NVIDIA Rumored to Start Production of GeForce RTX 30 Series SUPER GPUs for Desktops This Quarter

Tsing

The FPS Review
Staff member
Joined
May 6, 2019
Messages
11,180
Points
83
nvidia-geforce-rtx-3080-key-art-1024x576.jpg
Image: NVIDIA



It was recently reported that NVIDIA would be releasing SUPER-charged versions of its GeForce RTX 30 Series GPUs for the mobile market. According to prominent leaker Kopite7kimi, green team is actually planning to release new GeForce RTX 30 Series graphics cards for desktops with the SUPER branding as well. Described as part of an “Ampere Phase 2” initiative, NVIDIA’s new GeForce RTX 30 Series SUPER GPUs will smoothen the transition between its current Ampere generation with the next one, Lovelace, which is rumored to power the GeForce RTX 40 Series. While they won’t be leveraging any sort of node change, NVIDIA’s GeForce RTX 30 Series SUPER GPUs are said to feature wider bus interfaces and higher VRAM capacity for increased performance.



not only laptops https://t.co/eGSv8R89wP— kopite7kimi (@kopite7kimi)...
Click to expand...

Continue reading...


 
AH HAHAHAHAHAHAHA.. WAHAHAHAHAHAHAHAHAHA....

F U Nvidia.

Anyone here expect to ever see a NON super NON TI new video card come out after a few months? It'll all be super or nothing. And I'm betting a 2-400 dollar markup.
 
Grimlakin said:
Anyone here expect to ever see a NON super NON TI new video card come out after a few months? It'll all be super or nothing.
Click to expand...
The 'Super' lines recently have been a way to take advantage of better yields without lowering MSRPs. They may be priced higher than their non-Super predecessors, but they're usually also faster.

In the end it's more of a wash for price vs. performance in a lot of ways.

Grimlakin said:
And I'm betting a 2-400 dollar markup.
Click to expand...
Not that MSRPs are indicative of much of anything these days ;)
 
No, it's a sad and ridiculous time for PC hardware. Period.

The situation with other hardware is just getting worse and worse regardless of whether they say "things are improving". Improving for WHO? Not for enthusiasts or regular consumers, that is for GD sure.

Why in the world would you release ADDITIONAL product lines when you already cannot get the product lines you already have into the hands of the people they were meant for? F all these companies.
 
Darkbreeze said:
Why in the world would you release ADDITIONAL product lines when you already cannot get the product lines you already have into the hands of the people they were meant for? F all these companies.
Click to expand...
Because that's not at all how manufacturing works?
 
I have two minds about this:

First, I agree with @Darkbreeze, it is very insensitive to announce things in this manner. The problem now isn't that we need faster GPUs, it's that there are no GPUs to buy, and it isn't apparent as to how this addresses the issue.

On the other hand - more GPU SKUs shouldn't hurt availability, and ~should~ do nothing but help improve it - even if only slightly.
 
This is probably due to manufacturing lessons being learned. Since they can make the super chips now actually more efficiently than the old that's what they do.
 
Grimlakin said:
This is probably due to manufacturing lessons being learned. Since they can make the super chips now actually more efficiently than the old that's what they do.
Click to expand...
The 'Super' chips have been the same chips - the Super SKUs are usually either the same dies as before but with more cores enabled or the next size up die, perhaps with cores disabled versus the SKU that previously featured it.

Granted that's what they've done so far. They could always change it up, but the overall trend has been to take better advantage of yields rather than entire new dies being released.
 
LazyGamer said:
The 'Super' chips have been the same chips - the Super SKUs are usually either the same dies as before but with more cores enabled or the next size up die, perhaps with cores disabled versus the SKU that previously featured it.

Granted that's what they've done so far. They could always change it up, but the overall trend has been to take better advantage of yields rather than entire new dies being released.
Click to expand...

Ok so it's better yields as opposed to new dies.... pretty much the same effect in the end. It becomes more cost/risk efficient to make the better chips so they make those and basically wait for bad batches for the older ones or contract enforcement only to make and sell them.
 
Grimlakin said:
make the better chips
Click to expand...
It's generally the same chips in different cards, i.e., a card that had a chip with cores disabled may get a 'Super' version with everything enabled, and a card that started with a fully-enabled chip may get the next largest chip as its 'Super' version, perhaps with some cores disabled.

Sometimes there are new steppings so you get an 'A' after the 'GT104' (for example) chip, but these stepping are usually limited to numbers of cores enabled, higher clockspeeds, and perhaps lower voltages (or higher voltage tolerances for overclocking).

Main takeaway is that Nvidia isn't putting out completely new chips to make the 'Super' cards (and generally don't do that for Ti cards either). These suffix'd SKUs are marketing terms to allow for retail SKU flexibility to address competition and yields as necessary.


And I'll say that that's not necessarily a bad thing, which has been my point. The 2000-series Super release, for example, brought higher performance to lower MSRPs. Nvidia used a quick marketing maneuver to essentially provide a better value for many market segments, at least in terms of MSRP.
 
You must log in or register to reply here.
Become a Patron!
Back
Top