Digital Foundry Article Technical Discussion [2018]

[bitsandbytes]

Article layout as well.





Why waste the additional silicone or even select that particular GPU (Polaris) if this was [is] going to be an issue?

My take from the article is that Microsoft took their time (a learned lesson from XB1) and selected the appropriate hardware, then optimally configured the X hardware on utilizing more of the silicone strengths, in essence providing a more powerful design. The Pro just seems rushed in retrospect. Not dumping on it (I own one, as well as an X), not at all. Just that Sony could have made smarter choices on Pro's overall architectural approach.

My impression about PS4 Pro is that Sony were so preoccupied and paranoid about not breaking compatibility with the 700+ PS4 games.

 

[Shortbread]

Yield is necessarily the reason for disabling CUs, today all GPUs and all consoles disable CUs for yield. The consoles however cannot resell the perfect chip at a higher margin, so they have a compromise to make. I.e. the additional area used for redundancy ends up less expensive than the yield hit of requiring perfect chips.

The existence of a 470/570 SKU means there are chips with defects or other issues keeping them from the top specification. The console APU only has one bin, and as a component of a complete system it has to compete for a share of the overall cost.
Even if Polaris at this point has highly mature yields, they would have been lower early in the product life cycle. The console chip's specification is constant from the time of least mature yields until it stops production.

GCN chips also appear to usually disable CUs in increments that are a multiple SE count, so it may be that this is the least that can be disabled.

I guess my perspective was wrong. I was looking at the Pro from a lower volume product perspective. That the Pro wasn't replacing the base model PS4, thus allowing Pro more leeway on getting the best samples because it is a low volume item. I understand the reasons why CUs are disabled ...but as I stated, my thought process was stuck on "quality," rather than a mass volume product.

 

[sniffy]

The asymmetry of the second pair of shader engines on the Pro is interesting but I think it was probably an area saving move to limit the x axis more than anything else.

 

[Shortbread]

My impression about PS4 Pro is that Sony were so preoccupied and paranoid about not breaking compatibility with the 700+ PS4 games.

I believe Sony was more worried about pricing. And rightfully so, the PS3 launch price was a disaster. Microsoft took a chance with X pricing, because they felt the hardware was worth it (and it is I must say).

Remember there were two possible Pro designs (one with a beefier CPU), but Sony (as far as I know) didn't want to take the risk on a more costlier system.

 

[bitsandbytes]

I believe Sony was more worried about pricing. And rightfully so, the PS3 launch price was a disaster. Microsoft took a chance with X pricing, because they felt the hardware was worth it (and it is I must say).

Remember there were two possible Pro designs (one with a beefier CPU), but Sony (as far as I know) didn't want to take the risk on a more costlier system.

If you are referencing the "verified" GAF insider about PS4K then IMO he embellished (to put it kindly) whatever real knowledge he had. He ignored the real evidence like doc leaks and pretty much disappeared never to be heard of again...

Did Sony have two options in ~mid 2014 for their mid-gen console? Possibly/probably, but that decision would have been made years (at least two) before we heard any hint/rumour. IMO.

 

[liolio]

I hope DF is to compare the newly released AMD APU to the Xone S and PS4 Slim, it should be in the ballpark of the first one as far GPU is concerned and lead both on CPU performances.
It should beat my Alienware Alpha i3 overall too. QUite an interesting piece of kit that sell for 170€, mini gaming PC scene should get interesting.

EDIT cheap gaming laptop pushing 1366*768 could also get real.

 

[HMBR]

I hope DF is to compare the newly released AMD APU to the Xone S and PS4 Slim, it should be in the ballpark of the first one as far GPU is concerned and lead both on CPU performances.
It should beat my Alienware Alpha i3 overall too. QUite an interesting piece of kit that sell for 170€, mini gaming PC scene should get interesting.

EDIT cheap gaming laptop pushing 1366*768 could also get real.

I think the Xbox One achieves a better result than this, but it's hard to know for sure, since the XB uses dynamic resolution, but I think low is a good bit bellow the XB1 quality, still I certainly agree that it would be interesting to compare in more titles and more directly, the GPU and CPU side are certainly stronger, but the memory bandwidth is lower.

 

[liolio]

I think the Xbox One achieves a better result than this, but it's hard to know for sure, since the XB uses dynamic resolution, but I think low is a good bit bellow the XB1 quality, still I certainly agree that it would be interesting to compare in more titles and more directly, the GPU and CPU side are certainly stronger, but the memory bandwidth is lower.

Thanks for the input.
I dug through some more reviews and it seems that I went a little optimistic, the fast DDR4 and more bandwidth efficient ROPs are still not a match to the raw bandwidth both the X and PS4 SOCs are provided with.

Spoiler

I still wish somebody dig the topic, playing with power setting, memory speed, etc.)
Still pretty good performances, definitely workable for a cheap gaming laptop or mini PC. There may be a sweet (and cheap) spot for the APU as from 11 CU to 8 you lose ~10% performances. Couple with decent but not top end (costs...) DDR4 the 99€ may be the most dollars efficient option.

 

[BRiT]

Can Kingdom Come Deliverance's tech deliver its ambitious vision?
Every console version tested, with Xbox One X on top.

http://www.eurogamer.net/articles/digitalfoundry-2018-kingdom-come-deliverance-console-analysis

Visually speaking, the situation is straightforward on console, where both base and enhanced PlayStation 4 and Xbox One consoles share similar strengths and weaknesses. There's a decent degree of cross-platform conformity here: texture resolution is identical, as is shadow quality, and motion blur sampling. A real issue with the console releases is the presence of low resolution art in some cases - a problem that manifests right away on the title screen, just as it did on the notorious CryEngine-powered Lichdom Battlemage. Equally, low resolution assets show up across the game on the whole, if you stop to look too closely.

Speaking of loading, it's worth making a quick point here. This being an open world game in the vein of Grand Theft Auto or The Elder Scrolls series, you'll be sitting through a lengthy initial loading screen. The problem is, you'll be waiting up to two minutes to just get to the title screen on a standard PS4, before selecting your save from there, which sets off another lengthy wait. PS4 Pro cuts the initial wait down to 1.5 minutes, Xbox One lands at one minute, 13 seconds - while Xbox One X fastest at just 53 seconds.

This sequence follows if you actually load a game too, but thankfully you don't have to wait quite as long. Frustratingly this means another 28 seconds wait on an Xbox One X, and up to 44 seconds on the worst performer - again, the PS4. As a one-off it's not that bad, given the world is one, seamless sprawl you can walk across, from end to end. However, supposing you use a fast travel system in the game, this is roughly the sort of time-frame you can expect on each console - and that could affect your experience of it.

...

Overall, there's a simple pecking order if you're choosing to play Kingdom Come Deliverance on consoles. Xbox One X delivers the highest resolution - a 77 per cent boost over PS4 Pro, in fact - and smoother performance on top of that. However, between the sub-30fps drops, the obvious texture and geometry pop-in, and long loading times, it's fair to say that no platform is covered in glory here. It could be a simple matter of optimisation through patches that improves this situation - and we hope that these improvements are eventually delivered. Beyond the surrounding furore, Kingdom Come Deliverance actually has some great ideas behind it. It's a decent alternative to the Elder Scrolls, with some amazing backdrops and an interesting historical hook to its story - it's just the technology at its heart that's struggling to keep up. With this in mind, the PC version may be the way forward for those considering the game, and it's something we're planning to look at shortly.

 

[Tkumpathenurpahl]

However, between the sub-30fps drops, the obvious texture and geometry pop-in, and long loading times, it's fair to say that no platform is covered in glory here.

That's not ideal, but none of it's a deal breaker.

Plenty of great games drop below 30fps. Hell, my favourite game of all time, Shadow of the Colossus, routinely dropped to the low 20's on my PS2. It just made it more awesome "OMG my console can't even cope with the size of this bastard!"

Pop-in's annoying, but it never stopped me loving Uncharted 1.

And lengthy load times plagued the Witcher 3 too. That wasn't enough to stop me from spending every non-working hour of my life playing it. If anything, it just gave me ample opportunity to skin up ;D

With this in mind, the PC version may be the way forward for those considering the game

I don't mean to be an arse, but isn't that always the case?

That said, the global illumination is reason enough to get the PC version. Fit!!!

 

[function]

Overall, there's a simple pecking order if you're choosing to play Kingdom Come Deliverance on consoles. Xbox One X delivers the highest resolution - a 77 per cent boost over PS4 Pro, in fact - and smoother performance on top of that.

The idea that the X1X is a "brute force" system compared to the PS4Pro is simply not backed up by empirical evidence. It's a baseless zealot claim that unfortunately has gained traction, even here.

In terms of perf/Watt, perf/FLOP, perf/ROP, perf/mm^2 of die, perf/$$$ .... the X1X is clearly leading the PS4Pro and has been since its introduction.

MS did a huge amount of profiling work on real software, tested changes using virtual system configurations and made lots of tweaks to memory access, caches, and per-chip board power delivery to squeeze everything they could out of a market leading but still finite amount of resources.

If the X1X is "brute force" compared to the PS4Pro, how come it's beating it in every possible measurable metric of efficiency when it's not artificially limited by a resolution or frame rate cap?

 

[sniffy]

The idea that the X1X is a "brute force" system compared to the PS4Pro is simply not backed up by empirical evidence. It's a baseless zealot claim that unfortunately has gained traction, even here.

In terms of perf/Watt, perf/FLOP, perf/ROP, perf/mm^2 of die, perf/$$$ .... the X1X is clearly leading the PS4Pro and has been since its introduction.

MS did a huge amount of profiling work on real software, tested changes using virtual system configurations and made lots of tweaks to memory access, caches, and per-chip board power delivery to squeeze everything they could out of a market leading but still finite amount of resources.

If the X1X is "brute force" compared to the PS4Pro, how come it's beating it in every possible measurable metric of efficiency when it's not artificially limited by a resolution or frame rate cap?

Do people actually argue that? I definitely agree with you, everything from the SoC to the actual console build is far more impressive than the Pro. I'm still amazed they managed to get to 1172MHz inside of a 175W envelope - that includes 8 Jaguar's clocked as high as they'll go and 12GB GDDR5. As far as I'm aware the first GCN product to ship with clocks that high is Polaris is it not? I don't think either X1X or the Pro are strictly Polaris either, rather older GCN tech with new features added on. So that they got the CU's to go that quickly in such a reasonable power envelope is phenomenal.

 

[BRiT]

Do people actually argue that?

Yes, certain camps have argued that over here: https://forum.beyond3d.com/posts/2021833/ It also seems to have others trying to push it to become a mantra on other forums too, which is really sad.

 

[Tkumpathenurpahl]

Stop being so insecure.

The X1X is a great console. A great console that externally looks like a PS2 and internally looks like a PS4.5.

Why wouldn't you take what works and improve upon it? Microsoft did, and they did a damned fine job of it.

 

[Silent_Buddha]

Can Kingdom Come Deliverance's tech deliver its ambitious vision?
Every console version tested, with Xbox One X on top.

http://www.eurogamer.net/articles/digitalfoundry-2018-kingdom-come-deliverance-console-analysis

Visually speaking, the situation is straightforward on console, where both base and enhanced PlayStation 4 and Xbox One consoles share similar strengths and weaknesses. There's a decent degree of cross-platform conformity here: texture resolution is identical, as is shadow quality, and motion blur sampling. A real issue with the console releases is the presence of low resolution art in some cases - a problem that manifests right away on the title screen, just as it did on the notorious CryEngine-powered Lichdom Battlemage. Equally, low resolution assets show up across the game on the whole, if you stop to look too closely.

Speaking of loading, it's worth making a quick point here. This being an open world game in the vein of Grand Theft Auto or The Elder Scrolls series, you'll be sitting through a lengthy initial loading screen. The problem is, you'll be waiting up to two minutes to just get to the title screen on a standard PS4, before selecting your save from there, which sets off another lengthy wait. PS4 Pro cuts the initial wait down to 1.5 minutes, Xbox One lands at one minute, 13 seconds - while Xbox One X fastest at just 53 seconds.

This sequence follows if you actually load a game too, but thankfully you don't have to wait quite as long. Frustratingly this means another 28 seconds wait on an Xbox One X, and up to 44 seconds on the worst performer - again, the PS4. As a one-off it's not that bad, given the world is one, seamless sprawl you can walk across, from end to end. However, supposing you use a fast travel system in the game, this is roughly the sort of time-frame you can expect on each console - and that could affect your experience of it.

...

Overall, there's a simple pecking order if you're choosing to play Kingdom Come Deliverance on consoles. Xbox One X delivers the highest resolution - a 77 per cent boost over PS4 Pro, in fact - and smoother performance on top of that. However, between the sub-30fps drops, the obvious texture and geometry pop-in, and long loading times, it's fair to say that no platform is covered in glory here. It could be a simple matter of optimisation through patches that improves this situation - and we hope that these improvements are eventually delivered. Beyond the surrounding furore, Kingdom Come Deliverance actually has some great ideas behind it. It's a decent alternative to the Elder Scrolls, with some amazing backdrops and an interesting historical hook to its story - it's just the technology at its heart that's struggling to keep up. With this in mind, the PC version may be the way forward for those considering the game, and it's something we're planning to look at shortly.

OK, now I can understand why some people on the forum have downplayed the graphics in the game. Looks like the console version is significantly worse visually than the PC version. That's unfortunate as the PC version can be very beautiful at times. Especially when it is raining (so gorgeous) and overcast.

Also, I wonder why both XBO platforms load the game so much faster than the PS4 platforms? I though loading times in games were roughly equivalent between the 2 platforms in games? Have I been thinking the wrong thing this whole time? Do other games also exhibit this behavior?

Regards,
SB

 

[BRiT]

Here's DF's latest article on yet another remake ... http://www.eurogamer.net/articles/digitalfoundry-2018-secret-of-mana-3d-remake-analysis

Secret of Mana's 3D remake is good fun - but won't impress retro purists
PC, PS4, Pro and Vita versions tested.

A genuine Super NES classic, Secret of Mana holds a special place in the hearts of those that played it back in the day. Its blend of role-playing action, gorgeous visual design and evocative music remains a treat even today. The series has persisted across multiple generations since, but the original is still best. Or is it? Last week, Square-Enix released a 3D remake for PS4, PS4 Pro, PC and even PS Vita - and we've played them all.

But what makes this game special? Secret of Mana holds up today and was considered good enough to make the top 21 line-up for SNES mini, and this is perhaps surprising bearing in mind its troubled development. Originally designed for the Super NES's CD-ROM attachment, a collaboration with Sony, the game was originally lined up as a launch title for the Super Disc format - a vast adventure that would take full advantage of the extra disc space rather than simply slapping a few Redbook audio tracks on the disc like many other titles. When the CD project was killed off, Square management pushed the team to complete the game on cartridge instead. The team was forced to cut content and dialogue, reducing the game to the form we know today but despite obvious shortcomings resulting from this process, the game still went on to become a classic.

 

[Scott_Arm]

I thought ps4 load times were typically much faster than Xbox One?

 

[iroboto]

I thought ps4 load times were typically much faster than Xbox One?

i think they usually pretty close, so much so that it's never a measurable benchmark that comes up in DF analysis. Unless it's Pro/X vs the base models.

 

[Tkumpathenurpahl]

I was under the impression that all platforms were pretty similar, except for the X1X, which is consistently much faster?

 

[Recop]

MS did a huge amount of profiling work on real software, tested changes using virtual system configurations and made lots of tweaks to memory access, caches, and per-chip board power delivery to squeeze everything they could out of a market leading but still finite amount of resources.

We can make such claims for any console... what's certain though is that MS chose a design that is much closer to the PS4 for the X...

I thought ps4 load times were typically much faster than Xbox One?

This is true most of the time. As most games tend to run at a higher resolution on PS4. This game is just an exception to the usual rule.

But they aren't much faster though.

Edit : some data from DF about loading times.

http://www.eurogamer.net/articles/d...o-ssd-upgrade-does-sata-3-make-a-difference_9

http://www.eurogamer.net/articles/digitalfoundry-2016-should-you-upgrade-your-xbox-one-with-an-ssd

Last edit : finally the available data don't seem to support my claim and loading times seem to be highly variable according to the game. I don't know why i believed this myth was true... iroboto was right.

http://www.eurogamer.net/articles/digitalfoundry-2014-xbox-one-external-hard-drive-upgrade-guide

http://www.eurogamer.net/articles/digitalfoundry-2014-ps4-hard-drive-upgrade-guide