visual graphs included the VRAM size for the 3060 so people googling "RTX 3060 review" are likely going to find those older reviews from the usual reputable sources and assume that the 3060 referenced in the review matches the one they are buying, when that may not be the case.

 

This means every review outlet is going to have to start specifying VRAM capacity in their GPU names on their charts going forward in anticipation of Nvidia's shenanigans or risk potentially misleading their audience unintentionally. 

 

But they all do say it in the review.  If a consumer isn't confirming that the review is for the model they are looking at then I don't what to tell anyone.  Every review I have read mentions the ram, bandwidth, power configuration etc.  Sure they don;t always need to put it in the graphs, but if you read any review or scientific research paper you will quickly see that that is the norm.   The average consumer has somehow managed to work it out alright up until now. I am not sure why all of a sudden it is a problem.

19 hours ago, MageTank said:

Or, Nvidia can have my idea for free and bring back the xxx5 names to their GPUs. Seriously, there was nothing stopping them from calling this an RTX 3055, lol.

And doing that would not change anything, people would still have to look up reviews to know how it performs,  but going on the arguments in this thread the logic would be if they call it a 3055 then people will assume it performs better than a 3050 and worse than a 3060.  That logic quickly falls down as soon as you realise that the name will never accurately depict performance no matter what they call it.

 

 

6 hours ago, leadeater said:

If you are asking this question then you have listened to nothing and are not even trying to understand what the problem might be.

 

Or you just don't like my conclusion.  

6 hours ago, leadeater said:

Bad logic, these products should at least have something from the original manufacture. Nvidia should acknowledge with actual information what the product and performance difference might be. A spec sheet listing is not sufficient.

Why's that?  there is almost nothing that the manufacture can include in a name that will satisfy the rational for asking for a different name.   Because no matter what they call it you will still be reliant on reviews to know how good it is.  I think the problem is you know that the ram amount makes a perfectly reasonable identifier as it has done before and you cannot argue that people will look at a 12G review and confuse it for an 8G review.  If that's the case then you may as well argue that every version of graphics card should have a completely different name and contain no numbers. 

 

6 hours ago, leadeater said:

RTX 3060 by itself tells you it's an RTX 3060, 12GB and 8GB on the end tells you what the VRAM capacity is. Thinking that people are inherently going to think they will preform so differently is idiotic and goes against all understanding of Nvidia graphic card naming.

And yet I never actually argued that people will assume a difference,  I argued that people will know there is a difference, but will require a review to know what that difference is.  I've always assumed that when a product has a different spec for anything and difference in price that there will be a difference in performance.  ALWAYS.  no matter if its just the ram size or a GT postfix.   It really isn't that hard to understand that, after all these companies spend a lot of money researching how consumer think and how names effect sales.  The difference between misleading and this is that the better model has the better number on it.   

 

6 hours ago, leadeater said:

So you are just going to dismiss how easy it is to find the wrong information, that information about it from Nvidia does not exist and the time span where it did not exist from everyone.

So far I haven't found the wrong information, every single review and google search I have done has told me there is a difference or at least stipulated which version of the 3060 the review is for. Even for all the reviews when there was only the 12G and the ti model in existence.   Even every link that has been posted here has specifically mention the 12G or the 8G version when referring to the specifics of the information.

6 hours ago, leadeater said:

Here is a suggestion, instead of passing this off and calling people idiots you apply a little bit of actual thinking to the issue. Making such a mistake is simple and does not make such a person stupid.

I never said making a simple mistake makes someone stupid, I said going out and buying a product without doing ANY research on it is stupid.  This should be self evident to anyone with half a brain.  There is a big difference between a product name being easily confused and a complete lack of any attempt to ensure you are seeking out the correct information.   Especially when every review has all the specs and the full model names along with links to resellers and prices.  Only someone completely ignoring half of what they read will make that mistake, and very likely that person will make a poor choice regardless of what any company calls their product.

 

6 hours ago, leadeater said:

Don't ask me what I think makes anyone saying what you are to be...

Personal attacks only weaken your argument.

 

6 hours ago, leadeater said:

And it's been established and proven that those changes typically result in an actual model change i.e. RTX 3070 Ti. A change in ram size does not mean a change in memory bus, it can change not it will change.

Never said that, I said a change in ram size establishes a change in the product.  It can now be identified as something different so whatever else changed should not be a problem, i.e the consumer will find out everything that changed when they look up the specs of the model or reviews of the model.

 

6 hours ago, leadeater said:

Right so you ignored evidence proving this to be true. No I will not post it again. You have lost your right due to your own actions for me to bother giving you the effort to do it again.

 

I did not ignore anything,  I acknowledge that when the naming conventions are not honest it is a problem, but when they are honest it is not a problem.  You can post or not post anything you like using whatever reasoning you feel, but that will not change the fact that what you have posted does not prove anything as absolute as you are claiming against my postulations.

 

6 hours ago, leadeater said:

It's been proven it does, end of story. Historic evidence is there, go find it.

 

You not seeing it is a you problem not proof it's not a thing. Yet again I cannot fix a you problem.

You can't seem to fix a proof problem either.    You seem to be stuck in a dilemma that everything is black and white all the time.  Nvidia reused a GPU name and spec in a dishonest way before, as I have pointed out and said I agree with you on it,  but that doesn't mean every name is dishonest if the underlying specs aren't what you feel they should be.  So long as the name is unique to the product. 

 

6 hours ago, leadeater said:

Does not address the issue of existing review not including 12GB in the graphs. Does not address the issue of Nvidia not clearly supply equivalent information that exists for the 12GB original model. Does not address the issue of how easy it is to end up looking at the wrong one, large part due to the two former things.

The information is there, how can it be that information that is so easy to access both in the reviews and on the makers website equals not clearly supplying said information. 

You may as well tell the judge you didn't stop because the sign was not clearly there while looking at a photo of the sign being right there.   Honestly I got all the information that differentiates these two models straight form the 3060 page on the nvidia website, the 2 articles I read both listed the same differences in spec.  How is that not clearly identifying each product as being different.  the 12G version has been around since when start of this year end of last?  the 8G version has been out a month 2 at the best? I am reading a 3060 12G review that was released in june this year.   You can't just say "look at all these reviews of the 3060 that don;t differentiate between the 2 models when those reviews were all released long before the 8G version was anything.    You are conflating conditions that are not born of naming issues.

 

6 hours ago, leadeater said:

You argument is insufficient and is not better than any of proposed name others have given.

 

Notice how I do not agree with your point. Here's a hint, I likely will never agree with your point and you're supporting arguments you've tried to put forward are not good. If they were good I would have change my opinion, even slightly. Nothing you have said has changed anything in any way for me. Re-explaining again will not change the situation, I do understand what you have been saying but me understanding is not agreeing.

Hmm,  you say you understand but your rebuttals do not give that impression.  I say not reading any reviews when buying a GPU is stupid and you respond by accusing me of calling people who make mistakes stupid.  No it does not look like you understand what I am saying.

 

For the following I have snipped out bits that are just repeating SO i can make some more coherent remarks.

6 hours ago, leadeater said:

 you do actually not understand a single argument given to you because seemingly you don't want to.

 

 

I legitimately think you've lost the plot trying to defend your point that you are now saying things you know not to be the case that you would at any other time realize.

S0 you claim there is nothing I can say that will change your mind, you claim I do not understand anything and have lost the plot as it were,  so why are you even responding?  I firmly believe an ongoing discussion where you consistently re-evaluate and reword your arguments until a mutual understanding can be at least agreed upon if not the actual condition.  but if you are only here to make it your point or no point then I am not interested.

 

6 hours ago, leadeater said:

I can read exactly zero reviews and understand there is a large difference between those two products. Without a review of the RTX 3060 8GB I could not do the same, it is currently impossible to be informed in any about how different it might be to the RTX 3060 12GB.

 

No 128bit vs 192bit is not sufficient to inform me of that.

 

Yes I can, that were not true then Nvidia would be guilty of false advertising.

 

I should consult reviews but I do not need to consult them to know which is better. I can get other important information from reviews but that has nothing to do with this and as much as you want to try and bring that in to the conversation it is not relevant to the issue between the RTX 3060 8GB and RTX 3060 12GB naming.

 

They literally already do, right now and in the past. What on earth are you saying?

 

So you do not need reviews to know which card is better, but you also need nvidia to change the name so you can know which card is better without reviews?  And you think my arguments are piffle.

 

6 hours ago, leadeater said:

If this is the case then simply exist the conversation.

 

 

I think I will. These replies are getting too long and way to much of these posts have more to do with the type of argument rather than the specifics of the content. And I likely just as guilty,  I do respect your input on this forum, most of the time anyway. 

 

6 hours ago, leadeater said:

You would be lying if you said you hear people regularly say out loud the VRAM amount for any typical graphics card and you'd be lying if you said that VRAM capacity is normally included in review graphs.

In reviews,  yes.  If you want to talk about graphs, if there is only one GPU i.e 3060 in the review then they will not mention the ram (because there is no need, there is no confusion as you have already been given the details of that model). But when there are two that have different rams amounts they ALWAYS do.  i.e all the 1060 6G and 10603G reviews.  Here's an example:

 

https://www.gamersnexus.net/hwreviews/2604-gtx-1060-3gb-vs-6gb-benchmark-review/page-3

 

When I was considering a 1060 (went with R570 in the end), I looked at a lot of reviews for the 1060 both the 3G and the 6G versions.   I know that model gets raised a bit with naming issues, however that was another one where I never struggled to find the information and never confused the two.

 

[MageTank]

3 hours ago, mr moose said:

But they all do say it in the review.  If a consumer isn't confirming that the review is for the model they are looking at then I don't what to tell anyone.  Every review I have read mentions the ram, bandwidth, power configuration etc.  Sure they don;t always need to put it in the graphs, but if you read any review or scientific research paper you will quickly see that that is the norm.   The average consumer has somehow managed to work it out alright up until now. I am not sure why all of a sudden it is a problem.

If customers did their research before buying things, I'd likely be out of a job, lol. The simple truth is, our caveman brains process images better than text. We see graphs with nice colors and think (ha, bigger bar better) and the name next to it informs our decision. To be frank, the average consumer doesn't care why something is better, they only care that they are getting the best value for their money. Take this forum for example. You and I have both been around here long enough to have our fair share of arguments/debates with people. How often do we provide endless amounts of evidence with links, only to be countered by an image with a graph? lol.

 

3 hours ago, mr moose said:

And doing that would not change anything, people would still have to look up reviews to know how it performs,  but going on the arguments in this thread the logic would be if they call it a 3055 then people will assume it performs better than a 3050 and worse than a 3060.  That logic quickly falls down as soon as you realise that the name will never accurately depict performance no matter what they call it.

You are correct. It won't accurately depict performance, as in, a customer won't know how it performs relative to one another. However, if they follow that naming scheme and ensure that for example, a 3060 is always faster than a 3055 (even if only by a few %), then a customer can at least assume by the name alone that bigger number = better and be correct in that assumption.

 

I am not asking that they make this idiot proof. I'll say it again, the moment you try to make something idiot proof, they build a better idiot. That's an exercise in futility. That said, you can always make something easier for people and it have no negative consequences on your bottom line. Simplifying the naming conventions used on hardware would go a long way towards this goal. I don't just mean this in GPUs, we need to see this with CPUs and motherboard chipsets too.

[HenrySalayne]

5 hours ago, mr moose said:

And doing that would not change anything, people would still have to look up reviews to know how it performs

Exactly! And they will find new and original reviews covering this specific product! Thank you for finally getting to the point.

[leadeater]

18 hours ago, mr moose said:

Or you just don't like my conclusion. 

No, your conclusion doesn't at all include half of what I said and only covers the parts you want to harp on about. I have covered much more and your conclusion includes none of it.

 

18 hours ago, mr moose said:

S0 you claim there is nothing I can say that will change your mind, you claim I do not understand anything and have lost the plot as it were,  so why are you even responding?  I firmly believe an ongoing discussion where you consistently re-evaluate and reword your arguments until a mutual understanding can be at least agreed upon if not the actual condition.  but if you are only here to make it your point or no point then I am not interested

The problem is you won't even acknowledge or talk about any points I have made so why wouldn't I be completely uninterested in discussing it with you anymore. I have done exactly this multiple times yet you leave out entire arguments and reasoning and will only address it in a way that has to do with what your argument is about, your point that you want to make, which is of course in your interest to do but so it is mine.

 

I've said I do not agree with your reasoning and I think what you have presented is bad reasoning and logic. Either put forward something better or that opinion won't change. Arguing that as text RTX 3060 8GB and RTX 3060 12GB are different is not in dispute. So I would again ask you to go back and actually read what has been said and the past evidenced examples of why this has been a problem so is a problem not might be a problem.

 

And the reason as to why I told you this, to hopefully actually wake you up and realize that is the case so you can go back and actually read and attempt to understand what has been said and maybe redress the points. I'll know you have done this when you start addressing things that you have not done so until now which are massive parts of the discussion you continue to either ignore or not understand.

 

18 hours ago, mr moose said:

So you do not need reviews to know which card is better, but you also need nvidia to change the name so you can know which card is better without reviews?  And you think my arguments are piffle.

You may as well be arguing for naming products after celestial bodies then. No it is not nonsense. Those model names actually mean and represent something and have always. Not that naming them after celestial bodies would be a problem itself so long as the name isn't used twice for different products, however then yes I wouldn't be able to look at the name and know which is better from that.

 

RTX 3060 is a product name that represents a product configuration. RTX 3070 is a product name that represents a product configuration. Nvidia through their own product information and marketing along with history establishes that yes indeed an RTX 3070 would be faster than an RTX 3060 and I do not need reviews to know this.

 

It is you saying otherwise that is nonsense. Seemingly you do not understand what product naming is for and why. You were the one arguing that a product or model name cannot and could never inform people of the detailed product configuration when that is exactly what it does. Product naming becomes a problem when you start making variants like this so that name is no longer sufficient and accurate as it used to be.

 

18 hours ago, mr moose said:

You can't seem to fix a proof problem either.    You seem to be stuck in a dilemma that everything is black and white all the time.  Nvidia reused a GPU name and spec in a dishonest way before, as I have pointed out and said I agree with you on it,  but that doesn't mean every name is dishonest if the underlying specs aren't what you feel they should be.  So long as the name is unique to the product. 

Bad argument, doesn't address the actual issue. RTX 3060 has been used twice and for two different products therefore not unique.

 

18 hours ago, mr moose said:

I did not ignore anything,  I acknowledge that when the naming conventions are not honest it is a problem, but when they are honest it is not a problem.  You can post or not post anything you like using whatever reasoning you feel, but that will not change the fact that what you have posted does not prove anything as absolute as you are claiming against my postulations.

Yes it does and has. This is a you problem as stated. I have in fact proven it with historic evidence of literally the exact same situation.

 

18 hours ago, mr moose said:

Never said that, I said a change in ram size establishes a change in the product.  It can now be identified as something different so whatever else changed should not be a problem, i.e the consumer will find out everything that changed when they look up the specs of the model or reviews of the model.

GTX 1060 5GB reviews where please?

 

 

18 hours ago, mr moose said:

The information is there, how can it be that information that is so easy to access both in the reviews and on the makers website equals not clearly supplying said information. 

False it is not there. Spec sheet is not enough. You tell me how much extra production information on Nvidia website for the original product exists. Is the exactly equivalent amount present for his model? No right? Any answer other than no is a lie.

 

Come on now at least be a little honest. Just a tiny small amount. Not much to ask for.

 

18 hours ago, mr moose said:

How is that not clearly identifying each product as being different.  the 12G version has been around since when start of this year end of last?  the 8G version has been out a month 2 at the best? I am reading a 3060 12G review that was released in june this year.   You can't just say "look at all these reviews of the 3060 that don;t differentiate between the 2 models when those reviews were all released long before the 8G version was anything.    You are conflating conditions that are not born of naming issues.

That is specially part of the damn issue and why it's important to NOT use conflicting names like reusing RTX 3060 in a situation like this to avoid the issue.

You just said what the problem is and why/how it's a problem yet don't seem to care how easy it is to mitigate the problem by not using RXT 3060 naming.

 

Because the 8GB did not exist is exactly why avoiding naming conflicts is important. It is a naming issue because they choose not to prevent it from being an issue. And more specifically it's an issue because they are not the same product with different VRAM capacities. The naming is NOT representative of the product no matter how much you want to argue "uniqueness" in names. I'm sorry but RTX 3060 does actually represent something.

 

18 hours ago, mr moose said:

I've always assumed that when a product has a different spec for anything and difference in price that there will be a difference in performance.  ALWAYS.  no matter if its just the ram size or a GT postfix. 

And assuming VRAM capacity like this is like I said a bad assumption supporting misinformation and makes no attempt to make sure technical specifications mean what they actually are and people understand things correctly. I'm sorry but my opinion on what you just said makes you a bad PC advocate, ALWAYS.

 

Yes it matters.

 

18 hours ago, mr moose said:

I think I will. These replies are getting too long and way to much of these posts have more to do with the type of argument rather than the specifics of the content. And I likely just as guilty,  I do respect your input on this forum, most of the time anyway. 

If you had at least made a single attempt to address and understand arguments and points outside of your own then it wouldn't have gone that way. It's your own fault.

[leadeater]

12 hours ago, mr moose said:

In reviews,  yes.  If you want to talk about graphs, if there is only one GPU i.e 3060 in the review then they will not mention the ram (because there is no need, there is no confusion as you have already been given the details of that model). But when there are two that have different rams amounts they ALWAYS do.  i.e all the 1060 6G and 10603G reviews.  Here's an example:

 

https://www.gamersnexus.net/hwreviews/2604-gtx-1060-3gb-vs-6gb-benchmark-review/page-3

 

When I was considering a 1060 (went with R570 in the end), I looked at a lot of reviews for the 1060 both the 3G and the 6G versions.   I know that model gets raised a bit with naming issues, however that was another one where I never struggled to find the information and never confused the two.

Correct and the reason why that is done and why this example is less of a problem is both variants existed from day one so everyone knew it was necessary to display the data in that way. Yet outside of reviews people mostly just talk about GTX 1060 and don't stipulate which one. Even then with his example because the product configuration is different, not VRAM capacity, they shouldn't both be named GTX 1060 either.

 

The whole entire reason you never struggled was because both models existed and were released at the same time, were globally available and were officially documented fully and acklowdge by Nvidia. This is not the same as say for example the GTX 1060 5GB (which you as yet never addressed) nor this RTX 3060 8GB.

 

These situations simply are not the same which is why the outcomes are different.

[Godlygamer23]

The fact that there are people literally defending NVIDIA on this nonsense is astounding. Performance and VRAM amount should be two different things. They did the same nonsense with the 1060, where the VRAM amount difference led to a massive difference in performance because the GPU on the lower capacity model was more cutdown. If someone was looking at two PCs with two differing amounts of RAM(say 16GB vs 32GB), would you expect an inherent difference in bandwidth? Maybe, if they used crappier modules. But to me, I would expect them to use the same general modules, just with lesser capacity, unless it was very clear on the specifications that they were using different RAM modules with lower clock speeds and looser timings. 

 

In the case of graphics cards, if there is a major difference in performance that is not related to the VRAM being filled up and paging to system RAM, GPU core, memory bandwidth, etc, but simply due to the difference in capacity, I would absolutely expect the same performance. It's not clear in the model what the difference _actually_ is, and I think it would be better if they put "3060 192-bit" instead of just the lower capacity. It makes it more clear that there is an actual difference. 

 

The other thing to think about is the price difference - there isn't one. Which means the lower performing product is also a worse value, so you have effectively a more expensive product for the performance you're actually getting.

 

Again, the fact that people are _defending_ this is absolutely astounding.