Prototype The world most flexible SFF case? LOUQE Ghost S1

Hifihedgehog

Editor-in-chief of SFFPC.review
May 3, 2016
459
408
www.sffpc.review
I think they should make some changes for compatibility reasons, so maybe wait for them to ship first batch, and then jump in V2.
What sort of changes do you propose? I am curious. All I know is my GTX 1080 Ti will fit easily. Any improvements I would greatly look forward to. I must admit, however, that I am now also considering the EG Model One given its price advantage. However, if a V2 of the Ghost S1 brought worthwhile improvements that are worth the added wait and the added price, I might consider it.
 

rommel671

Trash Compacter
Aug 2, 2018
48
11

Hifihedgehog

Editor-in-chief of SFFPC.review
May 3, 2016
459
408
www.sffpc.review
That would mean that the DAN Cases A4-SFX is not compatible either. This and many other sub-10L cases have only space for at most 2.25- and 2.5-slot cards. I strongly think the problem with the RTX 20 series is the TDPs are too high. For this reason as well as price, many people like myself are just settling on 10 series cards or waiting until 7-nm cards hit.
 
Last edited:
  • Like
Reactions: Bangle

Bangle

Cable-Tie Ninja
Apr 12, 2018
147
112
I don't have the precise numbers, but my math tells me it's more expensive for the end customers if the inventory is warehoused outside of Asia.

The price of the Ghost has factored in VAT since they are shipping from EU. If they warehoused in Asia, they can lower the product price. So the way I look at it, why increase the price for everyone else just to make it easier for EU customers.

That being said, they are a EU based company, there might be some legal and tax advantages to do so, but again, it's at an extra cost to customers outside of EU.

We were able to lower the cost for DAN A4 for customers outside of the EU, by warehousing in Asia. I suspect the same for Ghost.
Either way its an extra cost for someone. Its usually EU customers that get screwed over. Plus surely it won't be that more expensive?? I think all louque are doing ship it to the EU themselves, and will pay the duty based on the 'manufacturing' cost (say $50) rather than the sale cost ($200). It won't be warehoused for longer than a week as the items have already been sold through pre-order and the 'warehousing' is the courier's warehouse - which the shipping costs cover this.
 
  • Like
Reactions: Tabbes

SashaLag

SFF Lingo Aficionado
Jun 10, 2018
127
111
80m cooler means ~ 22mm wider, then the width will be 160mm. At 162x188x322mm, it's 9.8L without Top Hats, the same width as the M1, but more boxy, uglier proportions. If you add top hat to that for 240 AIO, 12.8L, almost same size as the M1, but with less cooling and component compatibility and more expensive - what's the point then. Don't forget, the PCIe riser increase potential problem too.

Effectively, the back to back layout sacrifices CPU cooling and water cooling to be narrow (120-140mm) and to reach a smaller volume. You deviate from that, the portions are not attractive, and size vs performance no longer make sense.
what's the point of making it bigger?

Now (140mm) it sits just in betwenn a Dan A4-SFX (112mm width) and a NCASE M1 (157mm)...
- You want a low profile CPU heatsink? You buy Dan one,
- You want a taller but not too tall heatsink? You "buy" a Louqe one,
- You want even taller? NCASE!
- You want AIO? Either NCASE M1 or Louqe...

Increasing overall dimensions, increase cost which is already high... And we already have a jack of all trades (NCASE M1)! Besides, now its pricier than M1 but cost can go down, using the same thickness of NCASE panel...
 

Bangle

Cable-Tie Ninja
Apr 12, 2018
147
112
That would mean that the DAN Cases A4-SFX is not compatible either. This and many other sub-10L cases have only space for at most 2.25- and 2.5-slot cards. I strongly think the problem with RTX 20 series is the TDPs are too high. For this reason as well as price, many people are just settling on 10 series cards or waiting until 7-nm cards hit.

for the price Nvida are asking most people that can spend that kind of money already have a 1080Ti. And many people doubt that the RTX2080 - which is slightly more expensive than the 1080Ti - and is within most 1080Ti owners budgets - will be more that 20% faster. Ray tracing will also probably be limited to 1080p. So alot of people will skip the RTX 2xxxx in my opininion - As most people with a 1080 or 1080Ti probably don't game at 1080p.

Judging from the rumors etc, Nvida are doing their hardest to supress benchmarking figures. Which is never a good sign.
 
  • Like
Reactions: Hifihedgehog

Wahaha360

a.k.a W360
SFFLAB
NCASE
SSUPD
Feb 23, 2015
2,131
10,697
what's the point of making it bigger?

Now (140mm) it sits just in betwenn a Dan A4-SFX (112mm width) and a NCASE M1 (157mm)...
- You want a low profile CPU heatsink? You buy Dan one,
- You want a taller but not too tall heatsink? You "buy" a Louqe one,
- You want even taller? NCASE!
- You want AIO? Either NCASE M1 or Louqe...
...

I understand your argument, the problem is the new CPUs + GPUs coming out, higher TDP and bigger size.

Intel is putting a 8C/16T i9-9900K mainstream (prob ~140W at load), Nvidia decided to release a dual fan Ref GPU at higher TDP, forcing board partners to make 3 slot GPUs, and many bigger in all dimensions compare to last gen.

The advantage of supporting a 58mm tall CPU cooler vs 48mm is marginal with 9th Gen Intel and Ryzen 2nd Gen. That's why the A4 V3 *had to support 120mm AIO. If 58mm CPU cooler was enough, A4 would just be wider at 140mm.

In order for Intel and Nvidia to compete with AMD, they both decided to forgo power efficiency in favor of performance. Hotter and bigger component is the trend now, it's a problem for all SFF cases.

So my point it, this middle ground between the A4 and M1 is shaky with the new hotter and bigger components coming out.

Increasing overall dimensions, increase cost which is already high... And we already have a jack of all trades (NCASE M1)! Besides, now its pricier than M1 but cost can go down, using the same thickness of NCASE panel...

Thinner panels doesn't make the product cheaper. For these manually made cases, the cost is not the material, it's the labor / process, and anodizing.

Many anodizing shops in Asia are closed due to new environmental regulation, the end result is that we have gotten quotes where anodizing is half of the cost of the case.
 
Last edited:
  • Like
Reactions: Beardedswede

dondan

Shrink Ray Wielder
DAN Cases
Feb 23, 2015
1,977
8,379
9900k will be easier to cool as an 8700k because same TDP but die are soldered.

Furthermore I don't see the prob with RTX 2080ti. The Founders card will fit.
 

Beardedswede

Cable-Tie Ninja
Jun 9, 2018
191
158
what's the point of making it bigger?

Now (140mm) it sits just in betwenn a Dan A4-SFX (112mm width) and a NCASE M1 (157mm)...
- You want a low profile CPU heatsink? You buy Dan one,
- You want a taller but not too tall heatsink? You "buy" a Louqe one,
- You want even taller? NCASE!
- You want AIO? Either NCASE M1 or Louqe...

Increasing overall dimensions, increase cost which is already high... And we already have a jack of all trades (NCASE M1)! Besides, now its pricier than M1 but cost can go down, using the same thickness of NCASE panel...

For me. I dont care If does the same as ncase m1. I still love the dual camber, the design, material choises. Modularity. Changeable top plate.I think they should do it multipel sizes. Like all big case manufakturers do. Of course this is just my wish. :)
Im sure there are issues with that like wahahah360 said. I’m just saying what I want :)
 
Last edited:

Nanook

King of Cable Management
May 23, 2016
805
793
9900k will be easier to cool as an 8700k because same TDP but die are soldered.

Furthermore I don't see the prob with RTX 2080ti. The Founders card will fit.
Do you think that 8c and 6c would have similar TDP in practice? I think that 8c would run hotter. I feel that the listed TDP are optimistic
 

Hifihedgehog

Editor-in-chief of SFFPC.review
May 3, 2016
459
408
www.sffpc.review
9900k will be easier to cool as an 8700k because same TDP but die are soldered.
I have never disagreed with you Dan, but I have to disagree with you on this. Technically, the 8700K has nearly the same official TDP as the 7700K but real-world tests show that the 8700K shows a significant power and temperature increase compared to its four-core predecessors. I can only imagine where an eight-core die with higher boost clocks on the same 14-nm process will take Intel processors to in terms of actual cooling and power requirements.
 
Last edited:

dondan

Shrink Ray Wielder
DAN Cases
Feb 23, 2015
1,977
8,379
@Nanook, @Hifihedgehog : Both of you are missing one aspect Intel TIM. The 8700k use TIM and the 9900k is soldered that makes the difffence. If both CPUs are soldered or using TIM (like 7700k vs 8700k) you are right but not with this important difference.

A little example:

Intel i7 5820K (6 cores 12 threads) 140W TDP SOLDERED
Intel i7 7700k (4 cores 12 threads) 95W DTP TIM

It is easier to cool the 140W 5820K than the 7700k. I tested it with an Thermalright AXP100 on both CPUs. In the old Prime95 version without AVX the 5820K reached 70-75°C while the 7700K reached 85-90°C.

So a soldered CPU works on the same level as dellided CPU. The advantage is 15-20°C.
The new 9900K will be a great step for SFF because it will bring us back to the Sandy Bridge age. I think a 9900K will be coolable with an L9i and will reach 80-90°C in Prime95 and 70-80 in games & software.

TDP:
The real TDP will depend on the board you will use. Modern boards allow the CPU to run the max turbo on all cores this results in a much higher TDP between 120-150W. You can check that with running Prime95 and read the TDP in CoreTemp. Only if you setup Max Power Duration to 95W you will run the CPU on default specs. For example it doesn't matter if you buy 8700 (65W) or 8700K (95W) on default configuration the max clock difference on all cores is only 100Mhz and also the TDP is the nearly the same. Theoretical it is "easy" to cool an 8700 with the L9i if you limit the "Max Power Duration" to 65W.
 
Last edited:

Hifihedgehog

Editor-in-chief of SFFPC.review
May 3, 2016
459
408
www.sffpc.review
I will agree to disagree since delidding only can do so much in such tight quarters. From my experience in perusing the charts at AnandTech, the Core i7-8700K at full load and stock settings consumes more power (around 15-20W more!) than a Core i7-6900K and Ryzen 7 2700X and is close to the 5820K you mention. Intel’s listed TDPs are a stretching of the truth at best even when assuring that core enhancements and TDP settings are turned off and set to their proper stock settings.

Admittedly, replacing the infamous TIM “toothpaste” will greatly enhance the die thermals but then you have to consider how the 8700K’s 15-20W power draw delta affects surrounding board components that are critical to power delivery. This situation will only get worse with two added cores and higher stock clocks that will further increase the gap of this delta again leading to die-related thermal issues in SFF. Many including myself believe that the 9900K will compound these problems since it will consume more power than the 8700K and perhaps as much or more than even the 5820K given the core counts and frequencies it is targeting.