Nvidia GeForce RTX 3050 Review: Availability is not guaranteed

In other words, you are putting yourself in a corner. I don't see the point of defending a product that is ill suited for the target market. While it may sound harsh, but consumers and reviewers should be feeding back to the product maker. AMD's pitch is that they made a card "for gamers", and not suited for miners. In my mind, cutting the VRAM could alleviate the mining issue, but what is the point of cutting on the PCI-E bus, removing the video encoder/ decoder? Are miners utilising the full x16 or x8 bus in the first place or are they using the mining rig to also stream movie at the same time they are mining?
Like I already said, cutting encoders/decoders and PCIe bus helps to make chip smaller. To illustrate:

GTX1650: 200 mm²
RX470/480/570/580/590: 232 mm²
GTX3050: 276 mm²
6500XT: 107 mm²

There you have it.
This is just an excuse for AMD to cut cost in my opinion, and not" for gamers". Please remember, each card targets a specific market. This card again, is meant for budget gamers. If you are telling to me look for other options, what other options are there that is budget and don't sux this much when all I have is a PCI-E 3.0 x16 slot?
Not just cost but die area too. If you haven't noticed, wafers are short supply. Only way to get more chips per wafer is to reduce die area. That's why this card has only PCIe 4.0 x4. In case you don't have PCIe 4.0 capable motherboard, look for older cards.
Isn't it penalising budget gamers? When 1 reviewer says its bad, there could be something wrong with the reviewer. When almost all the reviews have fairly similar concerns with the product, there is certainly an issue with the product. Some reviews may be milder and trying to find some positives with the product, but the limitations certainly cannot be overlooked. I am no reviewer, but objectively speaking, the card does NOT suit most budget gamers, and certainly not some olive branch that AMD has stuck out to help budget gamers. And when you are telling me to look for other options, basically you acknowledged that it is not suitable for most people.
Those who see it has everything wrong just don't understand current market situation. AMD released card they could do with minimal effort, that is available and is probably best choice on that price range. Anything AMD could do better, would have meant drawbacks elsewhere:

More features = less availability
New design for desktop = RDNA3 comes later
More RAM = miners buy everything
No release 6500XT at all = ...

Basically, AMD made all right choices with 6500XT at current situation. Like it or not. Bashing 6500XT for everything just proves reviewers lack understanding.
And when you are telling me to look for other options, basically you acknowledged that it is not suitable for most people.
Please tell me what video card is suitable for most people.
 
May be you should try and be the reviewer? When you review something, it has to be compared with another product and some assumptions, I.e. pricing, etc. No reviewers will look price and availability in every single country before coming out with a conclusion of the product. If that is the case, the review will never be released because prices differs from country to country, so is the availability, or other moving factors.
See die sizes that I posted above. Reviewers complain about lack of features when chip is clearly designed to be ultra small. And that's evident just looking die size. Availability wise, this is ultra small chip that is also first AMD's GPU made with new manufacturing tech (TSMC 6nm). Reviewers could assume availability would be good. How about RTX3050? That's cut down RTX3060. Basically Nvidia sells every fully functional GA106 die as RTX3060 and those that are defective sell as RTX3050. Now, since RTX3060 is short supply, what do you expect RTX3050 availability to be? This is not rocket science.

Again, I find people try to defend AMD's RX 6500 XT by putting themselves in a "no choice" situation. Hey AMD, I have no choice, so I think the card is ok, and I will buy it since there are no options. But is that the right feedback? So you mean to say if AMD is going to continue to pull a fast one on you next time, you will welcome their cost cutting measures?
Please tell me what AMD could have done better with 6500XT? Like said above, only bad choices were available. Basically you're down with two options: release 6500XT or do not release it at all. I simply cannot see why more options is bad on current situation.

Not just Cost but also Capacity.
No wonder progress is stagnated because people don't expect a new generation product to work better than its predecessor, and happy with the lack of progress or even regression in this case. The product can certainly game if you have a PCI-E 4.0 or 5.0 slot in your system, but warnings from reviewers should also consider who is the target market of this product. People should not be using PCI-E 2.0, but 3.0 is common. So you pay xx amount of a card expecting 100% performance, only to be crippled to about 60 to 70% when using in an old system which the graphic card died and the only cheap card is the RX 6500 XT.
Those people who do not expect that also should have predicted current situation and buy card before shortage got too worse. I did that.

Those who have PCIe 3.0 system also get nice performance boost when they upgrade their antique motherboard. Remember, PCIe 3.0 is 10 year old technology, PCIe 3.0 supporting motherboards started to appear 2012. No need to defend it.

Reviewers should mention about that PCIe bus performance penalty but hey, even with PCIe 3.0 6500XT is faster than GTX1050Ti. Also that doesn't matter those who have PCIe 4.0 system. And certainly it doesn't make 6500XT trash.
Are you using one with PCI-E 3.0? How would you feel if you are? And don't forget, almost everybody uses a PC, but only a tiny fraction is up to date with new technology. This is almost like AMD pulling a fast one on unknowing users and hopefully people buying a new PC don't buy one with PCI-E 3.0 support only.
How many new systems really are PCIe 3.0 only? AMD have offered B550 for mainstream 1.5 years now, high end X570 even longer.

And again, PCIe 4.0 x8 would have meant larger die and worse availability. AMD had no chance to win here.
 
This is all I'm going to say about review scores and this particular comparison, noting that context is important (as in, reading the full review or at least the full conclusion) and Steve's video addresses the comparison in an extended form...

The Radeon received a 20/100 because even if it becomes available at the intended price, it's not worth buying, it's crippled, and we believe it's a bad (really bad) GPU release overall. We didn't know what availability was going to look like ahead of release (day one review), but we anticipated it was going to be inflated and probably not worth buying beyond X price... per the review's conclusion.

The RTX 3050 received an 80/100 because as a product, it is an overall good GPU. It performs relatively well (slotted in between the GTX 1660 Super and RTX 2060), it has DLSS to its advantage. etc. as said in the review. We didn't know what availability was going to look like ahead of release (day one review), but we anticipated it was going to be inflated and probably not worth buying beyond X price... per the review's conclusion.

We always take value into consideration when assessing a product, and in the case of GPUs, it's a crazy market. We have dozens of articles posted in the last year about this, so rather than repeating ourselves over and over, we understand our readers are aware of the situation and the overall context that applies when we are reviewing new GPUs.
 
The Radeon received a 20/100 because even if it becomes available at the intended price, it's not worth buying, it's crippled, and we believe it's a bad (really bad) GPU release overall. We didn't know what availability was going to look like ahead of release (day one review), but we anticipated it was going to be inflated and probably not worth buying beyond X price...
Perhaps that's true in some kind of alternate reality. In todays market no card can offer better performance for same price when leaving used market out of comparison. No matter how "bad" you consider it, unless you simply cannot tell better card for same price, then card is actually pretty good. And looking at review you cannot. "Better" options are used market and something that costs 100 dollars more and has worse availability. 20/100 for best card for certain price is simply absurd. That's a Fact. Your personal Opinion is another question but review scores should be more fact and less opinion.
 
Waiting for next 20/100 score for best product on certain price category :joy:
It's not the best product in it's category, arguably yes it is for a 'new' product because it's the only product, but that doesn't mean it's great by default and as we've pointed out there are better make-do second hand options. If your argument is it exists and therefore a score of 20/100 is unfair, well that's an extremely weak argument.

You appear to have unlimited time to constantly argue on forums so forgive me if I don't bother circling back to this one, but we've given you our justification for the reviews conclusion/score, if you don't like it that's fine, but it's not going to change anything. Our reviews are based on very in-depth testing and analysis.

There's about 2, maybe 3 review outlets globally that gave the 6500 XT a positive review, so they might be more inline with how you'd like products reviewed. PCworld & Zach's Tech Turf, both did far more limited testing that us and mostly brushed over the products issues, so have at it.
 
It's not the best product in it's category, arguably yes it is for a 'new' product because it's the only product, but that doesn't mean it's great by default and as we've pointed out there are better make-do second hand options. If your argument is it exists and therefore a score of 20/100 is unfair, well that's an extremely weak argument.
Your only argument here seems to be "there are better second hand options available". That is extremely weak argument because second hand market products usually:

1. have no warranty
2. you can't get them right away
3. you must search what's available
4. there might be condition issues and even working issues
5. buyer might not provide you what he promises
6. you didn't offer any proof that those "better deals" are in fact real ones

Many times you could say new low end offering is bad "because there are better offerings on second hand market". And at same time GTX3050 scored 80/100 despite there are better deals on second hand market too. That also means 6500XT got -60 points because of something else than "second hand market has better products". And you simply failed to tell why.
You appear to have unlimited time to constantly argue on forums so forgive me if I don't bother circling back to this one, but we've given you our justification for the reviews conclusion/score, if you don't like it that's fine, but it's not going to change anything. Our reviews are based on very in-depth testing and analysis.
You are completely missing one main aspect of 6500XT: small die size. On your article you argued about PCIe x4, lack of decoders/encoders etc. Now, let's look at die sizes:

RTX3050: 276 mm²
RX6500XT: 107 mm²

Now you can argue about lack of features but you cannot argue fact that 6500XT is very small chip. Also you cannot argue with fact that much smaller chip gives much more working chips per wafer assuming defect rate is constant. And here we go with feature aspect:

Add encoders/decoders: more die area
Add more PCIe: more die area
Add (something): more die area

More die area = less working chips per wafer = worse availability.

You say that it lacks features but at same time completely dismiss very small die size that leads to better availability. Your in-depth analysis missed this part totally.
There's about 2, maybe 3 review outlets globally that gave the 6500 XT a positive review, so they might be more inline with how you'd like products reviewed. PCworld & Zach's Tech Turf, both did far more limited testing that us and mostly brushed over the products issues, so have at it.
Like said above, your analysis missed very important aspect so defending yourself because most also missed same thing does not apply.

I have done this. I have reviewed products long time ago. I know it's time consuming. I know you have hard time benchmarking many products time to time. That also means sometimes errors happen, happened to me too. Main question is, how those errors are fixed.

And in reality we live, RX6500XT is selling pretty well (looking at some stores' units available info) and at same time is available too. Amazing on current situation. AMD certainly did Something right and despite you're saying it's not worth buying, people do buy it.
 
“Your only argument here seems to be "there are better second hand options available". That is extremely weak argument because second hand market products usually:”

No, again we have an entire review for you to read as well as a follow-up video linked above which explains all the reasons why we think the 6500 XT is a bad product.

In a nut shell the argument is there are better products that are now 6 years old and you can buy those for less, or newer products which are also better such as the 5500 XT and GTX 1650 Super for roughly the same price.

If you can’t beat products from 6 years ago, we’d rather you not release anything at all, especially if it’s inferior in almost every way. We’re not going to support companies like AMD taking advantage of the current market position. If you don’t think AMD’s making good money off the 6500 XT silicon, well frankly you’re a fool.

We’re aware of the die size and noted it in the review, the die size is not a pro for the consumer, it’s a pro for AMD as it allows them to maximise margins. Navi 24 was intended as an ultra-cheap solution for laptops, AMD clearly decided it would be a smart business choice to change gears and sell it for a lot more as a desktop product.

Had it been planned for the desktop it would be a little larger and included all the missing features.

“You say that it lacks features but at same time completely dismiss very small die size that leads to better availability. Your in-depth analysis missed this part totally.”

Yeah nah it really didn’t. You’re using that as a justification for AMD making a **** product.

Sales figures don’t mean a product isn’t rubbish. Anyway I'm done here, you can have the win, I was wrong along with almost every other experienced review outlet.

I've said all I want to, for now I look forward to comparing the 5500 XT and 6500 XT in 2 years time, in the latest games at 1080p low. It shall be very interesting to see just how good of an investment the 6500 XT turns out to be.
 
Last edited:
“Your only argument here seems to be "there are better second hand options available". That is extremely weak argument because second hand market products usually:”

No, again we have an entire review for you to read as well as a follow-up video linked above which explains all the reasons why we think the 6500 XT is a bad product.

In a nut shell the argument is there are better products that are now 6 years old and you can buy those for less, or newer products which are also better such as the 5500 XT and GTX 1650 Super for roughly the same price.

If you can’t beat products from 6 years ago, we’d rather you not release anything at all, especially if it’s inferior in almost every way. We’re not going to support companies like AMD taking advantage of the current market position. If you don’t think AMD’s making good money of the 6500 XT silicon, well frankly you’re a fool.
No matter how good or bad product it is, you simply have no better options at same price. Reality check again: you have limited budget for new graphic card (forgetting second hand market for good reasons). Options are: GTX1050Ti and RX6500XT. Those are only options, which one would you buy? Sometimes you just have to take better option, no matter how bad it is.

And while you say AMD should have not released product that is worse than 5 year old ones, you failed to tell what AMD could have done better. Let's see AMD's options:

- Create new RDNA chip for TSMC 6nm: well, RDNA3 should come later this year and RDNA2 launched 1 year and 3 months ago. Not something reasonable to expect. Also since TSMC 6nm is short supply, chip must be very small.

- "Re-launch" RX580 or RX5000-series cards: GF 12/14 nm is very short supply and so does TSMC 7nm. Not an option.

-Design new GPU for something new manufacturing tech (like TSMC 16nm): Even Nvidia decided to launch old products, not creating new ones...

- Add more features on mobile chip so it's better suitable for desktop too: this would add more die area, increase power consumption and make availability worse.

- Something else, like what?

You're saying AMD did something wrong but cannot tell what they should have done better. I already proved not to launch it at all would make situation worse.
We’re aware of the die size and noted it in the review, the die size is not a pro for the consumer, it’s a pro for AMD as it allows them to maximise margins. Navi 24 was intended as an ultra-cheap solution for laptops, AMD clearly decided it would be a smart business choice to change gears and sell it for a lot more as a desktop product.
Not just price but availability too. And as stated above, there are reasons for that.
Had it been planned for the desktop it would be a little larger and included all the missing features.
So you are saying AMD should create new RDNA2 chip now when RDNA2 launch is 15 months old and RDNA3 should come later this year? At same time Nvidia launched RTX3050 that is based on die that launched 11 months ago. No, your expectations are not reasonable.
“You say that it lacks features but at same time completely dismiss very small die size that leads to better availability. Your in-depth analysis missed this part totally.”

Yeah nah it really didn’t. You’re using that as a justification for AMD making a **** product.
No, I don't. You do it. You are saying AMD made bad product but fail to tell what reasonable alternatives AMD had. Even better if it's something Nvidia did.
Sales figures don’t mean a product isn’t rubbish. Anyway I'm done here, you can have the win, I was wrong a long with almost every other experienced review outlet.

I've said all I want to, for now I look forward to comparing the 5500 XT and 6500 XT in 2 years time, in the latest games at 1080p low. It shall be very interesting to see just how good of an investment the 6500 XT turns out to be.
You only see 6500XT as bad product and say AMD is greedy etc. Again, failing to say what AMD could have done better and be reasonable (don't expect AMD to design new RDNA2 die just for desktops at this time etc) and at same time bashing AMD for creating product that IS available like GTX3050 has, is like...

My point is that no matter bad product 6500XT is, it has good place on current market AND AMD was justified to launch it. Feel free to prove me wrong. So far you're only saying AMD was "greedy" and should not have released this product, all without any Real arguments.

I could say right away that on current prices vs previous 5500XT prices, 6500XT is bad investment. Tbh I'm not even interested to see how bad it looks since I know it looks bad. But, current situation is not normal and there are endless "bad deals" available, too bad those are best in this situation.
 
And while you say AMD should have not released product that is worse than 5 year old ones, you failed to tell what AMD could have done better. Let's see AMD's options:
It's a hardware review, not a job application for AMD product manager.

Here is the new 6500XT. Considerable effort is spent testing it. We learn that it has very disappointing performance and functionality. It seems perfectly appropriate to let readers know about that finding while not at the same time offering some sort of proposal for an alternate AMD business plan.
 
Reviewer is blaming AMD for doing something wrong and same time refusing to tell what they should have done better (being reasonable, of course):

No, it's not a lie. It's a fact. Believing fact is a lie is enough to make claim pointless.
The worst part is reviewers not getting the products actual position in the market.

The 3050 never was a 6500XT competitor. In terms of *actual* prices it competes against the 6600, just like the 3060 competes against the 6600XT and the 3060Ti against the 6700XT. This realization should not be anything new by now.

nVidia‘s current 6500XT competitors are the 1050Ti and 1650.
 
No one is buying the Nvidia GeForce RTX 3050 despite the reviews.

Why isn't this TechSpot's article headline?
Who knows, maybe not allowed by Nvidia marketing dept?

And its not only TS, all YT channels and tech sites seems to be under this “trance” that only nvidia produces good gpus and never, ever dare in saying one valid criticism (besides the little rant about that email when HU was temporarily “banned” by nvidia).

Just look at the many valid points provided above by HardReset and Irata and yet Steve and the staff simply ignored the possibility that they were bias or wrong.
 
No one is buying the Nvidia GeForce RTX 3050 despite the reviews.

Why isn't this TechSpot's article headline?

Well, to be fair Mindfactory did get one RTX 3050 model in (MSI Gaming X) for €499 and they did sell 80 so far. So that‘s more than zero and pretty respectable considering they are the only major German retailer who has them in stock (OK, that last part was sarcasm).

Funny enough, they sell a 6600 for €4 less.
 
Well, to be fair Mindfactory did get one RTX 3050 model in (MSI Gaming X) for €499 and they did sell 80 so far. So that‘s more than zero and pretty respectable considering they are the only major German retailer who has them in stock (OK, that last part was sarcasm).

Funny enough, they sell a 6600 for €4 less.
Sales data so far from Mindfactory:

RTX 3050: one model, price €499, total sales 80 units
6500XT: 5 models, prices €279/289€, total sales 215 units
 
Sales data so far from Mindfactory:

RTX 3050: one model, price €499, total sales 80 units
6500XT: 5 models, prices €279/289€, total sales 215 units
To add: Mindfactory only shows the models they have in stock, so sold out models are not included.

According to Tech Epiphany Mindfactory sold 280 units of the 6500XT in week three 2022 (Jan 18th-24th, so the release week) making it their #1 seller that week.

 
I think it’s disgusting that Techspot didn’t compare the 3050 to a 6500XT in the ray tracing tests because they didn’t want to “be so cruel”. You mean you don’t want to see AMD get its *** handed to it? I’ve seen 6500XTs on sale in shops now and they have “ray tracing capable” written on the back of the box. It’s the biggest lie I think I’ve ever seen written on the side of a graphics card box and Techspot for some reason isn’t calling that bullshit out. Especially as I have no doubt Techspot would call it out if it was the other way around. They would “be so cruel” to Nvidia, their beef with them is highly publicised.
raytracing is still not there yet it kills your speed so much you would have to be a masochist to use it
 
Mindfactory‘s week 4 sales by Tech Epiphany are out. RX 6500 XT sales are down to 150 and in fact lower than e.g. 6900XT sales.
Considering that the closer to msrp models are gone, that‘s understandable.

Also interesting are 1650 sales (140).
3050 showed up at 10 sold, top seller was the 3060Ti with 230 units.
 
Starting to see some angry reaction videos on the 3050 from e.g. The Good Old Gamer and Tech of Tomorrow. The latter pointed out that the $249 Evga 3050 he got retails for over $500 now, asking if it‘s even worth doing a review under these conditions.
 
Starting to see some angry reaction videos on the 3050 from e.g. The Good Old Gamer and Tech of Tomorrow. The latter pointed out that the $249 Evga 3050 he got retails for over $500 now, asking if it‘s even worth doing a review under these conditions.
Or maybe they havent received their loyalty check from nvidia? :)
 
Back