Discussion PPL - Performance Per Liter - Round 3 - Compute Edition...?!?

[Boil]

I would imagine people are waiting to see what was coming. I was/am.
I would have bought the 10900k, but havent seen one available, it also doesnt have pcie 4. And AMD should be announcing their new card too, soon . . .

I would propose to wait until 14th September to see if any other new entrants arrive, and then open ppl 4 when the nvidia 30 series should be available . . .



You mean this

personally I would agree and ignore that protrusion.

The nr200 site also ignores the protrusions . . . as you can see in the stated case volume. Its only there to inform . . (I hope they are including the feet, does anyone know?)

The history of this way of measuring was (if I remember correctly) down to the reviews by some sites including such bumps etc . .

But relatively for PPL we can ignore it, if everyone agrees . . . ..

Yes, that PCIe bracket bit...!

Cooler Master quoting 18.25L is misleading, as they are NOT including the feet (I cover the calculations in my post); the 'real' PPL bounding box volume is 19.45L... JUST making the NR200 SFF...

 

[Elaman]

Hi all,
Huge lurker and noob here, but I just wanted to mention that this forum is awesome and this thread in particular, is just crazy.

This being about compute, I want to point out how remarkable the results of @deprecatedcoder and @hereforthefeast seem to me.
Correct me if I'm wrong but these processors require only 65W, but of course it would be interesting to see how much power is actually being used while running the benchmark. Kind of which system is more efficient per watt.

Otherwise, I just can't wrap my mind around the fact that you can have a 105W processor running full steam in such a tiny case, without it getting absolutely fried ....!?!?!

 

[Aux]

please find below the closing data of PPL 3 including statistics

This marks the end of PPL 3 and no further entries should be made here . ..

PPL 4 will now be opening in expectation of SFF builds with Nvidia Ampere cards from the 17th Sept
We look forward to your entries

 

[thelaughingman]

@Aux I think Oct 1st is more reasonable, gonna take some time between Ampere order starts and people actually have the cards on hand

 

[Aux]

@thelaughingman

true

but If I were building a gaming / high end rig now (Im not, I currently need a fast server), I would be looking at the Ampere cards . ..

My next rig will be sometime next year, more late next year if not 2022 . . . unless of course I suddenly get lucky and win a lottery or something . . .. .

but if someone wants to post a rig for ppl 3, I can drop it into the sheet, no worries

did you have your watercooling update yet?

 

[thelaughingman]

@thelaughingman
did you have your watercooling update yet?

unfortunately no. the order got stuck in custom for so long that I moved to the other end of the country. now it is staying at my parent's so I can only do the update when I visit home ?

 

[petricor]

Holy crap... Unless my calculator is broken, this thing by @catawalks would appear to score 187k...

 

[petricor]

Holy crap... Unless my calculator is broken, this thing by @catawalks would appear to score 187k...

ignore - have overread an external GPU fitted to it... still pretty impressive...

 

[eddy178]

Men i was amazed with the Sancho build and im totally doing something similar. Need a lot of DIY, but damn, that petricor S4M takes the cake and eats it. The only way i could possibly do that build is to actually pay him to do it and send it to me lol

 

[Elaman]

Men i was amazed with the Sancho build and im totally doing something similar. Need a lot of DIY, but damn, that petricor S4M takes the cake and eats it. The only way i could possibly do that build is to actually pay him to do it and send it to me lol

Yeah, you need some tools at your disposal, and I feel like a lot of experience is also needed.
But it's not just the maxed-out S4M, I feel like we could make (e.g.) a top 10 by eliminating repeat entries by same author, I wouldn't be able to replicate any of those builds either. But that's a great thing!

 

[Boil]

... make (e.g.) a top 10 by eliminating repeat entries by same author ...

Why? All of the "same author" entries are different builds. Only allowing one build per "author" seems to go against the spirit of the "competition". We are all just trying to have fun here! And each round of the PPL threads goes until it hits 25 entries, so why bother with a "Top 10" list?

 

[rfarmer]

Why? All of the "same author" entries are different builds. Only allowing one build per "author" seems to go against the spirit of the "competition". We are all just trying to have fun here! And each round of the PPL threads goes until it hits 25 entries, so why bother with a "Top 10" list?

Agree with Boil. I have 2 entries here. First was fully watercooled in the Ghost with large tophat, 2nd entry was after switching to aircooling and changing my GPU. So my second entry not only had a smaller volume but upgraded GPU.

 

[thelaughingman]

Yeah, you need some tools at your disposal, and I feel like a lot of experience is also needed.
But it's not just the maxed-out S4M, I feel like we could make (e.g.) a top 10 by eliminating repeat entries by same author, I wouldn't be able to replicate any of those builds either. But that's a great thing!

Why? All of the "same author" entries are different builds. Only allowing one build per "author" seems to go against the spirit of the "competition". We are all just trying to have fun here! And each round of the PPL threads goes until it hits 25 entries, so why bother with a "Top 10" list?

Agree with Boil. I have 2 entries here. First was fully watercooled in the Ghost with large tophat, 2nd entry was after switching to aircooling and changing my GPU. So my second entry not only had a smaller volume but upgraded GPU.

I think at only 25 entries there's not point in truncating the table.

However if somehow in the future we have like 100+ entries each round or the benchmarks/format are consistent enough across rounds for a sufficiently large and comparable dataset, then I would be interested to see a 'battle of the cases' table, i.e the best score/setup for each case (some cases are inherently more popular and will have more entries than others)

 

[REVOCCASES]

This could make some potential champion I guess...

Log - G-unique DiY stuffs \\3.1L brickless ITX Gigabyte RTX2070 build ,no name yet

after years of diy i almost lost all passion posting a thread tell how small of a case i have designed and made,but i would love to introduce how to solve the power solution of the smallest brickless itx case. cableless Gstick :18cm*5cm*4cm 600watts capable,very silent,very high efficiency up to...

smallformfactor.net

 

[AURMEND]

once the G-Stick is here im gonna go back to first place. (not properly OCd)

 

[CubanLegend]

once the G-Stick is here im gonna go back to first place. (not properly OCd)

You need to go all-in and move up to a 5950x

 

[AURMEND]

You need to go all-in and move up to a 5950x

Nah, can hit CB R20 10k points with more Tuning. Need a 25% gain to even consider an upgrade

 

[Aux]

@AURMEND @CubanLegend

We are now on

Performance per litre - Round 4​

Discussion - Performance per litre - Round 4

I'm in the same boat, waiting on Big Navi and possible upgrade from 3900X to 5950X before joining this round ? then I can say peak-AMD achieved! Big Navi for me as well, but I think I am gonna stick with my 3900X for now. Zen 4 & RDNA3 though, that will be a fully new build! Probably AM5 /...

smallformfactor.net

thanks