Post Xbox One Two Scorpio, what should Sony do next? *spawn* (oh, and Nintendo?)
- Thread starter mpg1
- Start date
If it's held at the end of May it's possible we wouldn't hear about it until next week.
I think there is some wiggle room with Phil's comments, but perhaps you're right.
GrimThorne
Newcomer
You've just described the LIFE of a journalist. They live their lives mobile and are constantly jumping from one flight to the other. I dated a photographer for awhile and I can tell you that no one cares about framing an event around the schedule of a member of the press. She would get a call and she would have to go. If you can't make it they'll just send somebody else. That's it.
Now as for Microsoft holding an impromptu hardware announcement? That will depend upon just what is actually being revealed. An Xbox Slim requires no special announcement and can very easily be tucked into their E3 conference. But if the hardware announcement IS significant then Microsoft has already proven they can put together a last minute reveal.
They could simply proceed in the same way they did before the June 18, 2012 Microsoft Surface reveal. Those press invitations went out by email only 4 days before the event. The actual location of the venue (Milk Studios) wasn't shared with the invitees until 3 hours before it started. They were only told the event was in LA.
So press attendance and time is NOT a factor if Microsoft really has something to reveal.
XB1.5 and PS4.5 are going to be exactly the same, aren't they. Probably have announcements on the same day, and then the execs will meet in a bar and talk about what's going on and how they both have the same product...why the hell are we doing this...all this competitive stress...let's just be friends! And in two months well have the PlayStation Xbox (Xbox PlayStation in NA) that plays all games. And the consoling world will collapse...
I think that the mid cycle upgrade was always going to happen. New generations are simply far too expensive for them to do because you have to front-load so many new expenses to launch a new console from games to OS development. By continually upgrading their systems they can avoid this expense, and give themselves more marketing options if they want to split the system between high end and mainstream offerings.
Would DDR4 offer much of an improvement in bandwidth? Or would they need to consider going up to HBM/GDDR5 if they are thinking of upgrading the performance?
Would DDR4 offer much of an improvement in bandwidth? Or would they need to consider going up to HBM/GDDR5 if they are thinking of upgrading the performance?
DDR 4 could give them 50% more BW if they were prepared to pay a premium for the fast stuff. Taking into account a presumed 80% memory controller efficiency and 20GB/s cpu access, it might net you around 75% more BW for the GPU portion.
Sony have more than doubled GPU performance on a 25% increase in BW. Perhaps MS could treble it with DDR 4? I dunno. Pretty sure they'd need a larger and faster pool of esram (and more ROPs) to support it though.
Sony have more than doubled GPU performance on a 25% increase in BW. Perhaps MS could treble it with DDR 4? I dunno. Pretty sure they'd need a larger and faster pool of esram (and more ROPs) to support it though.
Last edited:
My personal hope and dream here is that perhaps this slim model is a better variant with better clocks but we aren't talking whole new gen level like Neo is.
And... Then over clock the OG XBO to meet the same performance spec hur
2017 xb1.5 when can go with newer tech gddr5x polaris, zen etc.
this year slim with buffed up spec on smaller die.
buffed(min):
hdmi 2 - to allow 4k streaming with updated hardware decoder/encoder.
up clocked cpu & gpu to make everything a little better.
even a 50Mhz gpu up clock would make a big difference(can't remember where it would place it TF wise)
leave esram, memory etc as is.
is anthing here for slim to unreasonable / problematic?
this year slim with buffed up spec on smaller die.
buffed(min):
hdmi 2 - to allow 4k streaming with updated hardware decoder/encoder.
up clocked cpu & gpu to make everything a little better.
even a 50Mhz gpu up clock would make a big difference(can't remember where it would place it TF wise)
leave esram, memory etc as is.
is anthing here for slim to unreasonable / problematic?
I have a feeling we will not be seeing 64MB of ESRAM for the shrinks that are expected... unless CU/ROPs are being cut/halved for more ESRAM space.
Or did ESRAM fabrication (module shrinks) become smaller? I haven't really kept up with this...
2.5 gHz GPU incoming!
I guess we really don't know if MS could make a fully BC nextbox with GDDR5. I'd be inclined to guess yes, given the frankly amazing job they've done with their last two passes at BC (with OG -> 360 we even had very tech minded professionals swearing blind on B3D that it wasn't BC but recompilation until the lead BC engineer listed what he'd done on his CV!). I'm sure it would be easier with a more similar memory arrangement though.
If MS got their 64MB of esram by doubling up the number of 8MB macros used on the X1 then we could be looking at well above 400 GB/s, and with advantages beyond just high bandwidth and low latency.
I'm not sure about the whole upclock argument. That was never something tried before - we never had a PS2 Slim with 20% faster CPU and GPU. The mid-gen upgrade is talking about new hardware and a substantial difference. A mostly irrelevant upclock would do little to imrpove the experience but would certainly pee off existing console owners. If there's not enough of a change to equate to a new product and offering, it's probably not worth doing.
I have a feeling we will not be seeing 64MB of ESRAM for the shrinks that are expected... unless CU/ROPs are being cut/halved for more ESRAM space.
Or did ESRAM fabrication (module shrinks) become smaller? I haven't really kept up with this...
While esram takes up die space, it doesn't necessarily use die area that would otherwise be used for CUs. That all comes down to cost, and what you save on die area you might lose by needing a wide external bus with more expensive DRAM. On 14 nm, 64 MB esram would take up about the same area as the 32MB on X1.
X1s 16 ROPs are unlikely to be due to die area IMO, as I understand them to take up little space. It's more likely down to either MS thinking 16 efficient ROPs would be enough, or the esram's four banks being in some way related to the four ROP quads.
Interestingly Lord sebbbi recently made an interesting post regarding ROPs. Basically, 16 ROPs are usually enough given PS4Bone BW limitations ... but there are exceptions (e.g. 16 ROPs suck for shadow map rendering!).
GCN has full fill rate when writing 64 bit render targets. Modern console games bit pack two 32 bit render targets to one 64 bit render target effectively doubling the fill rate (compared to most GPUs).
8 bytes/pixel (64 bpp) * 16 pixels/clock * 853 MHz = 109 GB/s.
Now this is just the ROP color writes. If you use depth buffering or sample any textures, you obviously need even more bandwidth to fully utilize 16 ROPs.
PS4 memory bandwith is 176 GB/s. Subtract the CPU (up to 20 GB/s) and convert theoretical max to actual achievable bandwidth and you see that 16 ROPs are enough. More than 16 ROPs give proper benefits only when you are not bandwidth bound. All 64 bpp modes are bandwidth bound on GCN (as GCN has full speed 64 bpp ROP output). This means that HDR output (4x 16 bit float, such as lighting, post processing, etc) and/or g-buffer rendering (bit packed to 64 bpp RTs) do not get any gains from extra ROPs. More than 16 ROPs gives you benefits when you are rendering to 32 bpp (R8G8B8A8, R11G11B10F, etc) target or when you are rendering depth only (such as shadow maps).
Also you can use compute shaders to completely avoid the ROP bottlenecks. Most games already do this for post processing. Also traditionally high bandwidth rasterizer jobs such as particle rendering can be nowadays done with compute shaders. See here: http://amd-dev.wpengine.netdna-cdn....dering-using-DirectCompute-Gareth-Thomas.ppsx. This technique requires zero memory bandwidth for backbuffer operations or blending, because it renders the particles in tiles using LDS (64 MB local memory inside CU) to hold the backbuffer tile for blending. The final output is written to memory (once) using raw memory writes instead of ROPs.
Having more ROPs is of course nice, but you can avoid the performance hit in most cases... Except in shadow map rendering. 16 ROPs sucks for shadow map rendering
That's an interesting idea. What if Microsoft offered a slim with just higher clocks on a smaller die?
2.1GHZ cpu
1000MHz gpu = 1.5TFlops
Doesn't seem like much but if they offered this thing at $199 it would be compelling. Probably wouldn't require any changes from developers either.
that's where I'm coming from, enough of an up clock not come in last in DF, or add to cost, shouldn't need much dev input, and not a big enough jump to pee off exciting owners.
we've never had an upclocked slim before, but there's a lot happening that's never before....
so why do it? Because the always last perception/reality can be sorted with few downsides
'Always last' doesn't matter to the consumers buying on the low price end. If XB1 slim is a match for PS4Slim at 1.8 TF somehow, and priced the same as PS4Slim, do you really think would be PS4 buyers would get an XB1 instead? Or do you think there are people out there wanting an XB1 but will refuse unless it's primary number is the equal to PS4's? People bought PS3's plenty enough despite it failing to win many DF faceoffs. The primary buying decisions aren't the numbers, but value, momentum, library, yadayada, and the poking about with specs isn't that important once you're past the early adopters keen to buy the most powerful machine (plenty of whom are fanboy enough that they'll believe their console of preference is the most powreful machine regardless!).
I think such an upclocked console wouldn't benefit MS sales at all, but would annoy the existing userbase. The only upside is new owners would get a better experience, and maybe some XB1 owners would upgrade. Whcih they often do with slims anyway. A slim just uber quiet and small would probably be more appealing to the masses than a larger, hotter, more powerful slim, IMO.
I think such an upclocked console wouldn't benefit MS sales at all, but would annoy the existing userbase. The only upside is new owners would get a better experience, and maybe some XB1 owners would upgrade. Whcih they often do with slims anyway. A slim just uber quiet and small would probably be more appealing to the masses than a larger, hotter, more powerful slim, IMO.
Similar threads
