By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Pemalite said:
Captain_Yuri said:

There's a reason why Teraflops really can't be compared between different architectures...

It can't even be compared with the same architectures.

Grab a Radeon 7750 DDR3 and compare it against the Radeon 7750 GDDR5, it's going to have half the performance.
https://www.hardware.fr/focus/76/amd-radeon-hd-7750-ddr3-test-cape-verde-etouffe.html

Grab a Geforce 1030 DDR4 and compare it against the Geforce 1030 GDDR5, it's going to have half the performance.
https://www.gamersnexus.net/hwreviews/3330-gt-1030-ddr4-vs-gt-1030-gddr5-benchmark-worst-graphics-card-2018

Identical GPU's, Identical Architecture, massive differences in capability.

Or take the...
Radeon 7970 @ 3.78 Teraflops.
Radeon RX 470 @ 3.73 Teraflops.

Both Graphics Core Next, Night and day difference...
https://www.anandtech.com/bench/product/1872?vs=1722

freedquaker said:

7.5 GB of usable RAM vs 9 GB of usable RAM, yes they are roughly on par. Let's not forget that the very same games run on PS4pro with only 6 GB of usable RAM as well as PS4 with a mere 5.5 GB.... The only difference is usually the resolution and texture quality. Well a 3 GB of difference between the PS4pro and XBX creates a difference in texture quality but a 1.5 GB will not, but merely resolution differences, which is why I am not so certain that Lockhart can run all XBX games at 4K resolution but will not have a problem running them at 1440p generally speaking.

Ram is Ram.

The Playstation 4 Pro *could* have shown much larger differences than just resolution/framerate if it wasn't memory constrained... It's a doubling+GST in GPU resources remember. - It was certainly memory constrained. - Could an SSD have mitigated that somewhat? Sure. But it's still not a replacement for lots of fast Ram...
There are plenty of PS4 Pro titles that didn't even get a resolution bump... Or employed a dynamic resolution and often sat around the 1080P mark anyway.

Microsoft saw through this issue and realized that in order to chase higher resolutions with higher quality assets, you needed more memory, hence the upgrade to 12GB.

Lockhart due to it's performance level will be better suited to 1080P.

freedquaker said:

RDNA 2 is roughly 50%+ faster than Polaris, but of course things vary; however, in the general scheme of things, pretty much 2020 and earlier games should run similarly, if not better, on Lockhart, than in XBX, even if they need to be scaled down to 1440p for one reason or another. In the next-gen compatible titles, however, it will be more performant than XBX due to use of faster RAM, faster SSD, and the utilization of the faster CPU.

Actually. RDNA2 isn't 50% faster than RDNA1. AMD has NEVER made that claim.

And RDNA is certainly not faster than Polaris on a per-flops basis... But on a performance per watt basis.

https://www.anandtech.com/bench/product/2525?vs=2578

Essentially a Radeon 5500XT @4.8-5.2 Teraflops is slightly better than a Radeon RX 570 @ 4.8-5.1 Teraflops... Most of the chip has been kept equal too, such as memory bandwidth.

But that is the key thing, it's performance "per watt", not just performance.

And that is entirely expected, RDNA is built on top of the foundations of GCN, uses the exact same instruction set, it's a hybrid GPU design.
https://en.wikipedia.org/wiki/RDNA_(microarchitecture)#Architecture

RDNA2 will take a full architectural deviation from Graphics Core Next. Finally. Yay.

But don't propagate the false claim that it will offer a 50% performance improvement, that is twisting AMD's wording.
https://www.techpowerup.com/264538/amd-rdna2-graphics-architecture-detailed-offers-50-perf-per-watt-over-rdna


freedquaker said:

Finally, to address some of the suggestions here, the price gap between XBSX will be around $100-$150, depending on what is included in the configuration. If it comes as a digital only console, we might be looking at a $250 4TF 10 GB Lockhart vs fully-fledged $400 XBSX (or for $350 and $500 respectively), which makes even more sense, competing with a regular $500 PS5 and a digital only $450 one, which will then make the competition very interesting between the two. In such a scenario, PS5 will still win in all Asia, and the non-English speaking world, while the English speaking would be home to a fierce competition.

I won't personally speculate on price, because there hasn't been any real concrete evidence on the various price points manufacturers are gunning for... Or how that will translate to certain market success/failures.

Dulfite said:
Can there be a third option? One without a disk drive that as just as powerful as the Series X, but slightly smaller or shorter and $50-150 cheaper?

Doesn't need to have a massive hardrive as I'm used to uninstalling and reinstalling games on my Switch and PC all the time. I only like to keep installed what I'm currently playing anyway and I am digital only now.

That is entirely possible, just enough storage for the OS+Future updates and a reliance on expandable storage.



I never said that AMD said that RDNA2 is 50% faster than RDNA, no, I said RDNA2 is 50%+ faster than Polaris, and this is not a claim made by AMD but others out there like Mark Cerny (literally said this in his presentation a couple of months back) and MANY PC reviewers in their benchmarks. If you search around, you will see that this is indeed the case. Even RDNA cards are much faster than Polaris, TFlops to TFlops.

A simple but popular comparison:
https://www.eurogamer.net/articles/digitalfoundry-2019-teraflop-face-off-current-gen-vs-next-gen

Also one example you have shown between DDR3 vs GDDR5 is very interesting and finds an average difference of 36% performance deficit with DDR3 (15% to 54%), which reminds me of the situation between XB1 and PS4 if it weren't for the ESRAM. However, the memory bandwidth difference between the XBX and the Lockhart would be relatively small and it will be mostly down to the architectural differences, as well as CPU and Disc bottlenecks.



Playstation 5 vs XBox Series Market Share Estimates

Regional Analysis  (only MS and Sony Consoles)
Europe     => XB1 : 23-24 % vs PS4 : 76-77%
N. America => XB1 :  49-52% vs PS4 : 48-51%
Global     => XB1 :  32-34% vs PS4 : 66-68%

Sales Estimations for 8th Generation Consoles

Next Gen Consoles Impressions and Estimates