Concept SENTRY 3.0: Development and Suggestions
- Thread starter ZombiPL
- Start date
steelfractal
when in doubt slap a delta on it
Silver Supporter
Bronze Supporter
M...M...M...M...Multi-Tier...Subscriber...
- Aug 15, 2019
- 423
- 531
steelfractal
when in doubt slap a delta on it
Silver Supporter
Bronze Supporter
M...M...M...M...Multi-Tier...Subscriber...
- Aug 15, 2019
- 423
- 531
It's just barely 3.5 years, I think, since we've sold last 2.0 batch post-campaign, not 6 years... And in the meantime we had covid pandemic.
We're working on the design step by step when we can, but it's still not a priority project for now.
And yet, you still have the rubber o-ring outside on the power button and not on the inside...Dr Zaber Sentry.. Fan waiting here patiently()
3 slot GPU support and 47mm air coolers
2.5 slot GPU with standard 2U risers and proper support for 48mm air coolers is what we're aiming for. If we'd want to support thicker cards, then it means longer riser or lower cpu cooler and (in both cases) no 2.5" drive support.
On top of that 3 slot cards would be so heavy that relying on just slot and bracket connection to handle the cards isn't something we can accept if we're promoting the case as something portable. Take a look at the problem of Gigabyte cards cracking around the pci-e lock - if something like this happens for someone while travelling with our case, he could argue that it's our fault that we're supporting such cards without the hold on the end of the card, and of course making a supporting piece that supports all shapes and thicknesses of the card would be both cumbersome and expensive to make.
Could you consider making the cpu cooler clearance a Little bit higher? Having 3mm from the blades to the case helps with turbulence.
For the gpu: it’s a good point I haven’t considered so I’ll trust you with that.
For the gpu: it’s a good point I haven’t considered so I’ll trust you with that.
We're aiming to support properly Cryorig C7, regardless of it's shortcoming, so it's a 47mm cooler that had 1mm of space to perforation. We will probably have +1.5~1.75 mm more so it'll be 2.5~2.75mm now which should be better.
Black Ridge loses out to AXP90-X47 Copper on AM5 tmk, and also doesn't affect RAM compatability. The BR is compatible with ASUS B650E-I, but you'll need to removed M.2 heatsink and use aftermarket one. They've just released rev.D with AM5 support. I'd certainly be curious to see the two compared with official brackets since I've only used DIY modified versions.
So, what do you guys think about Nvidia's new "SFF-Ready" scheme? Will you be designing with that in mind?
It's larger than I'd like, but it'd sure be easier to get more people onboard with small form factor computers if all they had to do was look for a little badge, rather than comparing dimensions and hoping the numbers are accurate.
It's larger than I'd like, but it'd sure be easier to get more people onboard with small form factor computers if all they had to do was look for a little badge, rather than comparing dimensions and hoping the numbers are accurate.
It's a missed opportunity to set things straight. The whole problem of not informing end user what the actual measurements are with a proper product spec sheet is missed, so still if you have a case that is not exactly those dimensions and thus certified, you will still need to ask someone who has the card physically to check the measurements for it.
We should be angry about this because it's nvidia treating the SFF community as idiots by forming the communication of this SFF-ready as idiotproofing for newbies that can't figure things out for themselves while even I who made the whole compatibility spreadsheets of cards can't be 100% sure until I have a specific card in my hands.
But there is a reason for not providing the spec sheet with exact dimensions - kind of like security by obscurity - if you get a drawing with dimensions and you get a card that doesn't match the measurements on the drawing, in some countries you can return the card including the seller having to pay you back for the shipping because product wasn't as advertised. If they don't provide this, along other details in other hardware, like for example exact memory chips in the card, in SSDs, in RAM sticks etc, they can switch suppliers, change the cooler size on the card and so on, and it won't have effect on returns because of mismatched spec if they don't provide that.
We would be fitting perfectly the cards within 304mm limit, except for the stupid new Nvidia Fire-Hazard [TM] connector and the space required for it.
Overall whole situation with how the cables are handled in GPUs is simply stupid, especially with the trend for hiding the cables, where Quadro cards have had for ages the power connectors at the far edge from the pci-e slot which makes it so that it fits slim HP/DELL workstations. Nvidia having the amount of control they have over the board partners could simply force everyone to comply with one setup of the power connector at that end, and then it would mean everyone is designing cases around similar requirements. And also they could have, I don't know, maybe say that SFF-Ready cards require spec sheet with dimensions marked on a drawing of the card?
Well, for one, the cards with 8-pin power connector can have those dimensions and still use some low profile connector, or simply be a pci-e spec reference sized card, so they simply can plug 8-pin and 6-pin connectors without problem into the card from the side. Also this means that cards above RTX 4070 are not going to fit into Sentry 2.0 because of that plug, but there are RTX 4070 cards which are a perfect fit for Sentry 2.0, exactly as we designed with the proper TDP for such case.
Also I think we shouldn't be conforming to to a bad standard like this 12(+4) pin connector that nvidia is forcing because they want to squeeze more power draw into the card because they have to have constant performance grow to make investors throw money at their buble, while also wanting the card to look clean when showed on the stage, while also not caring about the tri or quad headed hydra dongle sagging off the card in our cases.
So the GPU compartment will actually be the same size with same compatibility as per drawing on our website for the 2.0. Also we'll have to see if we're going to be officially supporting 2.5 slot cards because again, if we don't have proper dimensions shown, just saying the card is 50mm thick doesn't mean it's within 2.5 slot card outline. Technically there can be a 50mm card that is sticking out 5 mm outside of 2.5 slot outline if there's nothing sticking out at the back of PCB.
I doubt any other vendor will make a non-FE card that will not be huge for this wattage. And the FE card is whole inch taller than the pci-e reference card size.
It's interesting how the PCB of the card is in the middle making it so that there's pass-through in the front as well. I wonder how the pci-e connector PCB will be connected to that card PCB in the middle.
Hi there, I have a sentry 2.0. I am planning on building with 5080. Your manual says the highest length of the GPU is 305 mm. I saw one GPU with 306 mm and 50 mm thick. Can I cram that inside the Sentry 2.0? I couldn't find any info on the thickness allowance in the manual.
I agree, I find fascinating that they've managed to cram that wattage in that space with that 3D Vapor Chamber (it seems they are calling it that) and that PCB design must be quite pricey to manufacture. Looking forward to see a teardown (and hopefully a Gamers Nexus video with the NVIDIA thermal engineer with whom he already teardown the 3090 and 4090 FE)
Considering the 4090 FE was already the smallest air cooled 4090, and the tiny margin NVIDIA leaves these days to AIBs, I'm essentially sure no AIB will go up to the challenge.
If they did, probably most users looking for specifically a small one, would try to get a Founders, although as Founders models are generally popular, are frequently out of stock and that might be an opportunity, an AIB challenger would probably need to be considerably more expensive, and it is already enough expensive as it.
Not saying you should do a Sentry 3.0 now, but, it sure looks as an improving landscape for it.
I don't think NVIDIA would go through all this trouble just to make a one-off., flagship GPUs exploded in size due to the big increase in power (the why of these new heights in power draw is another thing), and that's why they made the new (finicky) power connector in the first place, in these 4 years they ended up releasing GPUs that can use all the new power budget the new connector allowed, (I mean one simply needs to move a slider to allow the 4090 FE to use 600W) so, given that I don't think they are going to release a double 12V2x6/12VHPWR connector GPU any time soon, and that they are showing they can air cool a thermal load of 600W in two slots, it seems this is they are going with this kind of format for the forseeable future on the FE models.
Curiosly enough back in june 2024, NVIDIA decided to launch the SFF-Ready program, in which they defined the category "SFF-Ready Enthusiast GeForce Cards"
with this size limitations:
In hindsight seems like they were preparing themselves.
Again, seems too much effort for just one-off of reasonable sized GPUs
Considering the 4090 FE was already the smallest air cooled 4090, and the tiny margin NVIDIA leaves these days to AIBs, I'm essentially sure no AIB will go up to the challenge.
If they did, probably most users looking for specifically a small one, would try to get a Founders, although as Founders models are generally popular, are frequently out of stock and that might be an opportunity, an AIB challenger would probably need to be considerably more expensive, and it is already enough expensive as it.
Not saying you should do a Sentry 3.0 now, but, it sure looks as an improving landscape for it.
I don't think NVIDIA would go through all this trouble just to make a one-off., flagship GPUs exploded in size due to the big increase in power (the why of these new heights in power draw is another thing), and that's why they made the new (finicky) power connector in the first place, in these 4 years they ended up releasing GPUs that can use all the new power budget the new connector allowed, (I mean one simply needs to move a slider to allow the 4090 FE to use 600W) so, given that I don't think they are going to release a double 12V2x6/12VHPWR connector GPU any time soon, and that they are showing they can air cool a thermal load of 600W in two slots, it seems this is they are going with this kind of format for the forseeable future on the FE models.
Curiosly enough back in june 2024, NVIDIA decided to launch the SFF-Ready program, in which they defined the category "SFF-Ready Enthusiast GeForce Cards"
with this size limitations:
- 151 mm maximum height including power cable bend radius
- 304 mm maximum length
- 50 mm or 2.5 slots maximum depth
In hindsight seems like they were preparing themselves.
Again, seems too much effort for just one-off of reasonable sized GPUs
I know it was mentioned before, but I said it like that in case someone happen to read just the last messages.
I agree with was pointed before, clearly has its problems, clearly they were planning ahead, they knew the hazard the new connector is in a constrained space, the new position of the power connector maybe be the solution, and they may even have thought of that redesign back when they released the program, questionable not explaining in detail the matter of the connector on SFF-Ready, I wonder if with the 50 series, they've plan to enforce guidelines for the AIBs that want the sticker on their GPUs
Similar threads
