Tuesday, January 29th 2019

AMD Radeon VII 3D Mark, Final Fantasy XV Benchmarks Surface - Beats and Loses to RTX 2080

Benchmarks of AMD's upcoming Radeon VII graphics card have surfaced, courtesy of the one and only, graphics card info and results leaker extraordinaire Tum Apisak. In these scores, and looking purely at the graphics portion of the benchmarks, AMD's solution really does seem to bring the fight to NVIDIA's RTX 2080 - no small feat, considering that it's mostly a shrunk-down version of AMD's previous-gen Vega with overcharged memory and core clocks.

The Radeon VII scores, according to Tum Apisak (take it with a grain of salt), 27400 on the FireStrike test; 13400 on the FIreStrike Extreme bench; 6800 on the FireStrike Ultra test; and finally, 8700 points on Time Spy. Consulting 3D Mark's database, it seems that factory-overclocked RTX 2080 graphics cards usually score around 27000 points on the FIreStrike base and 6400 points on the FireStrike Ultra tests, which means that at least in this synthetic scenario, AMD's graphics card ekes out a win.
Results also surfaced for Final Fantasy XV's integrated benchmark, where the tables are more than turned, however, with AMD's Radeon VII scoring just 300 points higher than the RTX 2070 graphics card, and 1200 points lower than the RTX 2080 AMD wants it to compete against on the standard preset at 2560 x 1440 resolution. I'm not a conspiracy theorist, but Final Fantasy XV is an NVIDIA optimized title, if you'll remember, which might help explain why AMD's Radeon VII plunges in performance as soon as you take it up to the high preset at the same resolution: it now stands a laughable 300 points above NVIDIA's GTX 980 Ti, which is... I'll let you fire off in the comments.
Sources: via VideoCardz, Komachi, Tum Apisak's Twitter
Add your own comment

54 Comments on AMD Radeon VII 3D Mark, Final Fantasy XV Benchmarks Surface - Beats and Loses to RTX 2080

#1
phanbuey
Unless I'm reading the graphs wrong... it looks like it's not beating the 2080 at all... Also that 3dmark it only passes it at 4K (16GB HBM), and is slower/equal at lower resolutions;

I think a "does not get stomped by a 2080" is closer to reality than "beats".
Posted on Reply
#2
Gasaraki
We'll wait for official reviews from tech sites, thanks.

"The Radeon VII scores, according to Tum Apisak (take it with a grain of salt), 27400 on the FireStrike test; 13400 on the FIreStrike Extreme bench; 6800 on the FireStryike Ultra test;" The picture clearly shows 6707.

How is 6707 = to 6800 TPU?
Posted on Reply
#3
Raevenlord
News Editor
phanbueyUnless I'm reading the graphs wrong... it looks like it's not beating the 2080 at all... Also that 3dmark it only catches it at 4K, and is slower at lower resolutions;

I think a "does not get stomped by a 2080" is closer to reality than "beats".
On 3D Mark. It's in the text.

Updated the title for clarity. "Winning is winning, by an inch or a mile", I recall someone saying that. being stomped is being stomped, though.
Posted on Reply
#4
NdMk2o1o
phanbueyUnless I'm reading the graphs wrong... it looks like it's not beating the 2080 at all... Also that 3dmark it only passes it at 4K (16GB HBM), and is slower/equal at lower resolutions;

I think a "does not get stomped by a 2080" is closer to reality than "beats".
Those are the FF IV benchmark tests, read the article.
Posted on Reply
#5
Tsukiyomi91
Why say it "beats" the 2080 when it was getting butchered all over the place, let alone embarrassingly getting beaten by a 1080Ti at 1440p?? So much for a Vega refresh & the use of HBM v2...
Posted on Reply
#6
phanbuey
AMD's own benchmarks had it within a few FPS margin of the 2080, and 6800 to 6400 is definitely within error.

Looks legit. I find the 3dmark Timespy is usually pretty indicative of real world performance.

Also FF IV benchmark looks shady as mentioned in the article.

Its going to be a nice card either way - just comes down to pricing. Nvidia might start cutting given the abysmal sales of the RTX series.
Posted on Reply
#8
phanbuey
VycyousThat score was also with a 2700X. Here it is next to a result with a 9900K, which has a lower graphics score, interestingly enough.

www.3dmark.com/compare/fs/18067351/fs/18071203#
Yeah that is odd - reproducible though, a bunch of reviews had this happen though, where the Rryzens are ever so slightly faster at 4k.
Posted on Reply
#9
Gasaraki
VycyousThat score was also with a 2700X. Here it is next to a result with a 9900K, which has a lower graphics score, interestingly enough.

www.3dmark.com/compare/fs/18067351/fs/18071203#
The OS version between the tests were different.

64-bit Windows 10 (10.0.17134) vs. 64-bit Windows 10 (10.0.17763)
Posted on Reply
#10
Vayra86
I'm not expecting many surprises here. Its likely going to end up within spitting range of a stock 2080 and lacks the OC headroom. Vega all over again.

As for this FFXV 'benchmark'... please. Stahp. Its completely unreliable.
Posted on Reply
#11
HD64G
GasarakiWe'll wait for official reviews from tech sites, thanks.

"The Radeon VII scores, according to Tum Apisak (take it with a grain of salt), 27400 on the FireStrike test; 13400 on the FIreStrike Extreme bench; 6800 on the FireStryike Ultra test;" The picture clearly shows 6707.

How is 6707 = to 6800 TPU?
6871 is the graphic score
Posted on Reply
#12
Vya Domus
FFXV , everyone's favorite benchmark.
Posted on Reply
#13
jabbadap
phanbueyAMD's own benchmarks had it within a few FPS margin of the 2080, and 6800 to 6400 is definitely within error.

Looks legit. I find the 3dmark Timespy is usually pretty indicative of real world performance.

Also FF IV benchmark looks shady as mentioned in the article.

Its going to be a nice card either way - just comes down to pricing. Nvidia might start cutting given the abysmal sales of the RTX series.
Uhm I find 8700 TimeSpy graphics score very underwhelming. Or did I miss something again? RTX 2080 looses even vanilla gtx1080ti on firestrike with hefty margin so beating that does not surprise me at all. But yeah better wait for real world benchmarks.
Posted on Reply
#14
Gasaraki
HD64G6871 is the graphic score
whoopees, wow, that is a weird scoring system.
Posted on Reply
#15
jabbadap
Gasarakiwhoopees, wow, that is a weird scoring system.
Not that it really matters, but graphics score tries to be close to same no matter what cpu is in use. So comparing different runs it's better to compare graphics score than overall score.
Posted on Reply
#16
Unregistered
Vya DomusFFXV , everyone's favorite benchmark.
Use geekbench for CPUs and you've got the perfect match for your cpu and gpu /s
Posted on Edit | Reply
#17
moproblems99
Am I really that blind? I don't see it beating the 2080 anywhere?
Posted on Reply
#18
siluro818
moproblems99Am I really that blind? I don't see it beating the 2080 anywhere?
Look at graphics score, which compares the cards directly, not at the overall PC performance score.

edit: Wait I just saw that they haven't posted all the screenshots here. Basically RVII gets ~6800 3DMark FS Ultra graphics score, while the 2080 gets ~6400 on the same benchmark.

It is to be expected really - Vega 64 beats RTX2070 on that one.
Posted on Reply
#20
B-Real
To the concerned ones: it's not worth calculating performance on a benchmark regarded by every tech site as a complete sh...
Posted on Reply
#21
EarthDog
B-RealTo the concerned ones: it's not worth calculating performance on a benchmark regarded by every tech site as a complete sh...
who said that...
Posted on Reply
#23
efikkan
AMD has promised 25% more performance at the same power, but 25% is not nearly enough to be on par with RTX 2080.


But then technically AMD has never explicitly stated that it's actually on par with RTX 2080.
Posted on Reply
#24
Imsochobo
Vayra86I'm not expecting many surprises here. Its likely going to end up within spitting range of a stock 2080 and lacks the OC headroom. Vega all over again.

As for this FFXV 'benchmark'... please. Stahp. Its completely unreliable.
Vega does not lack overclocking!
It just lacks everything else!
Posted on Reply
#25
phill
Nothing to see here until the actual review....

Not interested in leaks, it just causes nothing but trouble and comments from either side having a paddy... Best leave it be and wait till TPU reviews it for real...
Posted on Reply
Add your own comment
Nov 19th, 2024 00:30 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts