DAREALGUMMY
Member
Wasn't PS4 a half step?
Wasn't PS4 a half step?
threat of mobile devices have waned.
.
Those figures are just the rated power of the PSU, right? Sony just confusing things! I always thought the main reason Sony did this is that a PSU is most efficient at 40-50% load and therefore kicked out the least amount of heat in the smaller internal space of Sony consoles?
Although saying that Microsoft only have 245W PSU in an even smaller space so maybe not?
That's higher than I thought, guess I was confused with base model drawing somewhere in the 150w.
On the other hand it also bode well for PS5, there's only so much Sony can do to bring the consumption down.
The 'Max Draw' figures that manufacturers specify.
As far as the EU regulation goes, it regulates distinct operating modes, stand-by and forces the use of certain auto-shutdown features. They don't set a power limit for primary modes though.
That implies your cooling is near 100% efficient, which it is unlikely to be. It's how Microsoft managed to make the XBO X smaller than the XBO S despite the former consuming more power. The vapour chamber solution they use adds more expense though. These are all trade-offs and depend on what you wish to achieve. You may well just plop a heatsinc on an APU but that doesn't do much work spreading and dissipating heat. The goal is to distribute heat in as large a surface area as you can so that you can dissipate it into the air.
OG PS4 does not exceed 151W (KZ:SF).
Sources:
http://energyusecalculator.com/electricity_gameconsole.htm
http://www.psdevwiki.com/ps4/Power_Usage
https://arstechnica.com/gaming/2013/11/playstation-4-hardware-review-off-to-a-mixed-start/
IIRC, PS4 Pro can go up to 170W while running RoTR at 1080p60 (1080p30 Enhanced graphics mode goes up to 130W).
For reference, OG PS3 had a 380W PSU and it didn't exceed 200W at most.
HBM3/LCHBM is where it's at.
At the very least, we should expect 16 x 2GB GDDR6 chips, for a total of 32GB RAM at 256-bit bus (clamshell mode).
Why do people think this is unreasonable? By 2020 this should cost the same as 8GB of GDDR5 (16 x 512MB chips) in 2013 ($88).
Lithography progress alone (28nm -> 7nm) should easily enable a 4x jump (I know that DRAM chips use slightly different nodes, but my point still stands). 4x compared to what we used to get (16x every gen) is not that far-fetched IMHO.
Sony's numbers are PSU ratings.In fairly non-exhaustive tests conducted by third parties.
The numbers I quoted are directly from Sony themselves. Whilst based on what you posted (which I'm fully aware of) I personally don't expect "real" console usage to ever approach Sony's max. consumption numbers, I can only assume Sony's numbers are based on controlled stress tests with non-game code to test their cooling systems.
Fundamentally, the discussion about power consumption is as it relates to the design of the next-gen consoles. And so max. TDP target is what is important to console makers. In which case, the actual PS4 power consumption figures aren't all that relevant, since TDP will include design and safety margins well above this. That's essentially what I was pointing out with this post.
Sony's numbers are PSU ratings.
Imagine buying a PC with a 700W PSU. Are you going to reach 700W in real-world gaming usage? Nope, unless you have 3 GPUs in an SLI/CF setup (which does not exist in consoles).
Console APUs should be 75-100W at most.
Their software efforts have underperformed this gen, and the Xbox division doesn't have a blank check to do whatever they want. They have a focus on cost controls and profitability.
But there just is no reason for a new gen to start yet. The only one who might have been interested in that idea is MS. But for Sony and third parties, there is just nothing to make them change for another few years.
The "PS4 slim" (official CUH-2000 series) released in Sep 2016 reduced the rating of the powersupply to 165WThe PS4 PSU is rated at 250W (230W for the slim). Yet on Sony's site they quote 165 W maximum power consumption.
???
The PS4 PSU is rated at 250W (230W for the slim). Yet on Sony's site they quote 165 W maximum power consumption.
???
This was iterated on with the CUH-1200 model that reduced the PSU requirement to 230W (mainly because they halved the GDDR5 chips IIRC)
The PS4 PSU is rated at 250W (230W for the slim). Yet on Sony's site they quote 165 W maximum power consumption.
What's strange is that XBOX ONE X will have a 245W PSU. There's no way it's going to consume 120W only. It should be closer to 190-200W.Others have pointed out that various models of the PS4 have reduced demands on the PSU and the power supply used has been tweaked correspondingly. It's also worth noting that power supplies are invariably more efficient when not pushed anywhere near their limits. That means that a larger power supply will often generate less heat with the same load than a power supply that doesn't have as much headroom.
The PS4 PSU is rated at 250W (230W for the slim). Yet on Sony's site they quote 165 W maximum power consumption.
???
The "PS4 slim" (official CUH-2000 series) released in Sep 2016 reduced the rating of the powersupply to 165W
Source: https://en.wikipedia.org/wiki/PlayStation_4_technical_specifications#Power_usage
What's strange is that XBOX ONE X will have a 245W PSU. There's no way it's going to consume 120W only. It should be closer to 190-200W.
geordiemp said:Anyway, have we got any other gossip yet ?
It works in their favor, more units sold goes hand in hand with more software sold and potentially PS+ sub, all that is just pure bank and they'd get cheaper prices as well on components.
Coming back to what Pachter said about 50% 4K TV adaption rate, PS4 Pro will hold it's ground till a proper next-generation machine is released and the beauty of it = backward compatibility.
To be honest we would be better off with them waiting till 2020 or even 2021 IF it means we get a better machine, be it teraflops, memory, bandwidth or CPU instead of releasing too soon and risk a too big of a difference in raw power with XB2.
Came across something interesting Matt said in Microsoft revenue thread that stuck out and might worth considering as to when we will see a PS5.
What I make of the first quote is that Microsoft is locked in and they will have to ride this generation out at all cost, which is what we pretty much assumed.
At the same time because of that and because Sony is "winning" this generation they have absolutely no reason to release a PS5 earlier than maybe 1 year prior to XB2 launching.
Assuming the latter launches in 2021 then 2020 is where we'll see a PS5 (which lines up with all major components needed, process availability and maturity).
That would still give them a year's head-start, not to mention the whole BC thing in play that's a great sales pitch, more so with the vast library of exclusives PS4 has carrying over.
It works in their favor, more units sold goes hand in hand with more software sold and potentially PS+ sub, all that is just pure bank and they'd get cheaper prices as well on components.
Coming back to what Pachter said about 50% 4K TV adaption rate, PS4 Pro will hold it's ground till a proper next-generation machine is released and the beauty of it = backward compatibility.
To be honest we would be better off with them waiting till 2020 or even 2021 IF it means we get a better machine, be it teraflops, memory, bandwidth or CPU instead of releasing too soon and risk a too big of a difference in raw power with XB2.
Assuming what you say is true and the most likely scenario seems to be a PS5 in 2020. What should Microsoft do? Being a year behind PS5 and releasing the next Xbox in 2021 seems like a bad idea.
i bet they ditch backwards compatibility and just do psnow to grow their services
Maybe a half step. But it's still going to be marketed as a PS5. With PS5 only games. Consumers won't care and will still buy it.
We will see if it works in their favor.
Sony has yet to prove that they are capable of a seamless backwards compatibility solution. At the same time, Xbox One X is strong enough that it could be viewed as a next-gen console even though it is current-gen. When you look at it this way, MS will have a 2 or 3-year headstart against PS5 because we already know that their generational transition will be seamless. And they can and will release a new Xbox no later than one year after PS5.
So whatever Sony does, whether they are waiting or rushing things with PS5, it will always work in Microsoft's favor, that's how I look at it and therefore I think that releasing Xbox One X at this particular time was a smart decision.
We will see if it works in their favor.
Sony has yet to prove that they are capable of a seamless backwards compatibility solution. At the same time, Xbox One X is strong enough that it could be viewed as a next-gen console even though it is current-gen. When you look at it this way, MS will have a 2 or 3-year headstart against PS5 because we already know that their generational transition will be seamless. And they can and will release a new Xbox no later than one year after PS5.
So whatever Sony does, whether they are waiting or rushing things with PS5, it will always work in Microsoft's favor, that's how I look at it and therefore I think that releasing Xbox One X at this particular time was a smart decision.
It's bad because sales go to the forerunner for a year, and then price drop goes to the forerunner a year later.Why is it a bad idea? Provided that PS5 releases in 2020 or earlier, it seems like a great idea to me (to wait one year).
I actually dont care what the TF of Ps5 will be, 8, 12, 16, on my 55 inch 4K set 2160c looks good enough to me, its more about draw distance and pop in, but give me 60 frames anyday.
But is it OLED with HDR doe?
Next gen should be fun to see what they can do with "that" much power and pushing the envelope.
We will see if it works in their favor.
Sony has yet to prove that they are capable of a seamless backwards compatibility solution. At the same time, Xbox One X is strong enough that it could be viewed as a next-gen console even though it is current-gen. When you look at it this way, MS will have a 2 or 3-year headstart against PS5 because we already know that their generational transition will be seamless. And they can and will release a new Xbox no later than one year after PS5.
So whatever Sony does, whether they are waiting or rushing things with PS5, it will always work in Microsoft's favor, that's how I look at it and therefore I think that releasing Xbox One X at this particular time was a smart decision.
During AMD's Financial Analyst Day, the company announced their future plans for their Zen series of processors, with plans to release Zen 3 before or during 2020 using GlablaFoundries' 7nm+ manufacturing process.
This went over my head though I watched the conference that day...
Zen (14nm) 2017 > Zen refresh (14nm+) 2018 > Zen 2 (7nm) 2018 > Zen 2 refresh (7nm+) 2019 > Zen 3 (7nm+) 2019/2020, doesn't look like Zen 3 will make it in time if PlayStation launches in 2020, it would be a Zen 2 refresh.
Genesis-SNES, Dreamcast-PS2It's bad because sales go to the forerunner for a year, and then price drop goes to the forerunner a year later.
History has shown us multiple times that first to market is the better hand played.
Zen2 at 7nm is everything Ps5 needs, slap in some decent bandwidth and memory and TF will take what space is left and good to go.
Double masking vs EUV is more about number of process steps and fab costs, slight mods in Zen3 maybe upping clock in time for Ps5 pro lol.
After 7nm I can see all the enhancement work in reducing RC and speeding up clocks
I expect an IPC improvement on Zen2 to be somewhere around 15-20%, AMD seems optimistic/confident in their Zen architecture so maybe they'll surprise us like first Zen.
They also keep emphasizing that they have 2 teams working simultaneously on Zen 2 and Zen 3 because of data centers.
Which leads me to believe that they're heavily invested in the CPU side of the business, looking at what EPYC is doing I can see why.
Zen/AM4 platform will last till 2020 so that makes even more sense as to why they would want improved CPU's out ASAP since it's compatible with current mainboards.
I also believe that given the amount Sony will buy-in that they'll get a nice discount and price/performance, it's beneficial for AMD also and gaming as whole to push Zen.
When games start taking advantage of their 6-8 core CPU's in console space it might have an interesting impact on the PC market, at the very least it gives them a reason to "brag".
Probably worded it horribly, what I meant is that because games will be developed and optimized for 6-8 cores the PC counterparts might reap the benefits as well in the end.
I'd say we'll see a Zen2+ in PS5, I hope they clock it probably and not downclock it or settle for a measly 100Mhz boost like they did with PS4 Pro (especially if PSVR2 will be in play).
As for power consumption I'm not that worried looking at Digital Foundry's tests of the PS4 Pro: https://www.youtube.com/watch?v=0wNoCnPxTp4
PS4 Pro's CPU is about 500 MHz faster not 100 MHz.
I expect an IPC improvement on Zen2 to be somewhere around 15-20%, AMD seems optimistic/confident in their Zen architecture so maybe they'll surprise us like first Zen.
Zen is already 1.7 IPC of Jaguar
So any further small incremental IPC improvements from Zen to Zen 2 to Zen 3 is nothing compared to that of Zen over ANY jaguar.
I think geordiemp confuses IPC with single-threaded performance.Where did you get the 1.7 from, spot me a link if you can I'd like to read up, AFAIK AMD compared Zen to Excavator (52% IPC boost, don't know about Jaguar).
GPGPU and multi-threading are not going anywhere.Any incremental improvement is better than none and maybe they won't be inclined to heavily rely on GPGPU (safe to assume resources is being wasted that way that could be used elsewhere?).
Wasn't PS4 a half step?
I think geordiemp confuses IPC with single-threaded performance.
Single-threaded performance = IPC x frequency
Excavator is an entirely different uArch and it's less efficient than Jaguar, despite reaching higher frequencies (kinda like the P3 vs P4 comparison, if you get my drift). There's some speculation that inefficiency (less performance/more watts) was the reason it was rejected in current-gen consoles.No I confuse nothing, Zen is 1.52 IPC of Excavator.
http://www.anandtech.com/show/11143...or-under-330-preorder-today-on-sale-march-2nd
You have to dig a bit and juggle a bit of this and that to get excavator vs Jaguar, but in total its about 1.7 x.
Add add in clock improvements to 3 and 4 Ghz, its not even close. Do you think Intel dominating for so many years on just 20 % lol.
Go look at a die shot of Jaguar and see those tiny little zores compared to Zen.
Or check IPC of i5 vs any of the older AMD cores like Jaguar and you can see where 1.7 x and clock increases come into play.
If Jag was 80 % of Zen, nobody would give a damn and nobody would be crapping on everything Jaguar. Yes, its that bad, and yes, I5 class is that good.
One thing is for sure AMD is making great strides with Zen as a whole, I've never had an AMD CPU but something about Zen is swaying me.
Won't go into technical details that I don't understand so I'll leave that discussion for you guys to battle it out, though you probably saw these already.
https://community.amd.com/community...-more-performance-updates-for-ryzen-customers
https://community.amd.com/community...emory-oc-showdown-frequency-vs-memory-timings
It holds up with what users are experiencing, gaming performance has increased apparently, some article wrote it as follows "Ryzen's full potential wasn't unlocked at launch".
While it might not be relevant by the time PS5 is released a new architecture also comes with it's kinks and they are doing a great job ironing it out. Happy for the Red Team TBH.
I do wonder if the usage of GDDR memory would have an effect in any way since it's a lot faster than traditional RAM, thoughts?
Why is it a bad idea? Provided that PS5 releases in 2020 or earlier, it seems like a great idea to me (to wait one year).
We will see if it works in their favor.
Sony has yet to prove that they are capable of a seamless backwards compatibility solution. At the same time, Xbox One X is strong enough that it could be viewed as a next-gen console even though it is current-gen. When you look at it this way, MS will have a 2 or 3-year headstart against PS5 because we already know that their generational transition will be seamless. And they can and will release a new Xbox no later than one year after PS5.
So whatever Sony does, whether they are waiting or rushing things with PS5, it will always work in Microsoft's favor, that's how I look at it and therefore I think that releasing Xbox One X at this particular time was a smart decision.
As others have mentioned, being second has been bad historically. One year head start on sales and better price flexibility when your rival console launches. It will be interesting to see how well the Xbox One X sells.
I'm still curious what people think Microsoft should do. The general consensus is that they are screwed unless Sony makes a mistake? That can't be right. Maybe they speed up and try and launch something in 2020 to compete with when they think a PS5 will launch? Maybe with aggressive price drops on the Xbox One X they could ride it out until 2022 and release something much more powerful than the PS5?