NVIDIA 4000 Series

[CAT-THE-FIFTH]

The RTX4070TI isn't a 70 series card but a rebadged 60/60TI series card. The GPU is under 300MM2 too like a GA106. People keep forgetting Nvidia released an AD103,so the AD104 actually is now the same position as the GA106 in the lineup and the AD103 is basically where the GA104 was positioned.

But more importantly look at the amount of shaders and memory bandwidth relative to the top end dGPU in the range,ie,the RTX4090.

But even ignoring die size,just look at the shader count:

NVIDIA GeForce RTX 4090 Specs

NVIDIA AD102, 2520 MHz, 16384 Cores, 512 TMUs, 176 ROPs, 24576 MB GDDR6X, 1313 MHz, 384 bit

www.techpowerup.com

NVIDIA GeForce RTX 4070 Ti Specs

NVIDIA AD104, 2610 MHz, 7680 Cores, 240 TMUs, 80 ROPs, 12288 MB GDDR6X, 1313 MHz, 192 bit

www.techpowerup.com

The RTX4070TI has 47% of the shader count of the RTX4090,half the VRAM capacity and 50% of the memory bandwith.

Now look at the RTX3090TI vs the RTX3070TI:

NVIDIA GeForce RTX 3090 Ti Specs

NVIDIA GA102, 1860 MHz, 10752 Cores, 336 TMUs, 112 ROPs, 24576 MB GDDR6X, 1313 MHz, 384 bit

www.techpowerup.com

NVIDIA GeForce RTX 3070 Ti Specs

NVIDIA GA104, 1770 MHz, 6144 Cores, 192 TMUs, 96 ROPs, 8192 MB GDDR6X, 1188 MHz, 256 bit

www.techpowerup.com

The RTX3070TI has 57% of the shader count of the RTX3090TI and 60.3% of the memory bandwidth of the RTX3090TI.

Now look at how it is compared to the RTX3060TI:

NVIDIA GeForce RTX 3060 Ti Specs

NVIDIA GA104, 1665 MHz, 4864 Cores, 152 TMUs, 80 ROPs, 8192 MB GDDR6, 1750 MHz, 256 bit

www.techpowerup.com

It has 45.2% of the shader count of the RTX3090TI and 44% of the memory bandwidth.

The RTX4070TI relative to the RTX4090,is basically what the RTX3060TI was relative to the RTX3090TI for the previous generation,with a die size of the RTX3060/GA106 series.

 

[muon]

It's funny how no one gets hung up on die sizes when it comes to CPUs.

 

[sniperone]

I bought myself a PS5 to tide me over until PC hardware becomes a bit more sensible again. And frankly it feels like a bit of a bargain in comparison.

It’s an absolute bargain to be fair and I imagine many more pc gamers will jump ship if things do not improve

 

[Alt0153]

It’s an absolute bargain to be fair and I imagine many more pc gamers will jump ship if things do not improve

depends how many of us like paying £80 for a game or whatever console going rate is now

 

[RavenXXX2]

£800+ and you have to turn down IQ, nevermind futureproof, it's not present proof.

The memory expansion, on the other hand, is less impressive. 12 GB is 50 percent more VRAM than the GeForce RTX 3070 Ti, but that doesn't change the fact that the Radeon RX 7900 XT offers clearly more with 20 GB and 12 GB no longer consistently enable optimal operation today and in the future will be too little from time to time. For example, if you activate or switch DLSS in Spider-Man Remastered in Ultra HD including ray tracing afterwards, you will immediately run into a VRAM shortage. 12 GB in the current generation is probably pretty much what 8 GB was in the old one: mostly sufficient, but sometimes too little after 2 years at the latest.

 

[CAT-THE-FIFTH]

It's funny how no one gets hung up on die sizes when it comes to CPUs.

Because the die size relative to the top determines positioning.

During Ampere you had GA102,GA104 and GA106. AD103 had a limited release and never really found it's way to desktop.

Ada Lovelace has AD102,AD103 and AD104. AD104 is third in line in the Ada Lovelace lineup like GA106 was third in line. Both were under 300MM2. The AD104 is really a 106 series die of yesteryear.

Plus it gets even more interesting when you look at shader count.

The RTX4070TI has 47% of the shader count of the RTX4090,half the VRAM capacity and 50% of the memory bandwith. The RTX3060TI has 45.2% of the shader count of the RTX3090TI and 44% of the memory bandwidth.

In terms of relative performance to the top it gets more interesting.The reason Nvidia launched the RTX4070TI as AIB models,because they are pre-overclocked out of the box. But I did some number crunching,from TPU relative performance figures.

The RTX4070TI is around 70% of an RTX4090 at qHD and they are all AIB models. The RTX3060TI FE was 63% of an RTX3090TI and 68% of an RTX3090 at qHD. With AIB models a few extra percent could be added.The RTX3060TI FE launched at £369. AIB models started at £420ish.

 

[z10m]

It’s an absolute bargain to be fair and I imagine many more pc gamers will jump ship if things do not improve

Meh my PS5 is sitting in the corner collecting dust.
It wasn’t even turned on for the last 4 months and I wasted money on upgraded PS Plus subscription which now run out.

 

[CAT-THE-FIFTH]

What's really funny is that the RTX4070TI is barely faster at qHD and 4K than an RX6900XT and OcUK had these for around £600ish.

 

[z10m]

What's really funny is that the RTX4070TI is barely faster at qHD and 4K than an RX6900XT and OcUK had these for around £600ish.

It’s not funny it’s sad.
Anyway who cares what it’s called.? What matters is performance per $ and that as we all know is ****

 

[RavenXXX2]

What's really funny is that the RTX4070TI is barely faster at qHD and 4K than an RX6900XT and OcUK had these for around £600ish.

And not vram limited.

 

[Alt0153]

And not vram limited.

is 12gb enough?

 

[RavenXXX2]

is 12gb enough?

Barely enough, on the limit at 4k.

The memory expansion, on the other hand, is less impressive. 12 GB is 50 percent more VRAM than the GeForce RTX 3070 Ti, but that doesn't change the fact that the Radeon RX 7900 XT offers clearly more with 20 GB and 12 GB no longer consistently enable optimal operation today and in the future will be too little from time to time. For example, if you activate or switch DLSS in Spider-Man Remastered in Ultra HD including ray tracing afterwards, you will immediately run into a VRAM shortage. 12 GB in the current generation is probably pretty much what 8 GB was in the old one: mostly sufficient, but sometimes too little after 2 years at the latest.

 

[CAT-THE-FIFTH]

It’s not funny it’s sad.
Anyway who cares what it’s called.? What matters is performance per $ and that as we all know is ****

And not vram limited.

I find it funny as this generation is pathetic.

I did some quick calculations on TPU,and the RTX4070TI is significantly worse price/performance than my RTX3060TI.

So a tinier chip,less memory chips and less memory bandwidth and between 2.15X to 2.45X the price assuming an £800~£900 street price.

But the RTX4050 laptops with a tiny AD107 are going to start at $1000.

Nvidia has become worse than Apple.

 

[muon]

Anything below £900 will sell out tomorrow is my guess.

 

[azazl187]

Meh my PS5 is sitting in the corner collecting dust.
It wasn’t even turned on for the last 4 months and I wasted money on upgraded PS Plus subscription which now run out.

I think you maybe an exception to the rule there. I have a high spec pc and although i only use it from time to time to play gt7 I think it's a brilliant bit of kit for the price you pay. I think the value for money argument cannot be topped when it comes to the ps5

 

[Hotwired]

What's really funny is that the RTX4070TI is barely faster at qHD and 4K than an RX6900XT and OcUK had these for around £600ish.

I've been waiting to see if the usual suspects try to say DLSS and RT makes it worth that few hundred more. Not heard a squeak yet.

 

[Poneros]

8 GB was a real problem, but 12 GB is more than fine. This is one of the weaker arguments against the 4070 Ti imo. I have seen no real data to indicate otherwise, particularly when including DLSS (aka settings for actually playable framerates).

 

[K1LaW8tt5]

Barely enough, on the limit at 4k.

But 10gb was enough

 

[CAT-THE-FIFTH]

I've been waiting to see if the usual suspects try to say DLSS and RT makes it worth that few hundred more. Not heard a squeak yet.

I think they already did.

 

[Woodsta888]

Linus nailed it;

Nvidia will p*** you off 50% of the time.