Sunday, June 2nd 2024
NVIDIA Unveils New SFF-friendly Enthusiast GeForce RTX Graphics Card Standard
NVIDIA has taken note of the upward-and-outward trend for enthusiast-segment GeForce RTX gaming graphics cards to be huge, and out of sight for small form-factor gaming PC builds. This can be a problem, as SFF gaming PC builds are confined to exotic liquid cooling solutions that drive up costs, or to make do with mid-performance graphics cards, or simply give up the idea in favor of a next generation console, such as the upcoming ones from Xbox and PlayStation. To confront this trend, NVIDIA developed the new GeForce RTX SFF-Ready Enthusiast Graphics Card Guideline.
Put simply, the guideline calls for SFF-Ready advertised enthusiast segment graphics cards to be no larger than the specified dimensions. These are a maximum of 30.4 cm in length, a maximum of 15.1 cm in height, and a maximum of 5 cm or 2.5 slots in card thickness. Only cards at or smaller than these dimensions (304 mm x 151 mm x 50 mm) quality for the SFF-ready marker. It's also important to note that board partners cannot work their way around this by using AIO liquid cooling solutions where there's more to the card than its main component (i.e. tubing and a radiator). The total physical dimensions of the card cannot exceed the ones specified above. In addition to this, NVIDIA is now going to maintain a continuously updating list of graphics cards and cases that meet the SFF-Ready dimensions.
Put simply, the guideline calls for SFF-Ready advertised enthusiast segment graphics cards to be no larger than the specified dimensions. These are a maximum of 30.4 cm in length, a maximum of 15.1 cm in height, and a maximum of 5 cm or 2.5 slots in card thickness. Only cards at or smaller than these dimensions (304 mm x 151 mm x 50 mm) quality for the SFF-ready marker. It's also important to note that board partners cannot work their way around this by using AIO liquid cooling solutions where there's more to the card than its main component (i.e. tubing and a radiator). The total physical dimensions of the card cannot exceed the ones specified above. In addition to this, NVIDIA is now going to maintain a continuously updating list of graphics cards and cases that meet the SFF-Ready dimensions.
50 Comments on NVIDIA Unveils New SFF-friendly Enthusiast GeForce RTX Graphics Card Standard
Also, that’s a rather weird case list. No Jonsbo cases? Or ssupd? Ridge and Terra, but no Torrent Nano? Did they just use a dart board?
Top card shown below fits NVidia's new SFF guidelines, with the 4060 LP under it. And an ITX sized card pictured for good measure.
Wouldn't mind seeing RTX 4070 Lite introduced (with these Ti and Supers flying around, it's not gonna hurt anyway), a GPU of the vanilla 4070's SM and VRAM count (or perhaps 44 SM instead of 46 just for lulz) but locked to, say, 120 W, and built with the smallest possible cooling solution that's still reasonable for this heat output.
I miss those ITX dual slot cards. No idea why we stopped making them, there's plenty of GPUs that fit the power bill.
I also love how the defined dimensions were high end cards like 2 generations ago.
I remember when high end cards were single slot and only drew 40 watts. All these modern cards are power hogs!
THEY are (primarily) the ones who started making gigantic cards in the 1st place, so what exactly is the point of this statement anyway ???
Ngreediya/Jacket Man are such douchebags will someone please just tell them to stfu & gtfooh already.....:D
Like....you guys have ATX cases. They have 7 slots. You dont use sound cards anymore, you dont use separate network cards, Controller cards? LMFAO no. 99% of "gaming PCs" are full of empty space.
Do people really want 72dba blowers back? Just to safeguard the empty space in their PC? The mini ITX guys I understand, because "mini ITX" used to mean sub 10L PCs, but have bloated to hold these big GPUs. But then, go buy that 2 slot jet engine 4080 and be happy? I seem to remember this exact same back and foth during the Thermi days, with people insisting the 480 was too much and 300w GPUs were way too power hungry and dual slot coolers were obnoxiously big and everything used to be single slot and blah blah blah.
I have to sit back and reflect on how PCs have changed, and I mean the acronym., Personal Computers. As someone that went from EATX to mATX I look at my machine now and I will be honest I am biased. To me though this is what engineering and advancement looks like.
To nvidia its apparently 5 slot cards. How big does the rest of the system need to get to accommodate that? History repeats itself in almost all cases; but I am hoping we dont get to the size of refrigerators before we shrink agian.
pun not intentional
I'm guessing they are showing this standard off because nobody can agree on what a "SFF" GPU is . So now they have a standard. So case makers can say their case is Nvidia SFF GPU compatible" and people can buy a GPU knowing it will fit without flipping through dimensions on every card. Well, yes, because previous cards were held back by manufacturing nodes that could not scale clocks and die size high enough to necessitate a 3 slot cooler. Its not like GPUs like the GTX 280 left tons of room on the table for improvement. Just like GPUs such as the 9800 pro didnt need a dual slot cooler.
Now our limit is not the clock speed but rather our ability to extract heat from the ever shrinking transistors. Which necessitates a bigger cooler to extract that performance without overheating or sounding like a jet engine. As a side benefit, our cards are quieter then ever and cooler then ever. I remember what a big deal the first Twin Frozr cards were back in the fermi era. The fact you could cool a 250w GPU without a noisy AF squirrel cage fan was a huge improvement.
IDK why the existence of the 4090 convinces people that there is no innovation, the 4080 super has a 2 slot variant for purchase. Anyone who wants a 2 slot card can buy that. In fact, there is a dual slot 4090 in china with a reduced 260 watt TDP. It's the PERFECT card for living out the dreams of 2011 builds. Curiously, people are not scrambling to buy it, even through its a 260w dual slot 4090 GPU.....
RTX A2000 and RTX 4000 Ada exist. It's neither difficult physically nor efficiency-wise, all the companies are just lazy and greedy. Not everyone's strictly looking for 4080 level performance, cards absolutely do not need to be triple slot 300mm (looking at all you offensively large 4060s and 4070s).
Which would just hobble higher clocked versions of top end cards (ie. XTX, 4090). Going flow thru like Nvidia doesn't leave much room for an OC oriented PCB. And anything below say 4070 Ti has no right to be even close to that big in the first place.
So essentially Nvidia put in zero effort into this announcement, stated the obvious, and passed it off as the next best invention. Then what is the point of "SFF" branding if all larger 15-20L cases in the past 5 years could accommodate these regular sized cards to begin with?
So the point is that SFF is massively popular now and no longer to be dismissed as niche......but only if you are still building 10L+. Otherwise you're still an oddity. Got it. Remind me what was the point of this branding again?
Not to mention that their insistence on 12VHPWR means nothing was gained. Yes, their spec dictates a lower height card that isn't taller than the bracket. But most of those cards used 8-pin. Now with FE we have a lower height card, but you still have to treat it like a taller card for 12VHPWR clearance.
To make full use of the lower height, you'd have to buy an extra cablemod angled adapter. Which itself was fully recalled for some failures just a few months ago.
And yet in desktop DIY PCs we’ve been essentially stuck on the same-ish tower form factor for a long while now. You’d think that what we call SFF would not be this exotic niche for enthusiasts by now, but the norm. And yet it’s not. And I absolutely think that GPU sizes increasing drastically over the last decade is part of the problem.
Lel
This cannot continue without failing, it just can't.
Is 15 cm and one millimeter enough for height, or they are not take under consideration the curvature needed for their 16 pin connector?
By the way, not all atx cases can house their 16 pin connector cards without a vertical mounting and riser cable, let alone sff cases.
On the other hand, how cooler are those cards in those "sff" cases, the advertised cards and cases.
Ngridia- The Way Is Meant To Be Paid.
The PC wants to do it locally. Every other device that got shrunk over time, is aimed at doing things in the cloud. Even laptops; they're not equipped with sufficient storage to carry everything on local disk. To 'complete' a workflow on a laptop, you're generally looking at some external influence, whether peripherals, data, whatever.
So yeah things get smaller because they do less. But we want our PC's to keep doing more. Locally. Directly.
GPU sizes have increased over the last decade because we are no longer limited by production node. A GTX 780ti could only overclock so far. Newer nodes can clock higher, and handle higher power use, the major issue today is not clock speeds, it's thermal output, and thats why GPU coolers have gotten big, to handle the heat and also to run quieter then ever before. Even small cases today like the dan A4 can handle long triple slot GPUs. So...why not make em that way? The only market this doesnt really benefit is the LP community, but considering we got the 4060LP this year, even we're doing great! The point of this branding is to appeal to the 95% of the market that still use micro of full ATX chassis, or ITX chassis that are 10+ liters and have length or width restrictions.
I'd LOVE for more proper LP support, or proper 7.67" ITX card support. The R9 285 was a great ITX GPU, but with the 3000 series we just stopped getting them. Such a cooler could easily handle a 4060ti, maybe even a 4070, but nobody wants to build one. Maybe just nobody bought them?
At the least, we got the 4060LP, which has been great for LP builds. And there is the ada 4000, if you want to spend jalopy money on a GPU.
Today though? We are using a massively refined version of the tech that allowed the microprocessor to be a thing. It's refined enough to allow relatively powerful computers that fit in our pockets, but it's still the same technology. It's still silicon, and as years goes by, the cooling requirement just increases. We went from ambiant air cooling, to an heatsink, to active air cooling with bigger and bigger heatsink.
If we were willing to compromise, tiny desktop computers could be the norm, like something roughly equivalent to a Mac studio. Not a charts topper but powerfull enough to work and play within reason. But no, (especially when it comes to GPU) we want impressive graphics, high refresh rate, high resolution, all the while being quiet, and software workarounds are being shunned by many tech enthuisiast. This is the "PC gaming master race" not "compromise land". Make 1440p the high-end target, you'll see people say that GPU makers are lacking ambitions.
It's easy to say for a CPU that it's "fast enough" even if you limit it. The same can rarely be said about a GPU. Unless we figure out a novel way to make them...