• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

The better late than never OcUK RTX4070 Super review thread!

Hm.... i wonder if i can install the AMD Andriod app on it, if so i can play Forza on my couch through my TV while its being rendered on the Desktop.

What do you think about that Nvidia? Yeah its 2024....
 
Last edited:
Purgatory said:
If your TV is a google Android TV you can use SmartTube to have an add free youtube, I use this on TV instead of Youtube app :-

github.com

GitHub - yuliskov/SmartTube: SmartTube - an advanced player for set-top boxes and tvs running Android OS

SmartTube - an advanced player for set-top boxes and tvs running Android OS - yuliskov/SmartTube
github.com github.com
Click to expand...
Klo said:
You can use it on a Amazon Firestick, that's what I do.
Click to expand...

An update to this.

It worked, i had to put the FireTv stick in to Developer mode, install a downloader, put the SmartTube Github URL in, download, Install..... and it works great, Youtube on my TV and no more ads, thank you :)

If the amount of ads on Youtube was reasonable i wouldn't care and i wouldn't bother...
 
Last edited:
humbug said:
An update to this.

It worked, i had to put the FireTv stick in to Developer mode, install a downloader, put the SmartTube Github URL in, download, Install..... and it works great, Youtube on my TV and no more ads, thank you :)
Click to expand...
I'm glad, its a little convoluted to get working, but worth it!
 
@humbug I’m paying 90p a month for YouTube premium as I subscribed to it “while being in Argentina” :p
That gives me ad free experience on all my devices (especially important on Apple devices), YouTube music and higher bitrate for 1080p videos (1080p Premium).
Totally worth it. ;)
 
subbytna said:
I just paid £13 for a years worth of YT P. remium via Mumbai, just over a quid a month.
Click to expand...

z10m said:
@humbug I’m paying 90p a month for YouTube premium as I subscribed to it “while being in Argentina” :p
That gives me ad free experience on all my devices (especially important on Apple devices), YouTube music and higher bitrate for 1080p videos (1080p Premium).
Totally worth it. ;)
Click to expand...


He pays nothing.... :D

I don't mind ads, if they want to install American confectionary on to my browser and advertise i'm perfectly fine with that, but not at a frequency where it makes the content completely unwatchable, there is no contract in you being the product in return for a service in that, its just narcissistic abuse and its the reason people use adblockers.
Adblockers only exist because web pages became unreadable due to flashing and audio playing banner ads all over every website in place of any actual content.
As usual the people wanting to make money could never make enough of it and went way beyond what's reasonable, and they have the cheek to complain that adblockers are piracy? Adblockers aren't the problem, you are!

Its also one of the reasons why no one watches legacy linier commercial TV, out of an hours viewing you get 18 minutes of original content, its ridiculous, even if people don't understand why they are bored and annoyed with legacy commercial TV.

I'm 47, i'm of that generation where i relax on my couch during the evening and watch a large screen TV, except its not TV that i'm watching anymore, its Youtube and occasionally something from a streaming service.
 
Last edited:
humbug said:
He pays nothing.... :D

I don't mind ads, if they want to install American confectionary on to my browser and advertise i'm perfectly fine with that, but not at a frequency where it makes the content completely unwatchable, there is no contract in you being the product in return for a service in that, its just narcissistic abuse and its the reason people use adblockers.
Adblockers only exist because web pages became unreadable due to flashing and audio playing banner ads all over every website in place of any actual content.
As usual the people wanting to make money could never make enough of it and went way beyond what's reasonable, and they have the cheek to complain that adblockers are piracy? Adblockers aren't the problem, you are!

Its also one of the reasons why no one watches legacy linier commercial TV, out of an hours viewing you get 18 minutes of original content, its ridiculous, even if people don't understand why they are bored and annoyed with legacy commercial TV.

I'm 47, i'm of that generation where i relax on my couch during the evening and watch a large screen TV, except its not TV that i'm watching anymore, its Youtube and occasionally something from a streaming service.
Click to expand...
Yep I'm 41 and completely agree.
 
Can you imagine if Nvidia launched a 16 GB 4070 with 256 bus for £600?
That "dream" would be game over for AMD this gen.
Nvidia are swimming in money, surely an extra 4gb is worth it (annoyingly you can see on the 4070/4070 super motherboard as 2 empty spaces for mem modules).
RTX 2000/3000 users looking to upgrade, have already seen the effect of minimal vram and want their "Upgrade" to not have that issue on whats an expensive card for most. Big failure.
Unless that extra memory eats into cards above performance but I can't see that.
Also Massive difference in build quality of top end 7800 XT vs 4070/super lower end cards (they look like plastic crap)

Really Happy with my Red Devil 7800 XT but that "dream" card would make me jump ship.

Its fair to say they ain't selling that well - I wonder how many and how much its down from what they predicted.
 
Last edited:
kaiserc said:
Can you imagine if Nvidia launched a 16 GB 4070 with 256 bus for £600?
That "dream" would be game over for AMD this gen.
Nvidia are swimming in money, surely an extra 4gb is worth it (annoyingly you can see on the 4070/4070 super motherboard as 2 empty spaces for mem modules).
RTX 2000/3000 users looking to upgrade, have already seen the effect of minimal vram and want their "Upgrade" to not have that issue on whats an expensive card for most. Big failure.
Unless that extra memory eats into cards above performance but I can't see that.
Also Massive difference in build quality of top end 7800 XT vs 4070/super lower end cards (they look like plastic crap)

Really Happy with my Red Devil 7800 XT but that "dream" card would make me jump ship.

Its fair to say they ain't selling that well - I wonder how many and how much its down from what they predicted.
Click to expand...
Is it common for cards to have empty memory slots? If so couldn't you find a cool guy to solder a couple of extra ram chips? Cool guy is the best i can come up with right now, don't hate,be gentle. :D
 
Last edited:
kaiserc said:
Can you imagine if Nvidia launched a 16 GB 4070 with 256 bus for £600?
That "dream" would be game over for AMD this gen.
Nvidia are swimming in money, surely an extra 4gb is worth it (annoyingly you can see on the 4070/4070 super motherboard as 2 empty spaces for mem modules).
RTX 2000/3000 users looking to upgrade, have already seen the effect of minimal vram and want their "Upgrade" to not have that issue on whats an expensive card for most. Big failure.
Unless that extra memory eats into cards above performance but I can't see that.
Also Massive difference in build quality of top end 7800 XT vs 4070/super lower end cards (they look like plastic crap)
Click to expand...
Nvidia can’t simply solder on 4gb more VRAM, it’s down to the bus width and the AD104 chip has a 192 bit bus so can only accommodate 6/12/24gb of VRAM.

For Nvidia to give you 16gb they would have had to use the AD103 which is technically the 104 chip traditionally used on 70 class cards but because Nvidia saw that performance on TSMC 5nm was very good they decided to cash in and use the 60 class 106 die which was renamed to 104 for the 70 cards. This only came with a 192 bit bus similar to the 3060/2060/1060 that came before so realistically limited the cards to a max of 12gb.

Had Nvidia used the 103 for the 4070 AMD would have still been able to compete so no “game over” it would have just meant them selling the 7900XT for £600 which I’m sure they are quite capable of doing since they manage to set a similar figure for their high end 6000 cards built on the top chip but the reality is that AMD are chancers.
 
Last edited:
Joxeon said:
Nvidia can’t simply solder on 4gb more VRAM, it’s down to the bus width and the AD104 chip has a 192 bit bus so can only accommodate 6/12/24gb of VRAM.

For Nvidia to give you 16gb they would have had to use the AD103 which is technically the 104 chip traditionally used on 70 class cards but because Nvidia saw that performance on TSMC 5nm was very good they decided to cash in and use the 60 class 106 die which was renamed to 104 for the 70 cards. This only came with a 192 bit bus similar to the 3060/2060/1060 that came before so realistically limited the cards to a max of 12gb.

Had Nvidia used the 103 for the 4070 AMD would have still been able to compete so no “game over” it would have just meant them selling the 7900XT for £600 which I’m sure they are quite capable of doing since they manage to set a similar figure for their high end 6000 cards built on the top chip but the reality is that AMD are chancers.
Click to expand...

I find it interesting how NVidia squeezed the bus width on the 40-series.

With Ampere, there was a top-down progression of 64 bit "jumps", from 384 bits on the 3090 and 3080 Ti & 12GB, to 320 bits on the 3080, 256 bits on the 3070 and 3060 Ti and you had to drop down to the 3060 before it got to 192 bits.

With Ada, whilst the 4090 still gets the "full fat" 384, there's no 320 this time round and we drop straight to 256 for the 4080 with 192 on the 4070.
 
Vertigo said:
I find it interesting how NVidia squeezed the bus width on the 40-series.

With Ampere, there was a top-down progression of 64 bit "jumps", from 384 bits on the 3090 and 3080 Ti & 12GB, to 320 bits on the 3080, 256 bits on the 3070 and 3060 Ti and you had to drop down to the 3060 before it got to 192 bits.

With Ada, whilst the 4090 still gets the "full fat" 384, there's no 320 this time round and we drop straight to 256 for the 4080 with 192 on the 4070.
Click to expand...

What they are doing is saving on die space by dropping two 32Bit segments from the IMC, this also saves power so they can further save money by putting less power phases on the PCB, and of course less memory IC's, then they flog all of those savings for up to $900, because we don't really complain. Not in a way that hurts them.

The most vocal thing tech jurnoes say about it is that its power efficient, well yes its missing a quarter of it IMC, a couple of memory IC's and a bunch of power phases. Nvidia do something _____ and get praise for it. Genius.
 
Last edited:
Vertigo said:
I find it interesting how NVidia squeezed the bus width on the 40-series.

With Ampere, there was a top-down progression of 64 bit "jumps", from 384 bits on the 3090 and 3080 Ti & 12GB, to 320 bits on the 3080, 256 bits on the 3070 and 3060 Ti and you had to drop down to the 3060 before it got to 192 bits.

With Ada, whilst the 4090 still gets the "full fat" 384, there's no 320 this time round and we drop straight to 256 for the 4080 with 192 on the 4070.
Click to expand...
TSMC 5nm had such good performance that Nvidia decided to sell 70 class as 80 and 60 class and 70.

That would have been ok if they didn’t also jack the prices and turn what could have been one of Nvidia’s best generations in terms of price to performance into probably the worse ever.
 
Joxeon said:
TSMC 5nm had such good performance that Nvidia decided to sell 70 class as 80 and 60 class and 70.

That would have been ok if they didn’t also jack the prices and turn what could have been one of Nvidia’s best generations in terms of price to performance into probably the worse ever.
Click to expand...

4nm, its on TSMC 4nm, RDNA3 is on 5nm, and 6nm.
 
Last edited:
Basically nvidia thought they could get away with low specs and slightly better than last gen performance by using framegen to showcase ada smashing ampere. Sales didn't pan out though and this is why 5000 series will be another ampere moment without the crypto bs.
 
Joxeon said:
It’s an enhanced version of TSMC 5nm called N4 but it’s not 4nm, kind of like with tsmc 7nm having N7 N7P N7+ etc.

Click to expand...

That's talking about N4P, an enhancement of N4, which in turn is an enhancement of N5, which is an enhancement of N6 which is an enhancement of N7.... they can all trace their roots back to 28nm.

You're calling it 5nm because you want to imply there is no difference, that's idiotic, there is, its denser and more efficient than N5.
 
humbug said:
That's talking about N4P, an enhancement of N4, which in turn is an enhancement of N5, which is an enhancement of N6 which is an enhancement of N7.... they can all trace their roots back to 28nm.

You're calling it 5nm because you want to imply there is no difference, that's idiotic, there is, its denser and more efficient than N5.
Click to expand...
It’s not idiotic at all, it’s all written by TSMC themselves. They clearly say N4 is part of the 5nm family.

“TSMC continues to expand its 5nm technology family to meet diverse customer demands. These include N5P, N4, and N4P for better power, performance and density, N4X, a process tailored for high performance computing (HPC) applications, and N5A for automotive applications.”

 
Last edited:
You must log in or register to reply here.
Back
Top Bottom