Who cares about Xbox One Slim, Not Slim or Scorpio if you have a PC you don't need any of them.
This thread is worth it though for casharmy's full-on spaz attack we are witnessing him melt right down into the f***ing ground before our eyes. Excellent stuff.
Gotta love those PC people always inserting themselves into a conversation because almost LITERALLY no one cares. Always so insecure. Play your PC games and be happy.
I've noticed that too. You cannot have a conversation on this board about consoles without some PC zealot coming in and going full infomercial on all the benefits of PC gaming. Lol.
@Wickerman777: lol saying 4K 60 fps uncompressed no compromise.........I guess they be stuffing in a 9TFLOPS gpu in scorpio and calling it 6TFLOPS like on them xfx boxes....cause that be only way they do 4K 60 fps no compromise....
I wonder if it turn out like the xboner 720..............it only do 720p hahahaha. on ghosts and BF4
the xbone one two......
you watch they name it 10.......just to beat sony.
You're not the first one I've seen here play around with their words. They said something about 4k in one part and then something about 60 fps in another. They never said 4k/60. But you know that, of course.
@Wickerman777: dude phils own E3 words.....scorpio will do 4K 60fps UNCOMPRESSED NO COMPROMISE.....note the no compromise words....but even if he never said no compromise or 60FPS or uncompressed.......he said 4K and at 6 TFLOPS impossible...cards at 7 like the fury or 8 like the fury X and nano have a hard time doing a solid 40 in 50% of games that be needed to make 30 fps fluid.....and that means NO FIRST PERSON SHOOTERS AT ALL.....
again we see people try to say it do 4K then say well they never said that or back pedal down the hill....
PHIL EVEN SAID IF YOU DONT OWN A 4K TV DONT BUY SCORPIO>>>BUY THE ONE S.
so what it is it you xbox fans don't grasp here....that it wont cost 600-700 bucks or it somehow do 4K but normal games don't count only indies...or you somehow think in 1 year AMD will make more magic then Nvidia did with the 1070 ....we know Nvida tflops tend to do better in games then AMD ....granted DX 12 is AMD game of thrones but there not going to pull rabbits out of hats for a company that only sold 10 -15 million consoles vs sony solid 40......
what part of that don't you grasp......that sony and only sony would be in a stronger position most of all knowing AMD released vulkan a direct rival to Microsoft direct X....
you people seem to be caught up in this halo fanasty of xbox promises.....where is the halo series?
i seen the same exact boondoggle with xbox one before it was released....how is this any different this time around.
how many back pedals we seen now from E3 ...
I link another one....http://www.dsogaming.com/news/microsoft-backtracks-exclusives-now-claims-e3-2016-games-will-coming-windows-10/
more prime examples of what they said is not what they said...
oh so somehow phil will tickle me elmo the game developers into spending extra time and money of what the pc master race is begging for .......a bailout for the 1%.......
you have a better chance of asking congress for a bailout..lol
I doubt they put the effort into an area that to be honest wont help them sell a game you will already buy,,,infact if you look at what everyone doing on steam 60% of people are doing weak 1080 not even the pc master race with all there crap talk can claim there really doing it that much better then us on consoles...
4K tv are rare and so is 4K gaming...its an elite class so that to me says a lot where they will target there buget....1080 max and 1440 mid...that good enough to sell GTA 5 to you for 60 bucks is it not....3 years later...infact I hear in 2k it not much better looking ....I asked because to me I want neo to hit 1440 on GTA5.....but it be all about money for the developers.....
unless phil gone Chinese and plans to graft the entire industry with chips for macau I don't see it going to happen.....a nudge wont help ...you going to have to tug that bull by the nosering.......lets be clear........money talk ..bs walks...
unless ms plans on xbox scorpio branded titles and they charge 10 bucks more for GTA 5 or BF1 titles and they sell 10 million units in less then 2 years no one going to waste any time more then they need to on 4K......
that's the only way your going to move developers ...with 200 million more on a 1 billion doller title....
look at how things are now......
the division was what.....epic fail.....starwars battle front.....epic money gaff and fail.
these guys get by now passing slop off on the table we eat off of......how are you going to rain them all in?
I don't mean to sound crude here but there is the way things should be ....and the way it is......getting anything from dice without cash is hard enough now....look at the games they made and EA....look how they gaff you for a DLC.....look how they charge you for weapons etc upgrades so on.....you think they give 4K away for free when it cost them extra effort.
dude I so remember dec 2013 on BF4 the epic crash reports...servers,saves,,,crash stuck in wall....FOR MONTHS after this went on.....now they going to pull the 4K rabbit out there hat because phil asked nice?
unless trump is president and threatens them with extra water at gitmo I don't see it going to happen.....not unless you pay more for games a lot more...if they make 1 billion bucks on a sht ass launch what makes you think there going to change over night with UMP and some sweat talk......
you really think phil will make a wonder wall with game developers?
wow I so wanna buy that....oh wait I have a PS4......but mine in black...wait I run out and buy one so I can game at 900P or some oddball 1440 x 1080P size that only cheap digital cams use.
you know everyone going to coin it the xbox one sht right......and remember you heard it here first......from me....
@casharmy: Clueless analyst. If you're going to get a Scorpio, there's really no reason to get a XB1 Slim immediately. Obviously the XB1 Slim will replace the bulky original XB1. Eventually the XB1 S price will drop to under $300. When that happens it will be a good time to get an XB1 S. Of course if you don't own an XB1 at all, get the XB1 S now. Then you can trade it for the Scorpio 1-2yrs later.
wow I so wanna buy that....oh wait I have a PS4......but mine in black...wait I run out and buy one so I can game at 900P or some oddball 1440 x 1080P size that only cheap digital cams use.
you know everyone going to coin it the xbox one sht right......and remember you heard it here first......from me....
@casharmy: Which perspective do you want us to comment on? The report of the DFC analyst? Or the perspective of the two commentators who destroy the analyst's argument and literally describe the analyst as a fanboy?
wow I so wanna buy that....oh wait I have a PS4......but mine in black...wait I run out and buy one so I can game at 900P or some oddball 1440 x 1080P size that only cheap digital cams use.
you know everyone going to coin it the xbox one sht right......and remember you heard it here first......from me....
XBONE Shit model.
This sums up just about every one of your posts newbie/alt
When Scorpio hits the streets most of you Cows will be flooding the streets with your butthurt angry dripping tears of fear, sadness and more butthurt jelly
I'll buy an xbone S. And then if the Neo is something I'm excited for when it comes out I'll buy that too. And after that if the Scorpio comes out and it's interesting and worth the money i'll get one of those. See how easy life is if you don't become a zombie for Sony of Microsoft? It's not that hard. No one needs to be butthurt. Just don't be an idiot.
only a fanboy such as yourself would think that we are talking about the size of the X1 slim vs the original PS4. I was simply pointing out the fact that there will almost certainly be a slim version of the PS4. look at the photo i provided, the Ps2 slim was way thinner than the original. you implied there was no need for a PS4 slim as it was already slim, but so was the PS2 (and slimmer than OG Xbox For that matter) and look what happened there.
stay mad you big baby.
@casharmy said:
@clone01 said:
@sts106mat said:
@casharmy said:
lol isn't that a oxymoron?
If you own a PS4 the "slim" is already sitting right there! If you don't own a PS4 the "slim" is available where every PS4s are sold.
durrrr there will almost certainly be a slimmer version of the PS4
Cashangry isn't all that bright. While I think the XBox one is fugly, the PS4 isn't that much better. If he'd get his nose out of sony's ass, he'd realize that, too.
Here is Xbone S (the slim xbone)
Here is PS4 BASE model
Man you xbox clowns are at the bottom of the barrel when it comes to arguments and intelligence. Don't talk about being "bring" when you can clearly see the "slim" model of xbone is STILL bigger than the PS4 dimwit.
not really the S its wider but ps4 its much chukyer
and front to back ps4 is actualy same as xbox one if not longer
lol I love how boners get suked in every time....best was when tweek town ran that 10 TFLOPS scorpio story it was 5X neo power and they all licked there chops only to find out the dogs are not getting the baby back rib bones...
news flash ...just in xbox 1080 folds space and time........more to come.
@Wickerman777: dude phils own E3 words.....scorpio will do 4K 60fps UNCOMPRESSED NO COMPROMISE.....note the no compromise words....but even if he never said no compromise or 60FPS or uncompressed.......he said 4K and at 6 TFLOPS impossible...cards at 7 like the fury or 8 like the fury X and nano have a hard time doing a solid 40 in 50% of games that be needed to make 30 fps fluid.....and that means NO FIRST PERSON SHOOTERS AT ALL.....
again we see people try to say it do 4K then say well they never said that or back pedal down the hill....
PHIL EVEN SAID IF YOU DONT OWN A 4K TV DONT BUY SCORPIO>>>BUY THE ONE S.
so what it is it you xbox fans don't grasp here....that it wont cost 600-700 bucks or it somehow do 4K but normal games don't count only indies...or you somehow think in 1 year AMD will make more magic then Nvidia did with the 1070 ....we know Nvida tflops tend to do better in games then AMD ....granted DX 12 is AMD game of thrones but there not going to pull rabbits out of hats for a company that only sold 10 -15 million consoles vs sony solid 40......
what part of that don't you grasp......that sony and only sony would be in a stronger position most of all knowing AMD released vulkan a direct rival to Microsoft direct X....
you people seem to be caught up in this halo fanasty of xbox promises.....where is the halo series?
i seen the same exact boondoggle with xbox one before it was released....how is this any different this time around.
how many back pedals we seen now from E3 ...
I link another one....http://www.dsogaming.com/news/microsoft-backtracks-exclusives-now-claims-e3-2016-games-will-coming-windows-10/
more prime examples of what they said is not what they said...
What you don't grasp is FLOPS has very little to do with AMD and NVIDIA GPU frame rate difference i.e. effective memory bandwidth comes into play when there's enough FLOPS to particular workload.
As for RX-480, any overclock editions will be bounded by effective memory bandwidth.
For example
((256 bit x 9000Mhz) / 8) x Polaris's 75.79 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 297 GB/s
For reference RX-480
((256 bit x 8000Mhz) / 8) x Polaris's 75.79 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 264 GB/s
--
Scorpio's 320 GB/s memory bandwidth
((384 bit x 6700 Mhz) / 8) x Polaris's 75.79 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 333 GB/s
Comparison.
The memory bandwidth gap between Fury X and R9-290 = 1.266X (random textures)
With Fury X, it's memory compression is inferior to NVIDIA's Maxwell.
The FLOPS gap between Fury X and R9-290 = 1.48X
The frame rate gap between R9-290X and Fury X is 1.19X.
Random texture memory bandwidth gap's 1.266X factor is closer to frame rate gap's 1.19X.FLOPS gap between R9-290X (5.8 TFLOPS)and Fury X (8.6 TFLOPS) plays very little part with frame rate gap.
With 980 Ti (5.63 TFLOPS), it's superior memory compression enables it to match Fury X's results.
Conclusion: When there's enough FLOPS for a particular workload, effective memory bandwidth is better prediction method for higher grade GPUs.
-------------------
@wildaries said:
cards at 7 like the fury or 8 like the fury X and nano have a hard time doing a solid 40 in 50% of games that be needed to make 30 fps fluid.....and that means NO FIRST PERSON SHOOTERS AT ALL.....
That's FALSE. A single step down in graphics settings can easily enabled 4K gaming with minimal visual quality lost.
@wildaries said:
what part of that don't you grasp......that sony and only sony would be in a stronger position most of all knowing AMD released Vulkan a direct rival to Microsoft direct X....
For Xbox One, AMD and Microsoft co-developed DirectX12 e.g. DirectX12's Asynchronous Compute was modeled after AMD GCN's ACE units.
As for RX-480, any overclock editions will be bounded by effective memory bandwidth.
For example
((256 bit x 9000Mhz) / 8) x Polaris's 75.79 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 297 GB/s
For reference RX-480
((256 bit x 8000Mhz) / 8) x Polaris's 75.79 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 264 GB/s
--
Scorpio's 320 GB/s memory bandwidth
((384 bit x 6700 Mhz) / 8) x Polaris's 75.79 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 333 GB/s
Comparison.
lol So you still claiming Scorpio will have 320GB/s just for its GPU.? Hahhahahahaahaa
So how good were you arguments about ESRAM matching the PS4 memory,how good were you endless quoted of Rebellion on false ESRAM claims.?
Dude get this over your head once and for all,those 320GB/s are on a single POOL and are shared with the CPU and system,so regardless of having 320GB/s a good chunk will be for system and CPU,and the rest will have the same pitfalls any other memory have on effectiveness.
Memory compression on GCN has shit to do with how the CPU and system handle their portion of the bandwidth assign to it,and just like every other bandwidth it will be lower than the peak bandwidth available,again the Scorpio will perform like a RX480 more or less,just like the xbox one perform like a 7770 more or less.
@panda30 said:
not really the S its wider but ps4 its much chukyer
and front to back ps4 is actualy same as xbox one if not longer
lol those screen are on purpose compare in different ways to make the xbox one S look smaller.
First screen the XBOS is not lined all the way to the left side you can clearly see it..
The second screen you see the PS4 lined to the millimeter with the xbox one left side.
The 3rd screen show them side by side in a weird angle,so you can't make a correct measurement.
Yet the 4th screen show both the PS4 and XBO facing each other,those comparison are made to order and done to showcase the xbox one s as being smaller than the PS4 which apparently is not.
lol So you still claiming Scorpio will have 320GB/s just for its GPU.? Hahhahahahaahaa
So how good were you arguments about ESRAM matching the PS4 memory,how good were you endless quoted of Rebellion on false ESRAM claims.?
Dude get this over your head once and for all,those 320GB/s are on a single POOL and are shared with the CPU and system,so regardless of having 320GB/s a good chunk will be for system and CPU,and the rest will have the same pitfalls any other memory have on effectiveness.
Memory compression on GCN has shit to do with how the CPU and system handle their portion of the bandwidth assign to it,and just like every other bandwidth it will be lower than the peak bandwidth available,again the Scorpio will perform like a RX480 more or less,just like the xbox one perform like a 7770 more or less.
GO BACK TO PRIMARY SCHOOL TO STUPID CLOWN.
Polaris' memory compression is better than Fury's version which enables RX-480 to rival R9-290X/390/390X. I have already factored in Polaris' memory controller efficiency of 75 percent. R9-290X has memory controller efficiency of 82 percent.
YOU ARE A FOOL since you haven't factored in PCI-E 16X version 16 link (16 GB/s+16 GB/s=32 GB/s) from the CPU's side. PS4's CPU-to-memory link is only limited 20 GB/s.
Using PS4's worst case memory bandwidth efficiency of 65 percent
((384 bit x 6700 Mhz) / 8) x PS4's 65 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 284 GB/s
Scorpio's 320 GB/s memory bandwidth
((384 bit x 6700 Mhz) / 8) x Polaris's 75.79 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 333 GB/s
Furthermore, MS has stated "more than 320 GB/s memory bandwidth".
Btw, Intel AVX2's gather instructions has GPU like gather behavior which makes memory data fetch efficient i.e. a single instruction that can read/fetch multiple data from different locations, hence reducing individual load instructions i.e. that's reducing memory bandwidth usage by reducing instruction count.
Polaris' memory compression is better than Fury's version which enables RX-480 to rival R9-290X/390/390X. I have already factored in Polaris' memory controller efficiency of 75 percent. R9-290X has memory controller efficiency of 82 percent.
YOU ARE A FOOL since you haven't factored in PCI-E 16X version 16 link (16 GB/s+16 GB/s=32 GB/s) from the CPU's side. PS4's CPU-to-memory link is only limited 20 GB/s.
Using PS4's worst case memory bandwidth efficiency of 65 percent
((384 bit x 6700 Mhz) / 8) x PS4's 65 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 284 GB/s
Scorpio's 320 GB/s memory bandwidth
((384 bit x 6700 Mhz) / 8) x Polaris's 75.79 percent memory bandwidth efficiency) x Polaris's compression booster 1.36X = 333 GB/s
Furthermore, MS has stated "more than 320 GB/s memory bandwidth".
Btw, Intel AVX2's gather instructions has GPU like gather behavior which makes memory data fetch efficient i.e. a single instruction that can read/fetch multiple data from different locations, hence reducing individual load instructions i.e. that's reducing memory bandwidth usage by reducing instruction count.
"Read Memory Caching and PS4" section on reducing memory access usage.
WTF man is noty fu**ing memory compression alone what make the RX480 perform like a R390X what the fu** are you smoking Polaris has many improvements over GCN is not just memory compression.
NEW geometry processor.
NEW 4TH GENERATION GRAPHICS CORE.
NEW COMMAND PROCESSOR
NEW L2 CACHE.
NEW DISPLAY ENGINES
NEW MULTIMEDIA CORES.
NEW MEMORY CONTROLLER.
WTF man you are trying to make see that only memory compression is the only thing improving Polaris,you are a joke and a MS ass kisser of epic proportions.
The only thing i never fact is your bullshit theories on how system and CPU will run with absolutely ZERO 0000000 bandwidth as by your pathetic ass argument all the 320GB/s are for GPU and nothing for system this is pathetic you are a blind biased lemming who simple can't admit being wrong and want to argue shit that can't be argue..
I just have this question for you if you DARE answer it without fu**ing hiding or spinning things like you always do.
HOW MUCH BANDWIDTH SCORPIO CPU AND SYSTEM WILL HAVE ACCORDING TO YOU.
You did this already as well with the xbox one,claiming 68GB/s + 104GB/s from ESRAM and i corrected you and you claim as well that the CPU bandwidth could use zero as well,even when on xbox one the CPU path to memory is 30GB/s RESERVEDDDDDDDDDDDDDDDDDDDDDD.
But that is not the only miss you have on your endless quest to make the xbox one look as strong as the PS4.
@ronvalencia said:
My postings on PS4 > X1 runs into problems with the following statement.
Xbox One and PS4 have no advantage over the other says RedLynx Trials creative director Antti Ilvessuo believes both next-generation consoles to be on par in terms of power. "The Xbox One and PS4 are on a equal footing according to Trials creative director Antti Ilvessuo. Speaking to VideoGamer.com at E3, Ilvessuo said: " Obviously we have been developing this game for a while and you can see the comparisons. I would say if you know how to use the platform they are both very powerful. I don't see a benefit over the other with any of the consoles." Ilvessuo also explained that "It all depends on how you use the platform and how you use it right. That's the thing." On the subject of used games, the man behind Trials also believed that 'developers can't really worry about that [side of things]' and that it was a matter for 'different people' to sort out."
--------
MS haven't offically revealed X1's GPU clock speed.
YOU quoting a developer on Beyond3d claiming the PS4 doesn't have no advantage over the xbox one,funny enough Trials,which was superior on PS4 and was a double kick to your face as it is a virtual texture game which you claim works better in 2 memory pools..lol
Then you ending saying MS haven't officially reveal the X1 GPU clock is the icing on the cake,it could have being 1.1ghz for all we care 18CU at 800mhz >>>>>>>>> 12 even at 1.1ghz in Sea Island architecture,the 7790 is 14CU 1.79TF at 1027mhz,the 7850 is 16CU at 800mhz 1.76TF ,yet the 7850 tops the 7790.
But you knew speed didn't matter you just continue to defend the xbox as usual.
@ronvalencia said:
Semiaccurate.com already covered the yield issue about a year ago. We don't have a confirmation on down clock speed.
The Xbox Yukon leak fromhttp://www.pcmag.com/article2/0,2817,2405922,00.aspspec'ed for a 1Ghz GPU.
Can i say i see what you did there.?
But but but 1ghz...lol
@ronvalencia said:
7850 = 45.
W5000 = 33.
What you use to claim the gap would be,based on your failed W5000 arguments which compared a Pitcair GPU with 12CU and 32ROP and NO ESRAM with a Bonaire 12CU with ESRAM which while it had 104GB/s wasn't optimal for everything and was to small at 32MB,period it was proven,so how was the gap.?
Loading Video...
Oh it could be as small as a few frame or simple mind blowing huge...lol
WTF man is noty fu**ing memory compression alone what make the RX480 perform like a R390X what the fu** are you smoking Polaris has many improvements over GCN is not just memory compression.
NEW geometry processor. Minor improvement. NVIDIA GPUs already has triangle discard hardware.
NEW 4TH GENERATION GRAPHICS CORE e.g. Support for improved memory compression i.e. TMUs must know about.
NEW COMMAND PROCESSOR. Minor improvement. Perhaps less reliance on ACE units. R9-390 has 8 ACE units to maximize it's shader resources.
NEW L2 CACHE. Minor improvement e.g. L2 cache must handle the new compressed data.
NEW DISPLAY ENGINES. Not related to 3D compute i.e. it's not ROPS.
NEW MULTIMEDIA CORES. Not related to 3D compute i.e. H.265 10bit support for 4K videos
NEW MEMORY CONTROLLER. Support for higher speed GDDR5-8000.
Besides memory compression and effective memory bandwidth improvement (~similar level as R9-290X), Polaris' other improvement are relatively minor since RX-480's 5.83 TFLOPS and frame rate result falls into R9-390(5.12 TFLOPS) and R9-390X (5.9 TFLOPS).
Try again.
@tormentos said:
WTF man you are trying to make see that only memory compression is the only thing improving Polaris,you are a joke and a MS ass kisser of epic proportions.
You are not thinking correctly.
@tormentos said:
The only thing i never fact is your bullshit theories on how system and CPU will run with absolutely ZERO 0000000 bandwidth as by your pathetic ass argument all the 320GB/s are for GPU and nothing for system this is pathetic you are a blind biased lemming who simple can't admit being wrong and want to argue shit that can't be argue..
I didn't say it's zero incidents. You didn't read the article on PS4's CPU cache and software optimizations for it.
Cache's main job is to reduce CPU's memory access i.e. it's a form of hiding the latency.
Software optimizations are geared towards fitting game code within the L2 cache hence reducing cache spill over incidents. The approach is similar to SPE's local memory software optimizations. On modern X86 CPU, cache spill over doesn't equal massive performance stalls e.g. out of order processing attempt to execute non-related non-stalled instructions i.e. X86 CPU is more forgiving than IBM PPE/SPE.
On PC, Swiftshader (Direct3D9c software renderer) example has a setting to limit the cache spill over events hence maximizing compute performance i.e. set CPU's cache size value into Swiftshader's settings file.
Read http://www.intel.com/content/www/us/en/architecture-and-technology/64-ia-32-architectures-optimization-manual.html Starts on Page 277 on cache related optimizations.
Read http://stackoverflow.com/questions/1922249/c-cache-aware-programming as examples for cache aware programming.
This is why R7-265 is being accepted as close approximation for PS4's GPU solution.
X86 CPU has instruction compression i.e. fewer instructions for a given job when compared to competing RISC CPUs.
You're being ignorant which is not my problem.
@tormentos said:
HOW MUCH BANDWIDTH SCORPIO CPU AND SYSTEM WILL HAVE ACCORDING TO YOU.
To deliver Intel Core iX PC gaming with RX-480 OC level GPU, ZEN's memory access behavior must duplicate Intel's memory access behavior i.e. if ZEN has any additional memory access incidents, it's additional latency cost relative to Intel's solution. Additional high latency incidents = less performance. ZEN's memory access pattern and usage is capped to AM4 socket and DDR4 speed support limitations e.g. PS4's CPU memory bandwidth consumption was no different to AMD's desktop PC CPUs regardless of GDDR5 type i.e. it's copy and paste engineering.
I already gave you Intel Core iX's memory bandwidth usage numbers..
@tormentos said:
You did this already as well with the xbox one,claiming 68GB/s + 104GB/s from ESRAM and i corrected you and you claim as well that the CPU bandwidth could use zero as well,even when on xbox one the CPU path to memory is 30GB/s RESERVEDDDDDDDDDDDDDDDDDDDDDD.
What correction?
"Host Guest GPU MMU" can source it's data from 3 memory pools i.e. CPU's cache (30 GB/s), "DRAM CNTR" (peak 68 GB/s BW) and ESRAM (peak 204 GB/s BW).
@tormentos said:
But that is not the only miss you have on your endless quest to make the xbox one look as strong as the PS4.
During Hitman 2016's many NPC scene, XBO delivered superior performance over PS4's version. What happened to your generalizations?
I stated XBO has superior memory bandwidth over PS4 which is correct.
I stated XBO's higher memory bandwidth wouldn't solve GPU's lesser ALU power hence W5000's example vs retail 7850 and later R7-265. This deals with rendering the graphics not with controlling the dot co-ordinates which is usually done on CPU, but GpGPU can handle this workload(Froblins' crowd simulation) .
YOU quoting a developer on Beyond3d claiming the PS4 doesn't have no advantage over the xbox one,funny enough Trials,which was superior on PS4 and was a double kick to your face as it is a virtual texture game which you claim works better in 2 memory pools..lol
Then you ending saying MS haven't officially reveal the X1 GPU clock is the icing on the cake,it could have being 1.1ghz for all we care 18CU at 800mhz >>>>>>>>> 12 even at 1.1ghz in Sea Island architecture,the 7790 is 14CU 1.79TF at 1027mhz,the 7850 is 16CU at 800mhz 1.76TF ,yet the 7850 tops the 7790.
But you knew speed didn't matter you just continue to defend the xbox as usual.
Mad Max was similar between XBO and PS4 builds. We know Mad Max's developers has the XBO's 16 ROPS workaround with TMUs. This workaround is bring XBO's cut-down 7790 GPU into W5000 like handling. What happened to your generalizations?
Did Trials developer used ESRAM as thier textures partially resident target? Did they match Froza's usage?
From http://www.gamepur.com/news/16560-dev-explains-forza-horizon-2-1080p4x-msaa-management-says-controversial.html
For FH2, textures partially resident (PRT) was applied for ESRAM, hence maximising TMU's reads/write hence XBO is not limited to 7770's 72 GB/s like TMU read/write memory behavior.
For PS4's PRT is just to lessen HDD access. PS4's TMU access memory pool at single effective memory bandwidth.
For XBO 's PRT, it's maximize TMU's memory access AND lessen HDD access when used correctly. XBO's TMU access two memory pools with different effective memory bandwidth. This workaround is bring cut-down 7790 GPU into W5000 like handling.
During Hitman 2016's many NPC scene, XBO delivered superior performance over PS4's version. What happened to your generalizations?
I stated XBO has superior memory bandwidth over PS4 which is correct.
I stated XBO's higher memory bandwidth wouldn't solve GPU's lesser ALU power hence W5000's example vs retail 7850 and later R7-265.
W5000 is remains inferior retail 7850 and later R7-265.
@tormentos said:
Can i say i see what you did there.?
But but but 1ghz...lol
I merely cited a leak, I didn't create this leak and it's a speculation thread.
@tormentos said:
What you use to claim the gap would be,based on your failed W5000 arguments which compared a Pitcair GPU with 12CU and 32ROP and NO ESRAM with a Bonaire 12CU with ESRAM which while it had 104GB/s wasn't optimal for everything and was tosmall at 32MB,period it was proven,so how was the gap.?
You failed to factor in TMU workaround from Mad Max. http://www.eurogamer.net/articles/digitalfoundry-2015-mad-max-face-off
NEW geometry processor. Minor improvement. NVIDIA GPUs already has triangle discard hardware.
NEW 4TH GENERATION GRAPHICS CORE e.g. Support for improved memory compression i.e. TMUs must know about.
NEW COMMAND PROCESSOR. Minor improvement. Perhaps less reliance on ACE units. R9-390 has 8 ACE units to maximize it's shader resources.
NEW L2 CACHE. Minor improvement e.g. L2 cache must handle the new compressed data.
NEW DISPLAY ENGINES. Not related to 3D compute i.e. it's not ROPS.
NEW MULTIMEDIA CORES. Not related to 3D compute i.e. H.265 10bit support for 4K videos
NEW MEMORY CONTROLLER. Support for higher speed GDDR5-8000.
Besides memory compression and effective memory bandwidth improvement (~similar level as R9-290X), Polaris' other improvement are relatively minor since RX-480's 5.83 TFLOPS and frame rate result falls into R9-390(5.12 TFLOPS) and R9-390X (5.9 TFLOPS).
Try again.
You are not thinking correctly.
I didn't say it's zero incidents. You didn't read the article on PS4's CPU cache and software optimizations for it.
Cache's main job is to reduce CPU's memory access i.e. it's a form of hiding the latency.
Software optimizations are geared towards fitting game code within the L2 cache hence reducing cache spill over incidents. The approach is similar to SPE's local memory software optimizations. On modern X86 CPU, cache spill over doesn't equal massive performance stalls e.g. out of order processing attempt to execute non-related non-stalled instructions i.e. X86 CPU is more forgiving than IBM PPE/SPE.
On PC, Swiftshader (Direct3D9c software renderer) example has a setting to limit the cache spill over events hence maximizing compute performance i.e. set CPU's cache size value into Swiftshader's settings file.
Read http://www.intel.com/content/www/us/en/architecture-and-technology/64-ia-32-architectures-optimization-manual.html Starts on Page 277 on cache related optimizations.
Read http://stackoverflow.com/questions/1922249/c-cache-aware-programming as examples for cache aware programming.
This is why R7-265 is being accepted as close approximation for PS4's GPU solution.
X86 CPU has instruction compression i.e. fewer instructions for a given job when compared to competing RISC CPUs.
You're being ignorant which is not my problem.
To deliver Intel Core iX PC gaming with RX-480 OC level GPU, ZEN's memory access behavior must duplicate Intel's memory access behavior i.e. if ZEN has any additional memory access incidents, it's additional latency cost relative to Intel's solution. Additional high latency incidents = less performance. ZEN's memory access pattern and usage is capped to AM4 socket and DDR4 speed support limitations e.g. PS4's CPU memory bandwidth consumption was no different to AMD's desktop PC CPUs regardless of GDDR5 type i.e. it's copy and paste engineering.
I already gave you Intel Core iX's memory bandwidth usage numbers..
What correction?
"Host Guest GPU MMU" can source it's data from 3 memory pools i.e. CPU's cache (30 GB/s), "DRAM CNTR" (peak 68 GB/s BW) and ESRAM (peak 204 GB/s BW).
@tormentos said:
But that is not the only miss you have on your endless quest to make the xbox one look as strong as the PS4.
During Hitman 2016's many NPC scene, XBO delivered superior performance over PS4's version. What happened to your generalizations?
I stated XBO has superior memory bandwidth over PS4 which is correct.
I stated XBO's higher memory bandwidth wouldn't solve GPU's lesser ALU power hence W5000's example vs retail 7850 and later R7-265. This deals with rendering the graphics not with controlling the dot co-ordinates which is usually done on CPU, but GpGPU can handle this workload(Froblins' crowd simulation) .
Mad Max was similar between XBO and PS4 builds. We know Mad Max's developers has the XBO's 16 ROPS workaround with TMUs. This workaround is bring XBO's cut-down 7790 GPU into W5000 like handling. What happened to your generalizations?
Did Trials developer used ESRAM as thier textures partially resident target? Did they match Froza's usage?
From http://www.gamepur.com/news/16560-dev-explains-forza-horizon-2-1080p4x-msaa-management-says-controversial.html
For FH2, textures partially resident (PRT) was applied for ESRAM, hence maximising TMU's reads/write hence XBO is not limited to 7770's 72 GB/s like TMU read/write memory behavior.
For PS4's PRT is just to lessen HDD access. PS4's TMU access memory pool at single effective memory bandwidth.
For XBO 's PRT, it's maximize TMU's memory access AND lessen HDD access when used correctly. XBO's TMU access two memory pools with different effective memory bandwidth. This workaround is bring cut-down 7790 GPU into W5000 like handling.
During Hitman 2016's many NPC scene, XBO delivered superior performance over PS4's version. What happened to your generalizations?
I stated XBO has superior memory bandwidth over PS4 which is correct.
I stated XBO's higher memory bandwidth wouldn't solve GPU's lesser ALU power hence W5000's example vs retail 7850 and later R7-265.
W5000 is remains inferior retail 7850 and later R7-265.
I merely cited a leak, I didn't create this leak and it's a speculation thread.
You failed to factor in TMU workaround from Mad Max. http://www.eurogamer.net/articles/digitalfoundry-2015-mad-max-face-off
What happened to your generalizations?
Try again.
1-FACT POLARIS HAS 6 NEW IMPROVEMENTS OVER OLD GCN.
Downplaying them doesn't change the fact that memory compression alone isn't what improve performance,in fact it has new graphics cores which is were performance on AMD GPU come from,so your whole try harder comment is total bullshit.
So no it is you who doesn't speak correctly.
2-Zen vs Intel mean shit until AMD release their CPU period,you are assuming and your assumption,mean shit and are horribly biased in favor of AMD and MS xbox most of the time.
@ronvalencia said:
Split render target combines 32 MB ESRAM and DDR3 memory pools into a virtual single frame buffer surface which yields similar memory write bandwidth as PS4.
For 1920x1080p, 32MB ESRAM contains 70 percent of the workload while DDR3 contains 30 percent of the workload.
Theoretical peak for frame buffer write operations
XBO, 108 + 68 = 176 GB/s
The frame buffer on xbox one is not 68GB/s + 104GB/s because 30GB/s is offlimit to the GPU at all time,so is 38GB/s + 104GB/s and after that we must deduce the loss from optimal performance,since the whole 38GB/s can't be use as that would be PEAK so it is much lower,like i told you is not the first time you try to use the whole bandwidth to claim it as GPU memory.
@ronvalencia said:
PS4's memory operates at 176 GB/s
Sony's 140 GBs/ is in line with any AMD DDR3/GDDR5 memory controller efficiency. AMD R9-Fury throws GDDR5 and it's memory controllers in the bin.
Hahahaaa........... See so since Scorpio uses GDDR5 and AMD memory controllers that mean it has the same efficiency as the PS4 memory,so from from 176 to 140 is a what 20% loss in bandwidth.? Ok lets apply that to scorpio..
320 - 20% = 256GB/s that is what Scorpio will have for CPU and GPU,you know i am just using your theory of this mean that and that = this.
So applying the same loss since the PS4 is 1 unified memory pool of GDDR5 and is AMD the Memory controller most be the same,so that mean Scorpio will have a 20% loss in bandwidth.
This is you see what you are doing there you blind biased BUFFOON,you are using the xbox one 68GB/s bandwidth combined with ESRAM 104GB/s to claim it = Pitcairn 153Gb/s memory Bandwidth... You blind biased fanboy that isn't true because the xbox one has a 30GB/s reservation which the GPU CANT TOUCH YOU FOOL.
THE XBOX ONE GPU CAN'T USE 30GB/S FROM THE MAIN BANDWIDTH BECAUSE IT IS DEDICATED TO THE CPU IT CAN'T USE IT SO YOUR CLAIM WAS BULLSHIT,AND DONE TO SHOW SOME HOW THAT THE XBOX ONE GPU HAD THE SAME BANDWIDTH AS PITCAIRN WHICH IT JUST DOESN'T.
@tormentos said:
Evidence of your lack of knowledge...Hahahaaaaaaaaaaaaaaaaaaaaaaaa
The xbox one can't use 68GB/s + 104GB/s because CPU and system take 30GB/s from the main pool leaving only 38GB/s...lol
That fossy math of yours...
This was my reply to you ^^ on that thread.
@ronvalencia said:
Did you forget there's a link between CPU and GPU?
Have you realized CPU's memory consumption floats between 0 to 30 GB/s?
Have you realized the purpose why you have a CPU L2 cache?
What you reply with. Wait a fu**ing minute isn't that the same shit you told me about Scorpio bandwidth that somehow CPU bandwidth usage was from 0 to 30GB/s.? While you were trying to argue that Scorpio GPU has faster bandwidth than the RX480 on PC.?
You are to hard headed and you argue shit that SHOULD NOT BE ARGUE PERIOD.
Fact is both the xbox one and Scorpio have shared bandwidth that mean and no time during gameplay the GPU will get all the bandwidth and some how the CPU will have 0 bandwidth in fact the way it is now,that bandwidth even while share still reserved to the CPU and can't be step on by the GPU,is not how things work if the CPU has 30GB/s connection,and is using only 10GB/s the GPU will not step into the 20GB/s at any time because that whole bandwidth is reserved for a reason,first since its only 30GB/s that mean efficiency will dictated that the numbers is 20% lower already,so there is no way the GPU step there which is what you patheticly try to imply,trying to make see that since the CPU bandwidth usage can be between 0 and 30GB/s some how when is at 0 the GPU will use that bandwidth which is not true,and second there is not a fu**ing instance on consoles while gaming that the CPU is using 0 bandwidth so stop pulling shit from your ass.
Oh look Mad max was equal on both consoles,that prove its the xbox one is exactly as powerful as the PS4..lol
Man since launch there have being games that are = Need For Speed was 1080p 30FPS on both,it mean total shit is just parity which some developers will have on its games,and say nothing bad about the PS4 rather that parity was set as a goal from the offset for the game,any game that is 1080p on both platforms and perform the same or close it because the PS4 was hold back for parity PERIOD.
12FPS gap.
12FPS gap.
This doesn't occur on PS4 according to DF it self on 1:34 of the comparison,so the xbox one version was up to 4 frames faster for a couple of second the PS4 commands a 10 to 14 FPS lead in most of the game.
No it doesn't and that 104GB/s is 32MB which aren't even optimal for anything which is the problem with many xbox games.
Your bandwidth claims are as sad as MS ones,but but but the xbox GPU has 68GB/s +104GB/s the CPU run on magic and love.
@ronvalencia said:
Did you forget there's a link between CPU and GPU?
Have you realized CPU's memory consumption floats between 0 to 30 GB/s?
Have you realized the purpose why you have a CPU L2 cache?
Let me re quote this because this alone prove how blind and biased you are in favor of the freaking xbox brand,CPU cache has shit to do with the limitation imposed on bandwidth on xbox one so that the CPU had bandwidth to operate,because only on your biased world CPU run on love and magic and not bandwidth according to your 0 bandwidth argument.
Log in to comment