Post your Final Fantasy XV Benchmark Results

[EmperorOfFinland]

So i decided to test my RTX 2070 Super GPU overclocks after i got my ram overclocked and cpu tweaked.
After crashing at +70mhz coreclock i took it down to +50mhz and it was stable after that
Running +1000mhz memoryclocks from +900 that i was running earlier for a long time.
So this is ryzen 7 3800x running as stock without PBO as i dont see the point on Allcore oc especially as mine does not hit over 4.2ghz on safe voltages.
I think this is as far as this system goes as im prettymuch at max performance for daily usage.

 

[EmperorOfFinland]

So i was troubleshooting my Xeon X5675 system why it did not like to accept the 6x4GB (24GB) of ram. so when i got it working i just decided to run a benchmark on it after i got it stable. the result did surpise me since i was only 2500 points from the score of my 3800x computer with the same GPU. Pretty good from a CPU from 2011 that i paid 30€ for like couple months ago
I lessened the memory overclock for more allround stable setup but well. i guess the system does atleast work as expected.
The cpu runs at 4.4ghz with 1600mhz memory. GPU runs 50+ on core and 900+ on memory.

 

[harm9963]

Next post with 3090 ! coming soon

 

[T4C Fantasy]

ill only submit 30 series cards from now on and 6000 series cards.

been busy and will continue to be but im curious how new things perform

 

[olegdjus]

I decided to update the result to 2080 Ti. Now I will make more on the game pc with 3080.

• ASUS ROG MAXIMUS XII EXTREME;
• Intel Core i9-10900K 5100 MHz (Ring 4700 MHz);
• G.Skill Trident Z RGB 2х16 GB 4300 MHz cl16-17-17-34 2T;
• ASUS GeForce RTX 3080 TUF Gaming OC 2055/21600 MHz 0.95 V;
• Custom watercool system (MO-RA 420 Pro);
• Intel Optane SSD 900P 480 GB (Windows 10 Pro) + 5 TB SSD;
• ASUS ROG-THOR-1200P;
• PHANTEKS Eclipse P400A White TG.

 

[mrthanhnguyen]

I decided to update the result to 2080 Ti. Now I will make more on the game pc with 3080.

• ASUS ROG MAXIMUS XII EXTREME;
• Intel Core i9-10900K 5100 MHz (Ring 4700 MHz);
• G.Skill Trident Z RGB 2х16 GB 4300 MHz cl16-17-17-34 2T;
• ASUS GeForce RTX 3080 TUF Gaming OC 2055/21600 MHz 0.95 V;
• Custom watercool system (MO-RA 420 Pro);
• Intel Optane SSD 900P 480 GB (Windows 10 Pro) + 5 TB SSD;
• ASUS ROG-THOR-1200P;
• PHANTEKS Eclipse P400A White TG.

Why u downgrade cpu when upgrade gpu?

 

[olegdjus]

Why u downgrade cpu when upgrade gpu?

That's 2 different PC

 

[olegdjus]

• ASUS ROG Crosshair VIII Hero (Wi-Fi);
• AMD Ryzen 9 5950X 4900-4750 MHz;
• G.Skill Trident Z (F4-3200C14D-32GTZSW) 2x16 GB 3733 MHz CL14;
• ASUS GeForce RTX 3080 TUF;
• Custom watercool;
• Corsair RM1000i;
• LD PC-V4
21974

 

[lazord00d]

 

[agent_x007]

Stock cooling for GPU :

 

[lazord00d]

I have to assume looking at these scores that this benchmark/game doesn't utilize multiple graphics cards :-(

Not a true indicator of the actual power of many of these systems including mine if not. Seems like a lot of games are moving away from multi-gpu support but personally I'm not sure why.

I use my machine as a space heater that pays me back and draws less current than an actual space heater... Nicehash you my only friend..$

 

[steevebacon]

.

 

[Agent_D]

AMD Ryzen 5800X @ stock
ASRock B550 Taichi - Bios 1.70
32GB G.Skill (4x8GB SR) DDR4 3600MHz 16-16-16-36
ASRock Radeon 6900XT
Seasonic Titanium 750w

---------------------------

6900XT @ stock - SAM On

---------------------------

6900XT @ Auto Undervolt - SAM On

---------------------------

6900XT @ Rage Mode - SAM On

 

[Vayra86]

I have to assume looking at these scores that this benchmark/game doesn't utilize multiple graphics cards :-(

Not a true indicator of the actual power of many of these systems including mine if not. Seems like a lot of games are moving away from multi-gpu support but personally I'm not sure why.

I use my machine as a space heater that pays me back and draws less current than an actual space heater... Nicehash you my only friend..$

MGPU is dead and has been dying since 2015. Why? Not economically feasible. Look at today, there are wafer/fab shortages everywhere - but even beyond that, it was always a wasteful practice (scaling/overhead perf loss, two PCBs, two coolers, two chips). What did you really get: a higher TCO, worse frametimes and driver support, and sometimes NO support giving you half performance for your money, higher temps on the cards themselves limiting OC ability (and limited anyway to the slowest card), the list is long. And what are the pros (for gaming)? OK you can waste twice as much money on GPUs every gen - that's really it. We had an idea of it being cheaper at some point in time because two lower end GPUs would be faster than the single high end one... but yeah, its really just a meagre discount for a host of issues.

It does look awesome in a system though. Yes. It really does Even just monitoring two GPUs is more fun than one. But yeah. Geek hobbies like that don't last unless you really make an effort to preserve them.

 

[INSTG8R]

MGPU is dead and has been dying since 2015.

Not to mention this a less than reliable test depending on how many times you run it

 

[Vayra86]

Not to mention this a less than reliable test depending on how many times you run it

This bench is like the Geekbench for gaming really. Pretty unreliable and can be manipulated.

 

[INSTG8R]

This bench is like the Geekbench for gaming really. Pretty unreliable and can be manipulated.

This guy gets it!

 

[lazord00d]

I'd agree on a lot of that.. I have 2x rx590s that I paid $145 each for 1 year ago.. they sell for over $200 now when you can find them new. Even used on ebay they're worth more than I paid.. so I gained value, at least temporarily. Which used to be unheard of in the computing world. Also I mine with my cards so the more the merrier, and I run them stock except undervolted so they don't run hot.. with the market the way it's been I've been churning out a fair bit of coin with this and a couple other systems. The actual OPS ability of any given system isn't reflected well in this benchmark, which was my point.

 

[EarthDog]

This bench is like the Geekbench for gaming really. Pretty unreliable and can be manipulated.

Put this in the bin with Userbench......

.... but hey, some are glutton's for punishment and want to start duplicate threads on the same jenky ass benchmark... what can you do, eh? Reporting dupe threads in this section hasn't worked out so far.

 

[Tatty_Two]

In a way it's a shame, SLi/Crossfire used to be a gateway to performance for those who didn't want to, or couldn't spend big $$$ on a GPU, don't or can't buy a HD7970, then buy a HD7870, save up or wait and then maybe a year later add another for 7970+ performance but as some have said, that kind of went out the window with some people buying two more powerful cards at a time as well as it seemed less games being developed that would support MGPU.

 

[T4C Fantasy]

This bench is like the Geekbench for gaming really. Pretty unreliable and can be manipulated.

tl:dr all benchmarks are unreliable

CB is easily manipulated too, this benchmark is affected by both cpu and gpu like 3dmark which all of these benchmarks including 3dmark ranges heavily after each test. none are perfect, but as a person doing all these threads none of that is valid since they all have this issue, my 9900K in CB can range off 1k off a score on small changes and can range between 100 to 500 on no change, and 3dmark is just as ridiculous. Just find a good score and it will be good, ram speeds also affect the scores in this benchmark.

as far as it being a trash benchmark everyone says it is, so i guess it is by their standards, but it looks better than most benchmarks and gives you something to look at like 3dmark.

 

[Agent_D]

AMD Ryzen 5800X @ stock
ASRock B550 Taichi - Bios 1.70
32GB G.Skill (4x8GB SR) DDR4 3600MHz 16-16-16-36
Gigabyte (reference style) Radeon 6800
Seasonic Titanium 750w

Testing with stock 6800 HSF and with 6900XT 2.75 slot HSF

---------------------------

6800 @ Stock - SAM On

---------------------------

6800 @ Auto Undervolt - SAM On

---------------------------

6800 w/ 6900XT HSF - Stock - SAM On

---------------------------

6800 w/ 6900XT HSF - Auto Undervolt - SAM On

---------------------------

6800 w/ 6900XT HSF - Frequency slider set to 2600MHz (max) - SAM On

---------------------------

6800 w/ 6900XT HSF - Frequency slider set to 2600MHz (max) and VRAM slider set to 2150MHz (max) - SAM On

---------------------------

This was a fun experiment and the temperature drops from putting on the 6900XT 2.75 slot HSF were significant; even with frequency and memory sliders set to max, it saw ~10C temp drops over the stock HSF and at stock and auto undervolt settings, it was 15-20C cooler.

 

[T4C Fantasy]

AMD Ryzen 5800X @ stock
ASRock B550 Taichi - Bios 1.70
32GB G.Skill (4x8GB SR) DDR4 3600MHz 16-16-16-36
Gigabyte (reference style) Radeon 6800
Seasonic Titanium 750w

Testing with stock 6800 HSF and with 6900XT 2.75 slot HSF

---------------------------

6800 @ Stock - SAM On

---------------------------

6800 @ Auto Undervolt - SAM On

---------------------------

6800 w/ 6900XT HSF - Stock - SAM On

---------------------------

6800 w/ 6900XT HSF - Auto Undervolt - SAM On

---------------------------

6800 w/ 6900XT HSF - Frequency slider set to 2600MHz (max) - SAM On

---------------------------

6800 w/ 6900XT HSF - Frequency slider set to 2600MHz (max) and VRAM slider set to 2150MHz (max) - SAM On

---------------------------

This was a fun experiment and the temperature drops from putting on the 6900XT 2.75 slot HSF were significant; even with frequency and memory sliders set to max, it saw ~10C temp drops over the stock HSF and at stock and auto undervolt settings, it was 15-20C cooler.

did you optimize the 6900 XT too to try and beat the 3080s?

 

[Agent_D]

did you optimize the 6900 XT too to try and beat the 3080s?

I did not; I only had the 6900XT for 2 days as I was holding onto it for a friend, so I just did some brief testing.

 

[KevsXP]