• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Discussion Nvidia Blackwell in Q1-2025

Page 75 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
jpiniero said:
Those numbers don't make any sense.

The 5090 has 33% more SMs.The 5080 has 5%, 5070 Ti 6%, and 5070 4% (over 4070 NS). You'd expect the gap between the 4090 and 5090 to be 25-30% more than the other ones.
Click to expand...

Blackwell uses the same 4N process as Ada, and it looks like the RTX 5090 is clocked really low.
 
ToTTenTranz said:
Blackwell uses the same 4N process as Ada, and it looks like the RTX 5090 is clocked really low.
Click to expand...
Blackwell uses what they are calling 4NP. The most dense silicon of the 4/5nm TSMC silicon. It gives maybe 1-2% efficiency boost and about even on performance uplift over ADA. Based on AMD and Intel silicon, Nvidia is still using the most advanced silicon with Blackwell. The highlight for Nvidia is the 30gbps GDDR7 memory.
 
Hans Gruber said:
Blackwell uses what they are calling 4NP. The most dense silicon of the 4/5nm TSMC silicon. It gives maybe 1-2% efficiency boost and about even on performance uplift over ADA. Based on AMD and Intel silicon, Nvidia is still using the most advanced silicon with Blackwell. The highlight for Nvidia is the 30gbps GDDR7 memory.
Click to expand...
*Scoffs in arrow lake*

ARL: am I a joke to you??
 
Hans Gruber said:
Blackwell uses what they are calling 4NP.
Click to expand...
ITS N4P AHHHHHHHHH
AMD USES THE EXACT SAME THING
STOP FALLING FOR MARKETING
Hans Gruber said:
The most dense silicon of the 4/5nm TSMC silicon. It gives maybe 1-2% efficiency boost and about even on performance uplift over ADA.
Click to expand...
Check the diff from N5P/N4 to N4P, it is the same.
Hans Gruber said:
Based on AMD and Intel silicon, Nvidia is still using the most advanced silicon with Blackwell.
Click to expand...
Wrong, same stuff, different DTCO.
Hans Gruber said:
The highlight for Nvidia is the 30gbps GDDR7 memory.
Click to expand...
And the perf bump seems to be under the bandwidth bump, which is underwhelming.
Of course offline renderers and GEMM spammers will get the most of it.
I think the reason the part with the biggest spec bump is being reviewed first and the part with the least is reviewed on launch day should be obvious.
 
jpiniero said:
Those numbers don't make any sense.

The 5090 has 33% more SMs.The 5080 has 5%, 5070 Ti 6%, and 5070 4% (over 4070 NS). You'd expect the gap between the 4090 and 5090 to be 25-30% more than the other ones.
Click to expand...
If the TPU preliminary spec differences are accurate, it's a combination of factors.

Also the pixel throughput increase is almost nothing on 5090 while 5080 gets quite a boost over the predecessor. The 5070 Ti gets an enormous amount of pixel throughput boost. The outlier seems to be 5070 over 4070.

But your assumption is that the 4090 has the same bottleneck as the 4080 and 4070. When you are cutting things down you don't get to cut down things that make it optimal in performance for the low end part.

Things are also going to change depending on the workload. Is it 1080p vs 1440p vs 4K? Higher memory bandwidth benefits are greater on 4K resolutions than on lower ones for example. Does it use Ray Tracing? What about DLAA and DLSS? The gains for 5090 is greater than 5080 when DLSS MFG is enabled.

Prescott on the high end was mediocre over Northwood. However, Prescott's architectural improvements such as the coalesced write buffers improved performance greatly for the cache starved 128KB L2 Celeron parts. It was actually reviewed decently, partly because the Northwood successor was so sucky. It was 25-30% faster per clock than Northwood. Oftentimes the strive to push on the high end benefits low end chips disproportionately - one can say Moore's Law is about benefitting the common man.
branch_suggestion said:
And the perf bump seems to be under the bandwidth bump, which is underwhelming.
Click to expand...
Which is logical if Moore's Law scaling stops, while memory continues to improve. If you have two identical chips and you have an option to use next gen memory standard you'll use it.
 
Last edited:
xpea said:

DF Direct Weekly #196 - CES 2025 Special! - Nvidia, AMD, Intel Highlights + More!​



Nice sections on Mega geometry, ReSTIR PT vs Lumen GI and the insane Black State demo. Also a second view on AMD CES PT demo.
Click to expand...
1736827551576.png

This is impressive. Honestly the best part of new cards is not the 50 series but the DLSS4 features.
 
poke01 said:
View attachment 114875

This is impressive. Honestly the best part of new cards is not the 50 series but the DLSS4 features.
Click to expand...

-Nvidia marketing is really something else.

"Remember that thing we told you was awesome and you absolutely had to buy our cards to use it? It's trash. Garbage. You should be ashamed of yourself for liking it.

So anyway, here is the new version you can't live without..."

Takes me back to Turing and how Ray Tracing was here, photorealistic lighting. The end of rendering.

Now we're 3 gens in and RT was garbage and you need Ray Reconstruction but what you really want is Path Tracing and...
 
GodisanAtheist said:
-Nvidia marketing is really something else.

"Remember that thing we told you was awesome and you absolutely had to buy our cards to use it? It's trash. Garbage. You should be ashamed of yourself for liking it.

So anyway, here is the new version you can't live without..."

Takes me back to Turing and how Ray Tracing was here, photorealistic lighting. The end of rendering.

Now we're 3 gens in and RT was garbage and you need Ray Reconstruction but what you really want is Path Tracing and...
Click to expand...
bit dramatic, these are companies their job is sell products/services. I don't care about whatever bull NV says. I see an improvement, I make a note of it and this improvement will help users with older cards
 
GodisanAtheist said:
-Nvidia marketing is really something else.

"Remember that thing we told you was awesome and you absolutely had to buy our cards to use it? It's trash. Garbage. You should be ashamed of yourself for liking it.

So anyway, here is the new version you can't live without..."

Takes me back to Turing and how Ray Tracing was here, photorealistic lighting. The end of rendering.

Now we're 3 gens in and RT was garbage and you need Ray Reconstruction but what you really want is Path Tracing and...
Click to expand...

Ray Reconstruction is software that works on all RTX cards.

Basically a free update for everyone, that improves things. How Dare They???

I agree that FG/MFG gating is sketchy, and FG/MFG marketing is beyond sketchy, but not everything NVidia does is automatically sketchy.
 
Heartbreaker said:
Ray Reconstruction is software that works on all RTX cards.

Basically a free update for everyone, that improves things. How Dare They???

I agree that FG/MFG gating is sketchy, and FG/MFG marketing is beyond sketchy, but not everything NVidia does is automatically sketchy.
Click to expand...

-Ray Reconstruction sort of feels like it's giving up the gig on Ray Tracing. Don't ray trace, just sort of do the semblance of Ray Tracing and then use deep learning to fill in the gaps and get a final result.

RT was sold as "no cheap trick lighting" and it's all just going back to cheap tricks..

Not a clean bill of sale.

I have no personal beef with the tech, I mentioned a few pages back that we'll ultimately just end up with a "AI-Tracing" where there won't even have to be rays cast, AI will just insert shadows and lighting etc where it thinks they should be based on the image presented to the viewer.
 
The problem with RR was that it introduced ghosting and a oily look when it was turned on. DLSS4 RR reduces this with its new AI model.
GodisanAtheist said:
Ray Reconstruction sort of feels like it's giving up the gig on Ray Tracing.
Click to expand...
Full RT is hard on current hardware and RR "helps" to reduce the performance impact.
Good thing its optional because it still needs improvement.
 
GodisanAtheist said:
-Ray Reconstruction sort of feels like it's giving up the gig on Ray Tracing. Don't ray trace, just sort of do the semblance of Ray Tracing and then use deep learning to fill in the gaps and get a final result.

RT was sold as "no cheap trick lighting" and it's all just going back to cheap tricks..
Click to expand...
Ray Tracing inherently adds noise to the frame because of how light rays traverse and end up. It is just one of those things that is accepted as a challenge for an overall more realistic lighting if the scene is set for it. From the days of pixelated graphics to anti-aliasing and texture filtering, and limitations likewise in video capture, every new technique brought a step change even if it created another set of problems to solve.
Nvidia relies on well trained ML models for denoising that work alongside advanced RT. It's by no means giving up or limiting RT in this case. Quite the opposite.
 
adroc_thurston said:
Pretty sure kimi already said 128M L2 for the full die.
Click to expand...
Big increase in all resources and they could not get 50%+ proper non-fake perf out of it ...

Seen report (rumour) that Rubin is pushed into Q3-Q4 2025 on N3, clearly Blackwell is not doing well on N4 and needs some kick up the bottom, maybe even design was not optimal.

Perhaps this whole 50 series will be shortlived...
 
GodisanAtheist said:
-Nvidia marketing is really something else.
Click to expand...
It's straight up lying nowadays, accepted by the public as being ok. Intel claiming Arrowlake mobile beats competition, secretly hiding that it takes more than twice the power to do so, while Nvidia is pretending copying frames is more performance.

The single most important reason for more frames is for increased responsiveness. If you don't have it, then you might as well not have it.
 
DavidC1 said:
It's straight up lying nowadays, accepted by the public as being ok. Intel claiming Arrowlake mobile beats competition, secretly hiding that it takes more than twice the power to do so, while Nvidia is pretending copying frames is more performance.

The single most important reason for more frames is for increased responsiveness. If you don't have it, then you might as well not have it.
Click to expand...
Everyone does it these days, remember the days when keynotes didn’t include fudged up graphs or fake frames or ballooning power consumption to cover up performance deficits.
 
poke01 said:
Everyone does it these days, remember the days when keynotes didn’t include fudged up graphs or fake frames or ballooning power consumption to cover up performance deficits.
Click to expand...
If anti-trust agencies think they are actually doing something good for the society, this is what they should fix. Nvidia has so much mindshare that lying through the teeth is acceptable.

Forget the whole "too big to fail" deal. Short term pain, long term gain. Stealing without using the literal word.
 
DavidC1 said:
The single most important reason for more frames is for increased responsiveness
Click to expand...
Eh, smoother picture is smoother picture. The problem is NV selling interpolation as a more perf technique (it's not).
poke01 said:
remember the days when keynotes didn’t include fudged up graphs or fake frames or ballooning power consumption to cover up performance deficits.
Click to expand...
Well it's a byproduct of Blackwell client being really weak.
They're yet to talk any concrete SM changes.
 
DavidC1 said:
If anti-trust agencies think they are actually doing something good for the society, this is what they should fix. Nvidia has so much mindshare that lying through the teeth is acceptable.

Forget the whole "too big to fail" deal. Short term pain, long term gain. Stealing without using the literal word.
Click to expand...
Yes I do think companies should be fined large amounts for deceptive marketing.
Just look at the M4 Pro vs Strix halo marketing from AMD and YouTubers influencers milking this with clickbait titles.

IMG_1480.jpeg

No mention of power consumption and talks of leadership when the comparison is only done against M4 pro and not the M4 Max. I bet the M4 Max uses less power than the Strix Halo and is still more powerful in 3D work.

Of course, there’s Nvidia with its fake frames BS and 5070 is as powerful as 4090 fake marketing.
 
poke01 said:
No mention of power consumption and talks of leadership when the comparison is only done against M4 pro and not the M4 Max.
Click to expand...
They've already said it's 48-50 vs 55W.
Which is funny given M4's have a sizeable nuff node advantage.
poke01 said:
I bet the M4 Max uses less power than the Strix Halo and is still more powerful in 3D work.
Click to expand...
Pipe down champ, Max has twice the membw.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top