• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Next-Gen PS5 & XSX |OT| Console tEch threaD

Status
Not open for further replies.

pawel86ck

Banned
And there you go, the reason why 2.5 x is left vague. LOL.

Timdog will be posting about effective 2.5 x 16 GB memory next. MS really need to clean up their act.
It's very simple concept, and they have explained it very well. Nothing is left vague.

This innovation results in approximately 2.5x the effective I/O throughput and memory usage above and beyond the raw hardware capabilities on average. SFS provides an effective multiplier on available system memory and I/O bandwidth, resulting in significantly more memory and I/O throughput available to make your game richer and more immersive.

2.4 GB/s x2.5 multiplier from their quote results in 6 GB/s RAW (12 GB compressed). So Xbox One X or XSX without SFS technology would need to use 6 GB/s SSD in order to achieve the same results. SFS also multiply RAM because the need to load 2.5x less data.

I wonder now if PS5 has similar technology?
 

Handy Fake

Member
It's very simple concept, and they have explained it very well. Nothing is left vague.



2.4 GB/s x2.5 multiplier from their quote results in 6 GB/s RAW (12 GB compressed). So Xbox One X or XSX without SFS technology would need to use 6 GB/s SSD in order to achieve the same results. SFS also multiply RAM because the need to load 2.5x less data.

I wonder now if PS5 has similar technology?
Now no.
 

geordiemp

Member
It's very simple concept, and they have explained it very well. Nothing is left vague.



2.4 GB/s x2.5 multiplier from their quote results in 6 GB/s RAW (12 GB compressed). So Xbox One X or XSX without SFS technology would need to use 6 GB/s SSD in order to achieve the same results. SFS also multiply RAM because the need to load 2.5x less data.

I wonder now if PS5 has similar technology?

Thats not how it works, SFS IF used loads in a smaller texture in RAM which if it gets rendered on screen, system streams in its big brother and blends it in late, so you bring in the late textures by blending rather than pop in a few frames later than they first come into view..

If thats what your talking about, all the time, I dont think thats good as every turn of view will display low textures initially and your handling small then blending big textures in all the time. if that saves 2.5 x memory it is not ideal is it.

Ps5 will just load in high textures straight away lol, does not need a SFS contingency plan.

IF SFS is to cover hickups and late delivery, then that makes more sense to me. Ps5 has been designed around not needing to do that so....
 
Last edited:
PS5 peak is 22 GB/s, so yeah stop BS

Where have they said that's the peak? All I have seen is 5.5GB/s raw and around 9GB/s compressed. I just speculated that this isn't something stated to be constant (like the Xbox) and I believe (like any consumer SSD a person can put in the PS5) that the SSD will very possible throttle itself to some speeds a good amount lower than the 5.5GB/s. And I know that for the most part, the raw performance of the PS5 SSD will be better than the Series X SSD, especially for relatively short things such as initial loading of a level, but I also have to wonder if in a hypothetical consistent SSD data culling situation where the SSD and PS5 system are being taxed hard whether or not the PS5 will throttle to a speed slower or as slow as XSX SSD (which is designed to not throttle) and then on top of that, I am wondering if such theoretical throttling would even be noticeable to the end user compared to XSX. I highly doubt that will be the case, but I felt that since we are discussing the XSX SSD, I would ponder that hypothetical and see if that's even remotely feasible.

It's no different than people wondering if the PS5's higher clocked GPU will be able to outperform the Series X GPU that is following a more wide and slow approach (although I wonder if calling a 1.8Ghz GPU could even be considered slow) compared the PS5's slim and fast approach. It's all just a question of hypothetical technological questions wondering about the different approaches to power the systems are using
 

Thirty7ven

Banned
It's very simple concept, and they have explained it very well. Nothing is left vague.



2.4 GB/s x2.5 multiplier from their quote results in 6 GB/s RAW (12 GB compressed). So Xbox One X or XSX without SFS technology would need to use 6 GB/s SSD in order to achieve the same results. SFS also multiply RAM because the need to load 2.5x less data.

I wonder now if PS5 has similar technology?

Yes it’s called sampler feedback, virtual texturing etc and that tech will keep improving.

SFS only difference is in preventing artifacts that might show up by using hw filters. Considering that the problem they are addressing Is how fast and just in time loading of a texture, and considering what we saw from U5, the PS5 speed seems more than enough to address the needs of virtual texturing in a way that prevents pop in and artifacts.

Also this is just textures. It only applies to textures. And much like Bcpack, alternatives exist and will continue to evolve.
 
Last edited:

pasterpl

Member
The foundation of the Xbox Velocity Architecture is our custom, 1TB NVME SSD, delivering 2.4 GB/s of raw I/O throughput, more than 40x the throughput of Xbox One. Traditional SSDs used in PCs often reduce performance as thermals increase or while performing drive maintenance. The custom NVME SSD in Xbox Series X is designed for consistent, sustained performance as opposed to peak performance. Developers have a guaranteed level of I/O performance at all times and they can reliably design and optimize their games removing the barriers and constraints they have to work around today.

i found this interesting, seems like consistent performance is key for ms with xbsex. Just wondering how cooling will compare between xbsex and ps5. It almost like ms knows something about ps5.
 
D

Deleted member 775630

Unconfirmed Member
Exactly, 2.5 x less memory used than the last gen, 40 x faster than last gen. Fine. Thats how I read it as well.

WE agree.

You got posters in here saying 2.5 x 4.8 Gbs - so it was not clear was it.

2.5 x what was already stated for XSX is now the meme - go read above, there are plenty.

Its like Cerny saying ps5 100 x faster then being unclear - he said over ps4 HDD.
I understand why people say 2.5x4.8GB/s though. The idea is that compared to the old systems on top of the read speed of 4.8GB/s you now also don't need to read as much data, more concretely 2.5x less data into memory. So if you would have a scene the old system could load 1.6GB (4.8/3) into that scene in a second, while the new system is able to load the complete 4.8GB into that scene.

IF, big if because I just don't know, Sony wouldn't have a similar solution in the PS5, that 8-9GB/s should also be divided by 3, because the GPU won't use all of it, which would actually be only 3GB of textures in a scene. But again, I don't know if Sony has something similar to SFS.
 

geordiemp

Member
I understand why people say 2.5x4.8GB/s though. The idea is that compared to the old systems on top of the read speed of 4.8GB/s you now also don't need to read as much data, more concretely 2.5x less data into memory. So if you would have a scene the old system could load 1.6GB (4.8/3) into that scene in a second, while the new system is able to load the complete 4.8GB into that scene.

IF, big if because I just don't know, Sony wouldn't have a similar solution in the PS5, that 8-9GB/s should also be divided by 3, because the GPU won't use all of it, which would actually be only 3GB of textures in a scene. But again, I don't know if Sony has something similar to SFS.

Because Sony dont need SFS as they designed their IO system to be insanely fast. Different solutions.

I have given this more thought, SFS IF USED is lower quality textures / Mips in memory and if its rendered brings in the high quality one and blends in. This is a longer pipeline so there is a cost, but that would use less instantaneous memory,

It would also mean every time you turn the camera, low textures appear and blend into higher ones a few frames later - a blended pop in that takes cycles. Do you think MS will use this as the new norm ?

I dont see how that is performant, its a longer pipe line for rendering...and if you constantly moving camera view everything would be low res ? I cannot get my head around why...

And cant imagine its as good as having the right texture ready in the first place and prefetching the next few seconds of gameplay.

SFS sounds like a backup plan if MS did not get their IO speeds up (BUT THATS JUST MY OPINION). I just thought of SFS as a safety net if a high res texture arrived late, but maybe not, who knows.
 
Last edited:
Xbox make advertisement with good taste .... did this really happen? I loved the music
LOL wut?

That's saying the STREAMING POOL for the demo is 768MB.

Meaning they used upwards of (TOTAL RAM - 768MB) to render what is seen on screen at any given moment. For one single "view" that's probably 12GB of RAM, with compression some amount under that on disk.

The demo was likely MASSIVE.

The tech will certainly be used, but it's going to be extremely difficult to truly push the limits of the PS5 I/O often. As pushing those limits for detail require a lot of disk space.
Maybe I am misinterpreting something but if they talking of memory pool for streaming still many data 768 MB I mean I current game like Killzone Shadow Fall use around 1.5 GB.

That talk of how much more data this engine with new SSD streaming era is capable to feed our systems.

suMW5Zg.jpg

Also something I read in this forum and era is the people think the bandwidth of this consoles is only for streaming the assets from the
SSD but also the SSD is working like some kind of ram (we see this already so is not weird), so that means you will read that data much
frequently that we use to do now, so is not like you need 5 GB of data of the current session be load in one second probably you only
need another small amount of data but should be as fast is possible so here is where the bandwidth matters.
 

Vae_Victis

Banned
I'm still deeply confused by that 100GB thing.

Through the massive increase in I/O throughput, hardware accelerated decompression, DirectStorage, and the significant increases in efficiency provided by Sampler Feedback Streaming, the Xbox Velocity Architecture enables the Xbox Series X to deliver effective performance well beyond the raw hardware specs, providing direct, instant, low level access to more than 100GB of game data stored on the SSD just in time for when the game requires it.

What does that mean? How do they get to that number? If the maximum transfer speed is comfortably below 10 GB/s, wouldn't you still need more than 10 seconds to access 100GB of data? Isn't it a bit of a stretch to call 10+ seconds "instant"?

Is this all just a meaningless figure of speech to say "we can read a game from the Drive faster than we used to"?
 

pawel86ck

Banned
Thats not how it works, SFS IF used loads in a smaller texture in RAM which if it gets rendered on screen, system streams in its big brother and blends it in late, so you bring in the late textures by blending rather than pop in a few frames later than they first come into view..

If thats what your talking about, all the time, I dont think thats good as every turn of view will display low textures initially and your handling small then blending big textures in all the time. if that saves 2.5 x memory it is not ideal is it.

Ps5 will just load in high textures straight away lol, does not need a SFS contingency plan.

IF SFS is to cover hickups and late delivery, then that makes more sense to me. Ps5 has been designed around not needing to do that so....
You say it's not how it works, but I have quoted their engineer saying SFS multiply effective I/O throughput and memory usage. Without SFS they would need 6 GB/s SSD (2.4 GB/s x2.5) in order to achieve similar I/O throughput. If you think MS engineer is lying then that's fine, but that's what he said.
 
Last edited:
D

Deleted member 775630

Unconfirmed Member
Because Sony dont need SFS as they designed their IO system to be insanely fast. Different solutions.

I have given this more thought, SFS IF USED is lower quality textures / Mips in memory and if its rendered brings in the high quality one and blends in. This is a longer pipeline so there is a cost, but that would use less memory,

It woudl also mean every time you turn the camera, low textures appear and blend into higher ones a few frames later - a blended pop in. Do you think MS will use this as the new norm ? I dont see how that is performant, its a longer pipe line for rendering...

I just thought of SFS as a safety net if a high res texture arrived late, but maybe not.
Well if Sony doesn't have this, then their GPU would only use 1/3 of the textures streamed to there. So like I said that 8-9GB turns into 2.6-3GB, which is less than 4.8GB of the XSX.

On the mipmaps in memory. You are correct that this is mostly for the lower quality textures in the distance, BUT because they only get the low quality textures there is a lot more room for high quality textures that are close by in the scene. I don't think this means turning the camera would mean you first see low quality and then suddenly the high quality pops-in because they specifically say "on demand, just in time", so you don't notice this.
 

But never say was not realtime, we say is so heavy that is not even close to be 4k and not is even in 30 fps and
is just running in a "similar" PC as the XSX.

Of course was realtime who will render offline a trailer just to hit that resolution and framerate.

They say a million triangles occupies the same amount as a 4k Texture. Wish they just come out and address this so we can move on from the concern.
But why if your a regular user you will not understand of what they are talking about, I saw people trying to use anything Epic says for the console war. Like the people
who see "running" the demo in the viewport of the UE and saying look is running and is not in PS5, for me this last people just have brain damage.

If you are a indie dev or from a modest studio you still working in pass your project (removing issues) from 4.24.3 to 4.25.1 a for the time of release of UE5 you cannot
hold your project so much time.

In the last scenario you are from a important AAA or AA studio and you already know more details but probably are under NDA or your work there you don't need to
know now.
 
Last edited:

sircaw

Banned
2.4 GB/s * 2.5 = 6 GB/s

Microsoft’s Xbox Series X SSD is fast, and can use similar hardware-based compression, but its speeds are 2.4GB/s, or 6GB/s with compression.


Does that article say BRUTE FORCE.
 
Because Sony dont need SFS as they designed their IO system to be insanely fast. Different solutions.

I have given this more thought, SFS IF USED is lower quality textures / Mips in memory and if its rendered brings in the high quality one and blends in. This is a longer pipeline so there is a cost, but that would use less memory,

It woudl also mean every time you turn the camera, low textures appear and blend into higher ones a few frames later - a blended pop in. Do you think MS will use this as the new norm ? I dont see how that is performant, its a longer pipe line for rendering...

I just thought of SFS as a safety net if a high res texture arrived late, but maybe not.

What are you doing here Geordie??? Not like this bro. Not like this. You have to dumb it down & keep it simple. Leave this loop convo. There's no talking points. Xbox cant compete posting this 2 months later.SSD talk was May. Its over.They cant compete 6 Levels of priority ,super low latency, kraken ,oodle. Its over. they marketed loosing points 2 months after defeat. Leave this loop convo. Keep it simple like this.



 
Last edited:

THE:MILKMAN

Member
So when I parse out all the jargon and acronyms around SFS it boils down to it meaning less RAM data is parked in memory at any one time? Just like with PS5 as explained in the below screenshot:

ETZ4EqLWoAIwqmI.jpg:large


So with XSX it = ~2.5 times efficiency of RAM and on PS5 it is ~5-6 times RAM?
 
Last edited:

Darklor01

Might need to stop sniffing glue
I have recently acquired an internal memo to XBOX Marketing. It shows what happens when you have achieved full marketing penetration.
0PS5Ni3H5wfWf4AyJFyEtHpK8UBoqyVObBrVdq3CwXygyRdFF0QLKkfrU8nbNXUaPqecue7jiZ9p_maf5pU--3xxA40NLrjq4X2tEMaU2tglrOm3jBQyKml7r6f_rmvk9F2j2djvSuoBNG-Au0ZTy6HNaOK4fZYBUJYjO4SYsca4qAIWYsLJhcNErC0nhxAqaZ31soVQQ9WwKhE6NFNo5CuDl_CMofRlAEU8GwTpmQx-ok_FcimAsoulppmq4Lw-riHQZbjCBFPX9XH-TktDLTL6iAq4bh0KD5JnxUkfDqADi9-_Oj_UAaTVVKwLFh50qcxT76C-3eZrbQKcqAvBrfcDV0-qGHAAhOasWap8-vr6mm5MoDLhl_NLE4d3uOve3S8NYQAHe5GNcwGwO2CkCK7jOtCaz-9qLlq0XqMvsW31LSgMRnHVtUa5USzGRdkO6uK7pF3-fmQiTuNt1sk9F62oXN8Tgm4jAe9LI5DUB7UAHKSIPMwIXgSE4sLugPUL-pz59L0jGlbsJMLwrZkzczsmT2AzVg8mbl-w_kdZ33xxtSS7pGRQxL6TFySDO4UMt-nzgVM8nP0UKYyTYh75LillzcypwbSfJ_QqsCtyRhN0bQZRYzYBFP4ITB93Cc-a18JbLGUGP-wEpKyHLj8GBso4otqRfSbsSOLNb6NCtBUuppv3YQbo8bFyOzGN=w1229-h1207-no


Marketing is dangerous. Very dangerous.

(For the record, I hope to own both systems at some point ;) )
 

Thirty7ven

Banned
Well if Sony doesn't have this, then their GPU would only use 1/3 of the textures streamed to there. So like I said that 8-9GB turns into 2.6-3GB, which is less than 4.8GB of the XSX.

On the mipmaps in memory. You are correct that this is mostly for the lower quality textures in the distance, BUT because they only get the low quality textures there is a lot more room for high quality textures that are close by in the scene. I don't think this means turning the camera would mean you first see low quality and then suddenly the high quality pops-in because they specifically say "on demand, just in time", so you don't notice this.

This isn’t true. You can still do the same for PS5 or PC, just load a lower res texture and then when it gets closer you load the normal higher res version. MS answer to lack of bandwidth are HW filters that eliminates pop in/artifacts in that process. But pop in/artifacts are a consequence of speed or lack thereof.
 

geordiemp

Member
Well if Sony doesn't have this, then their GPU would only use 1/3 of the textures streamed to there. So like I said that 8-9GB turns into 2.6-3GB, which is less than 4.8GB of the XSX.

On the mipmaps in memory. You are correct that this is mostly for the lower quality textures in the distance, BUT because they only get the low quality textures there is a lot more room for high quality textures that are close by in the scene. I don't think this means turning the camera would mean you first see low quality and then suddenly the high quality pops-in because they specifically say "on demand, just in time", so you don't notice this.

Discussing not loading into RAM whats not in view, both consoles can do that anyway. Cerny has already said the ps5 new geometry engine culls before calculation, and we saw UE5 only load and display whats in view. So nothing new here.

How to handle LODs efficiently when in view, XSX loading small then big is not as efficient as just loading big, thats clear.

Anyway, UE5 Nanite wont use LODs so if that tech takes off then ..........optimising LOD transitions will be a thing of the past,.
 
Last edited:

Lethal01

Member


They are taking about saving memory amount by not loading textures that are not used in the older way of streaming data.

It does not make streaming quicker. The mental gymnastics is quite amusing.


Obviously not literally, but effectively. If you only need to load half the amount of textures you effectively have the same effect as streaming at twice the speed.
 

IntentionalPun

Ask me about my wife's perfect butthole
Discussing not loading into RAM whats not in view, both consoles can do that anyway. Cerny has already said the ps5 new geometry engine culls before calculation, and we saw UE5 only load and display whats in view. So nothing new here.

How to handle LODs efficiently when in view, XSX loading small then big is not as efficient as just loading big, thats clear.

Anyway, UE5 Nanite wont use LODs so if that tech takes off then ..........optimising LOD transitions will be a thing of the past,.
Not using LOD uses more RAM though.

Not like that can just be ignored completely. PS5 does not have more RAM than XSX, but does have more bandwidth to pull data in faster meaning more can be used for what is on screen but with a lot of far away objects in a scene they might go beyond that and still need to do LOD.
 

Brudda26

Member
Jesus SFS is part of the compression pipeline that is where the 6GB/s number from the engineer comes from. In absolute perfect scenario that's the absolute limit no more than that. The multiplier is for the raw speed. Not the compressed speeds as SFS is part of the compression pipeline its how its achieving that 6GB/s.
 
oh shit!!!
Are you AegonSnake ? from the other place
unknown.png


You heard it here first guys, 12Gb/s average. Fastest SSD in the west.
So now are they using the theoretical limit of 6GB/s ?Just wow. :lollipop_neutral: (2.4*2.5=6)
Yes, but that means that normally textures are being loaded and only 1/3rd is being used (according to Microsoft). So if you wouldn't have SFS you could load 4.8GB/s of compressed assets (in this case textures) and only 1.6GB (1/3) would be used by the GPU. While now the full 4.8GB will be used. Or on average 2.5x as much as we would normally be able to load in.

In easier terms. An old system not supporting SFS would actually need an SSD with 12GB/s compressed read speeds to load in the same amount of textures as the XSX with SFS and an SSD with 4.8GB/s compressed read speeds.
We don't live in a world without Tiled resources/Partially Resident Textures so your example is not real is this year and not even in almost the last decade.

They cannot use SFS for every time they want, they said that, we already had a discussion of this, SF is an improvement of Tiled resources which already
exists in Xbox one. You are dev accordingly to you so this not should be a strong enough argument to reach your conclusion of 12 GB/s.
unless you care about framerate and resolution, then xbsex seems better choice
If you use the theoretical TF as only argument the delta will close of 2160 vs 1920 more of less so yeah could exist but I don't see it as deal breaker.

The framerate you don't know this is even more complicated than the resolution because could be limit for the GPU (any of its parts), CPU, SSD, API
this consoles are closed enough to not see a game running to 60 vs 30 fps and even is possible we see games perform better on one console and then
be the opposite.
 

sircaw

Banned
You know, seriously speaking here, its nice to see some new details released.

Its nice to see the tech people who have been starved of information being able to spar with one another.

This forum at it's best.
As for it's worst that's why i am here. Patrolling.🦈
 

pawel86ck

Banned
Dude, it doesn't work that way, VA results, with everything included, is 2.4(raw)/4.8(compressed)/6.0(theoretical max). They've been saying these numbers since the beginning.
Jason Ronald has specifially mentioned SFS multiply raw hardware capabilities. Are you suggesting MS have already taken into account SFS gains into 2.4 GB/s SSD speed? If that's the case their true SSD raw speed would be 960 MB/s.
 
Last edited:
It's very simple concept, and they have explained it very well. Nothing is left vague.



2.4 GB/s x2.5 multiplier from their quote results in 6 GB/s RAW (12 GB compressed). So Xbox One X or XSX without SFS technology would need to use 6 GB/s SSD in order to achieve the same results. SFS also multiply RAM because the need to load 2.5x less data.

I wonder now if PS5 has similar technology?
In which year do we live? 2005/2006 when we load all the textures.

What is next we load all the models from a scene with the same level of detail.
What are you doing here Geordie??? Not like this bro. Not like this. You have to dumb it down & keep it simple. Leave this loop convo. There's no talking points. Xbox cant compete posting this months later.SSD talk was May. Its over.They cant compete 6 Levels of priority ,super low latency, kraken ,oodle. Its over. they marketed loosing points 2 months after defeat. Leave this loop convo. Keep it simple like this.




Not please not. Please guys think dont just quote fanboys from Twitter.
 
Last edited:

Brudda26

Member
Jason Ronald has specifially mentioned SFS multiply raw hardware capabilities. Are you suggesting MS have already taken into account SFS gains into 2.4 GB/s SSD speed? If that's the case their true SSD raw speed would be 960 MB/s.
Yes hardware capabilities, compression is part of the hardware capabilities. SFS is part of the compression pipeline and is how they can achieve in the ideal scenarios 6GB/s max
 

pawel86ck

Banned
In which year do we live? 2005/2006 where we load all the textures.

What is next we load all the models of a scene with same level of detail.

Not please not. Please guys think dont just quote fanboys from Twitter.
Jason Ronald has explained In the article what's the difference between standard mipmaps and their SFS implementation.

these larger mipmaps require a significant amount of memory compared to the lower resolution mips that can be used if the object is further away in the scene. Today, developers must load an entire mip level in memory even in cases where they may only sample a very small portion of the overall texture. A single scene often includes thousands of different textures resulting in a significant loss in effective memory and I/O bandwidth utilization due to inefficient usage. With this insight, we were able to create and add new capabilities to the Xbox Series X GPU which enables it to only load the sub portions of a mip level into memory, on demand, just in time for when the GPU requires the data. This innovation results in approximately 2.5x the effective I/O throughput and memory usage above and beyond the raw hardware capabilities on average.

If PS5 doesn't have similar feature then we have a new winner in I/O throughput 🤣
 

Neo Blaster

Member
Jason Ronald has specifially mentioned SFS multiply raw hardware capabilities. Are you suggesting MS have already taken into account SFS gains into 2.4 GB/s SSD speed? If that's the case their true SSD raw speed would be 960 MB/s.
No, I'm suggesting they are applying this 2.5x multiplier to the raw SSD speed, so 2.4 x 2.5 equals 6, exactly the theoretical max they already disclosed in their specs.
 

pawel86ck

Banned
Yes hardware capabilities, compression is part of the hardware capabilities. SFS is part of the compression pipeline and is how they can achieve in the ideal scenarios 6GB/s max
Are you suggesting 4.8-6GB/s decompression speed given by MS already takes into account SFS gains (2.5x)? If that's the case then XSX SSD raw speed and compression must be really bad. For now I'm more convinced Ronald was talking here about SSD 2.4 GB/s raw speed multiplier.
 
Last edited:
This exactly this, everything mentioned today is the exactly the same stuff we heard months ago just different.

Bro thats whats sad about it all. People still running with it. The convo is not serious & its not a good read anymore. This shit was over in MAY. And discussed fully in here. It was over 2 months ago. And people in still here talking about SSD's.

You can only banta & troll at this point. Yeah I think Im tapping out of here after August show. The tear down is not worth talking points. Its either 3D stacked chips or not. Wont matter after that all details are out. Only the next GDC will reveal details.
 
Last edited:

DaGwaphics

Member
People can be so sensitive to marketing words. I can imagine the reactions to dumping that packet of instant oatmeal into a bowl.

giphy.gif


Or watching a show where every five minutes a commercial advertises a best, most reliable, most powerful, best value family sedan by a different maker.
 
Status
Not open for further replies.
Top Bottom