Discussion Performance per litre - Round 4

[Boil]

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 / DDR5 / PCIe Gen5...?

 

[Lucas]

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!

Also waiting for Big Navi, and then I'm gonna pray they will make a 3070 / big navi below 180mm so we can crunch some points in here! Also looking for the 5950X. Hope to run the 3070 and 5950X on 500W, gonna see how that will turn out.

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 / DDR5 / PCIe Gen5...?

I've been looking at DDR5 etc as well, but from my understanding we will have to wait until mid 2021 at least for that.

 

[Boil]

Also waiting for Big Navi, and then I'm gonna pray they will make a 3070 / big navi below 180mm so we can crunch some points in here! Also looking for the 5950X. Hope to run the 3070 and 5950X on 500W, gonna see how that will turn out.

Big Navi Nano would be awesome, especially if there is a full-cover water block to match...!

I've been looking at DDR5 etc as well, but from my understanding we will have to wait until mid 2021 at least for that.

I will most likely wait until 2022 or so for a fully new system; by then any issues with the new AM5 (X670 chipset?) platform should be smoothed out, and same for DDR5 RAM (and hopefully pricing will not be so crazy as well)...

Also wanting to secure a Winter One with SPK for future Zen X/RDNAX build...!

 

[Lucas]

General Entry Rules

Great competition Aux, I wonder whether there are any Google sheet / excel list of all the entries from round 1-4. That is, a complete hall of fame. I know the calculations have changed some, but there should be no issue to automatically recalculate with the known hardware. I'm writing a study where I would like this anyhow, so I could do it if I can get hold of the spreadsheet.

 

[Aux]

Great competition Aux, I wonder whether there are any Google sheet / excel list of all the entries from round 1-4. That is, a complete hall of fame. I know the calculations have changed some, but there should be no issue to automatically recalculate with the known hardware. I'm writing a study where I would like this anyhow, so I could do it if I can get hold of the spreadsheet.

All of the data is available on the forum, through the various ppl rounds. I have several excel sheets with the data in, but not with the same format.

 

[Lucas]

All of the data is available on the forum, through the various ppl rounds. I have several excel sheets with the data in, but not with the same format.

Yes, I've seen pictures of the data in the other threads. Could you share the excel sheets with me?

 

[Boil]

@Aux

Cinebench R23 is now out...! Reserve for Round Five, or wipe the board on Round Four & start over...?!? ;^p

 

[riba2233]

@Aux

Cinebench R23 is now out...! Reserve for Round Five, or wipe the board on Round Four & start over...?!? ;^p

R23 is actually more interesting for sff since it lasts 10 min, so it will test the cooling better, it is not so forgiving as R20.

 

[Boil]

R23 is actually more interesting for sff since it lasts 10 min, so it will test the cooling better, it is not so forgiving as R20.

If you go to the File menu then check the Advanced Benchmarks options, it will give you a drop-down on the interface with the choices:

Off
10 minutes (throttle testing)
30 minutes (stability testing)

30 minute test, with temp monitoring, would be really interesting to see in some of the smaller SFF chassis offerings...!

 

[Aux]

@Aux

Cinebench R23 is now out...! Reserve for Round Five, or wipe the board on Round Four & start over...?!? ;^p

Yeah, I see it. Could go to a round five.

Is this in general agreement with everyone?

 

[Lucas]

Yeah, I see it. Could go to a round five.

Is this in general agreement with everyone?

Looks good. With only one entry this round you could even ask Diamorif if he could redo his entry with more upto-date software.
Also, not sure if you saw my last message, would it be possible for you to share the excel sheet with me for my studies?

 

[Aux]

If you go to the File menu then check the Advanced Benchmarks options, it will give you a drop-down on the interface with the choices:

Off
10 minutes (throttle testing)
30 minutes (stability testing)

30 minute test, with temp monitoring, would be really interesting to see in some of the smaller SFF chassis offerings...!

So what are we proposing, using the results from R23 on a 30 minute stability testing?

I for one, would be interested in such results.

Is everyone interested in this competition happy with that?
Probable path would be to ask @Diamorif to recalculate with r23. Would that be ok?

@Lucas how do you propose I send you the sheet?

 

[Boil]

So what are we proposing, using the results from R23 on a 30 minute stability testing?

I for one, would be interested in such results.

Is everyone interested in this competition happy with that?
Probable path would be to ask @Diamorif to recalculate with r23. Would that be ok?

If AMD would get their drivers in order in regards to Vulkan, then OctaneBench could become the perfect GPU Compute-based component for PPL testing...?

And then we could have a PPL competition that equally includes Linux, Windows, & ASi Macs...!

The REAL deal though, that would be a cross-platform (Linux/Windows/ASi Mac) benchmark that also can run on either AMD or Nvidia GPUs; so everyone can "run what they brung"... And I mean a benchmark that would stress BOTH CPU & GPU at the same time, all while monitoring / reporting assorted temps & clock speeds...!

Then we could really, Really, REALLY see who has the best PPL...!

Because a quick single run of R20 & Timespy will probably not show any "extended runtime" throttling issues...?!?

 

[Lucas]

So what are we proposing, using the results from R23 on a 30 minute stability testing?

I for one, would be interested in such results.

Is everyone interested in this competition happy with that?
Probable path would be to ask @Diamorif to recalculate with r23. Would that be ok?

@Lucas how do you propose I send you the sheet?

Thanks, you could send it through email, I PM you! If you have it in google drive, you could give me viewing access and I can then copy it.

 

[aromachi]

Here's another entry. Very late to the party, and it will be destroyed by current hardware, but I don't mind. Been a long time fan of the PPL competition and wanted to be a part of it.

Case: fully custom sandwich-style with external fan. 115mm x 192mm x 246mm. Screws holding side panels added 2.5mm to all sides. Corner tabs added 1mm to top and bottom. +.239L for external fan.
Total volume= 5.67L

CPU: R5 1600
GPU: Zotac 1660 Ti

Cinebench: 2985
Octane: 135.50
Timespy: 6486

((2985 * 6486)/(5.67^1.5))/100 = 14339.92
((2985 * 135.50)/(5.67^1.5))/100 = 299.58

 

[Boil]

Case: fully custom sandwich-style with external fan. 115mm x 192mm x 246mm. Screws holding side panels added 2.5mm to all sides. Corner tabs added 1mm to top and bottom. +.239L for external fan.
Total volume= 5.67L

PPL volume is measured as a bounding box, so that protruding CPU cooler is gonna increase your overall volume measurement more than the notated 0.239L, which will change your overall PPL calculation...

 

[aromachi]

PPL volume is measured as a bounding box, so that protruding CPU cooler is gonna increase your overall volume measurement more than the notated 0.239L, which will change your overall PPL calculation...

Yeah, you're totally right. That really destroys my numbers. I'll probably just withdraw my entry at this point.

 

[confusis]

Meh, with a few extra bucks on an aftermarket cooler, it'd fit fine. Consider it a hotrod build. I'll allow it. (And have done similar in the past )

 

[Boil]

Yeah, you're totally right. That really destroys my numbers. I'll probably just withdraw my entry at this point.

No need to remove your entry, just adjust the numbers properly...! It's all for fun anyway...! ;^p

Meh, with a few extra bucks on an aftermarket cooler, it'd fit fine. Consider it a hotrod build. I'll allow it. (And have done similar in the past )

Bounding box is bounding box...! Otherwise, a "Hot Rod Exemption" could lead to a BeQuiet! tower cooler sticking out of an Inwin Chopin...! ;^p

 

[aromachi]

No need to remove your entry, just adjust the numbers properly...! It's all for fun anyway...! ;^p

I think that would do too much damage to my pride Maybe I'll have a new build by PPL Round 7. Can't promise that I won't try pulling the same edgy shenanigans though..