Unigine Heaven 4.0 Benchmark Scores Part 2

[NMflyer]

Question: Heaven 4.0 DirectX 11 benchmark recently started jumping to the desktop during execution of Scene 25/26. Is anyone else experiencing this behavior?

RTX 3080 Ti FE
i9-12900KS
32 GB DDR4-4000 C16

 

[altec.gs]

AMD Ryzen 5 1600 3.2Ghz
16 GB DDR4 @ 3.2Ghz Cl16 (2x8)
Sapphire RX 6800 (non XT) Pulse OC 16GB (standard version)
1TB 980 Samsung

 

[b.infinite]

Hi guys I'm testing my new GPU with default stock frequencies, but I don't understand why the "Min FPS" is so low, it worries me a little. What do you think about it?

View attachment 297364

Overclock might be too high. Do a baseline run then increase and note scores to find sweet spot

 

[Marcus Vinicus]

See further system info in my signature below

 

[AVATARAT]

A little update with a new CPU

7700x
6800 XT
2x32GB DDR5@6200MHz

 

[QuietBob]

Tested the slowest DX11 compatible GPU @ 1080p lowest. Coupled with the slowest 64-bit Intel desktop CPU from 2004, the single core Celeron D 326 (Prescott). Interestingly, at this resolution it held back the HD5450 in only a handful of scenes:



Posted from the actual machine

 

[DR4G00N]

Behold the raw power of quadfire HD 5970's!... at 1360x768... That's all my TV would do.
Using 22.10 NimeZ driver.

For reference my GTX 1050 scored 1531 pts @ 1936/1850MHz with under 75W.

 

[MrNobodie]

I've played powerpoints with more slides per seconde than what the FX 580 gave me there.

 

[sounik]

Rtx 3080 ftw3 ultra overclocked for everyday use.
1080p and 1440p scores.

 

[:D:D]

i7-6800K @ 3.8GHz + GTX 1080 FE @ 12.5/101.25 = 295

While GPU clock is set for 12.5MHz and reported by GPU-z, I get the feeling Nvidia is telling porky pies.

 

[Vayra86]

i7-6800K @ 3.8GHz + GTX 1080 FE @ 12.5/101.25 = 295
View attachment 312479

While GPU clock is set for 12.5MHz and reported by GPU-z, I get the feeling Nvidia is telling porky pies.

Aren't you running off the IGP or something? Card is idle and perf looks to match. If this is all the card'll do, its bricked...

 

[:D:D]

No integrated graphics. Could have just posted another GTX 1080 average OC'd bench but thought this was more interesting. I believe results should have been lower for 12.5MHz. Running with a set gpu clock of 115MHz shows the same result as 12.5MHz so something fishy going on with how nvidia are handling and reporting clocks. With gpu clocks at 600MHz and 1200MHz and mem at 1377MHz the bench shows about twice the performance gain (memory bandwidth gets cut off with the lower gpu clocks and higher memory clock).

 

[:D:D]

Here's a GTX 1050 Ti result since can only see one other in the DB. CPU at 4GHz.

 

[NoiseBox]

Hello,

I thought I'd chip in here with a few unusual systems. My main system is a rather old Dell Dell Precision T7610 (dual Ivy-Bridge Xeons, 64GB ECC Registered DDR3 RAM, and a 3060 12GB GPU)
I'm not much of a gamer so this system does just fine for me. Here is the score:

___
My next system is a 2013 Mac Pro, running Windows 10 via BootCamp.
It's a quad-core Xeon E5-1620 v2 (4c/8t@3.70Ghz) with dual Radeon D300 2GB GPUs running in Crossfire mode (Unigine seems to see it as such anyway) and 16GB RAM. A slightly better CPU and more RAM is on the way, and I will re-run the Benchmark and see if that makes any difference in scores.
In the meantime, here is that result:

Cheers!

 

[Jetster]

Hello,

I thought I'd chip in here with a few unusual systems. My main system is a rather old Dell Dell Precision T7610 (dual Ivy-Bridge Xeons, 64GB ECC Registered DDR3 RAM, and a 3060 12GB GPU)
I'm not much of a gamer so this system does just fine for me. Here is the score:
View attachment 323642
___
My next system is a 2013 Mac Pro, running Windows 10 via BootCamp.
It's a quad-core Xeon E5-1620 v2 (4c/8t@3.70Ghz) with dual Radeon D300 2GB GPUs running in Crossfire mode (Unigine seems to see it as such anyway) and 16GB RAM. A slightly better CPU and more RAM is on the way, and I will re-run the Benchmark and see if that makes any difference in scores.
In the meantime, here is that result:
View attachment 323649


Cheers!

Run the test with the settings as described by the op in the first posts

 

[tugrul_SIMD]

rtx4070 ventus 2x at 66% power (133Watts)

rtx 4060ti gigabyte oc 135% power limit (~180Watt)

Even when 4070 is underpowered & badly cooled, a 4060 with +35% extra power can't come close to it in terms of tesselation or cache performance.

 

[DR4G00N]

RTX 3060M 6GB @ ~1900MHz Core / 1750MHz Mem, 80W TGP.

 

[Veseleil]

Run the test with the settings as described by the op in the first posts

Really, many benchmark threads became useless because of this (the absence of reading focus or/and ignorance), and kinda like a door stops.

 

[NoiseBox]

Run the test with the settings as described by the op in the first posts

Here ya go, the Mac Pro is in use right now, but my Precision is free:

Really, many benchmark threads became useless because of this (the absence of reading focus or/and ignorance), and kinda like a door stops.

Yeah sorry about that, I looked at several pagaes of benchmarks, and missed the subtle differences on the screenshots, so I ran it at the default settings.

As promised, the Mac Pro 2013 with TPU thread requirements.

My new(ish) CPU came in this weekend, and I'm planning on upgrading the Mac Pro today. From what I have read it shouldn't make a difference in the scores, but I will test all the same.