So 3080 is about 30% faster than 3070 whilst having 50% more cores, hmm! Doesn't seem quite right to me, although I did hear somewhere that something to do with this architecture means it doesn't scale particularly well with increased cores. I bought 3080 (still in preorder queue) as I was sure it was gonna be better value as 50% more cores for less than 50% more money......if these benchmarks hold up then this might not have been such a great deal for me.
It seems like another bogus fanboy post but this might hold a little more water since it does appears that the RTX 3070 are out in the wild.
Either way, I'm definitely not getting a RTX 3000 series. I'm done with NVIDIA this generation. And if AMD turns out to be just hype than performance OVER the 30 series, I'll wait for Big NAVI 2 (RDNA 3) and I'll take a peek at Hopper.
True or false? Below. It does matches some of the rumors for the 3070 and these benchmarks in this article.
[youtube=z_B0mLuz3JY]
So 3080 is about 30% faster than 3070 whilst having 50% more cores, hmm! Doesn't seem quite right to me, although I did hear somewhere that something to do with this architecture means it doesn't scale particularly well with increased cores. I bought 3080 (still in preorder queue) as I was sure it was gonna be better value as 50% more cores for less than 50% more money......if these benchmarks hold up then this might not have been such a great deal for me.
Ashes test shows 3080 as 13/16/20% faster than 3070.
In Time Spy mean graphics (fps) results for 3080 are 26/28% higher than 3070.
In Time Spy Extreme mean graphics (fps) results for 3080 are 23/23% higher than 3070.
In FS Extreme mean graphics (fps) results for 3080 are 28/18% higher than 3070.
In FS Ultra mean graphics (fps) results for 3080 are 29/18% higher than 3070.
And from mean results I picked only results that used top intel's CPUs.
(3DMark is synthetic test with rather small memory footprint.)
Obsidian_Oasis:
It seems like another bogus fanboy post but this might hold a little more water since it does appears that the RTX 3070 are out in the wild.
Either way, I'm definitely not getting a RTX 3000 series. I'm done with NVIDIA this generation. And if AMD turns out to be just hype than performance OVER the 30 series, I'll wait for Big NAVI 2 (RDNA 3) and I'll take a peek at Hopper.
True or false? Below. It does matches some of the rumors for the 3070 and these benchmarks in this article.
[youtube=z_B0mLuz3JY]
Video has slight problem. Each of those systems should have no issue with 60fps recording. But video is 30fps. And on top of it some scenes look like they are under 15fps.
To me, entire thing looks like generic gameplay recorded on midrange PC overlayed with fake system stats.
There is likely very good reason why there are no live fps statistics on screen at all.
Ashes test shows 3080 as 13/16/20% faster than 3070.
In Time Spy mean graphics (fps) results for 3080 are 26/28% higher than 3070.
In Time Spy Extreme mean graphics (fps) results for 3080 are 23/23% higher than 3070.
In FS Extreme mean graphics (fps) results for 3080 are 28/18% higher than 3070.
In FS Ultra mean graphics (fps) results for 3080 are 29/18% higher than 3070.
And from mean results I picked only results that used top intel's CPUs.
(3DMark is synthetic test with rather small memory footprint.)
Video has slight problem. Each of those systems should have no issue with 60fps recording. But video is 30fps. And on top of it some scenes look like they are under 15fps.
To me, entire thing looks like generic gameplay recorded on midrange PC overlayed with fake system stats.
There is likely very good reason why there are no live fps statistics on screen at all.
Yeah, I saw that too, but there's no denying that this YouTuber do have an early access code for Assassin's Creed: Valhalla, so, there's definitely something here.
Dev / Beta tester / AIB partner under another account, etc.?
So 3080 is about 30% faster than 3070 whilst having 50% more cores, hmm! Doesn't seem quite right to me, although I did hear somewhere that something to do with this architecture means it doesn't scale particularly well with increased cores. I bought 3080 (still in preorder queue) as I was sure it was gonna be better value as 50% more cores for less than 50% more money......if these benchmarks hold up then this might not have been such a great deal for me.
No great deal comes from pre ordering something, people haven't learned that, yet.
I wonder if AMDs RT solutions is mostly software rendering than hardware? The consoles will have RT but unless I've missed it, Sony / MS has yet to announce how it would render RT. Dedicated RT cores? GPU acceleration? Software-based RT?
Because whatever the RT solution for the consoles are, I have a feeling it's going to be same solution for Big NAVI which concerns me, especially if Igor's 3D Mark's results are true. 🙁
So 3080 is about 30% faster than 3070 whilst having 50% more cores, hmm! Doesn't seem quite right to me, although I did hear somewhere that something to do with this architecture means it doesn't scale particularly well with increased cores. I bought 3080 (still in preorder queue) as I was sure it was gonna be better value as 50% more cores for less than 50% more money......if these benchmarks hold up then this might not have been such a great deal for me.
Not everything scales linearly, and not everything use all the cores. Not all operations are shader core centric too.
I wonder if AMDs RT solutions is mostly software rendering than hardware? The consoles will have RT but unless I've missed it, Sony / MS has yet to announce how it would render RT. Dedicated RT cores? GPU acceleration? Software-based RT?
Because whatever the RT solution for the consoles are, I have a feeling it's going to be same solution for Big NAVI which concerns me, especially if Igor's 3D Mark's results are true. 🙁
😕