Orbis (PS4) devkit docs leak; 8GBRAM, 2.2GBVRAM, new Controller


Recommended Posts

That's most likely true but there's always hope. The more RAM, the better.

Not really, there are several rumours floating about including rumours from current developers that the Durango RAM setup is less than ideal.

Take it with a grain of salt but some have said that the framebuffer can only register 1Gb of ram per frame, 2Gb for 3D.

Also have a read here, a developers eye view of both consoles, he isn't privy to next gen specs but gives his opinion on the current rumoured specs. Below is Durango.

Working here assuming the Eurogamer Article is close to correct. On this platform I'd be concerned with memory bandwidth. Only DDR3 for system/GPU memory pared with 32MB of "ESRAM" sounds troubling. 32MB of ESRAM is only really enough to do forward shading with MSAA using only 32-bits/pixel color with 2xMSAA at 1080p or 4xMSAA at 720p. Anything else to ESRAM would require tiling and resolves like on the Xbox360 (which would likely be a DMA copy on 720) or attempting to use the slow DDR3 as a render target. I'd bet most titles attempting deferred shading will be stuck at 720p with only poor post process AA (like FXAA). If this GPU is pre-GCN with a serious performance gap to PS4, then this next Xbox will act like a boat anchor, dragging down the min-spec target for cross-platform next-generation games.

With how good the Sony engineers have been at shrinking the PS3 OS but continuing to add more to it, I'd be much happier with faster memory (and CPU/GPU ;)) over more. 4GB is much more than the PS3 with it's strangely split 512mb.

Or I will put it this way, we're not seeing more than 4GB of DDR5 as it'll give near another ?500 PS3, which I hardly see Sony wanting this time around. I think it'll launch at ?349/399 across two SKUs.

Not really, there are several rumours floating about including rumours from current developers that the Durango RAM setup is less than ideal.

Take it with a grain of salt but some have said that the framebuffer can only register 1Gb of ram per frame, 2Gb for 3D.

Also have a read here, a developers eye view of both consoles, he isn't privy to next gen specs but gives his opinion on the current rumoured specs. Below is Durango.

That's very troubling. I really hope Microsoft doesn't drop the ball with Durango. I assumed the GPU would be similar to the PS4's GPU (based on AMD's GCN architecture). It just doesn't make any sense for them to use something older like AMD's VLIW4 architecture (e.g. Radeon HD 6000 series).

Not really, there are several rumours floating about including rumours from current developers that the Durango RAM setup is less than ideal.

Take it with a grain of salt but some have said that the framebuffer can only register 1Gb of ram per frame, 2Gb for 3D.

Also have a read here, a developers eye view of both consoles, he isn't privy to next gen specs but gives his opinion on the current rumoured specs. Below is Durango.

That's one thing i was worried about since long before these reports on the specs started coming. That Microsoft would include some kind of edram again, and also once again not make it big enough.

With how good the Sony engineers have been at shrinking the PS3 OS but continuing to add more to it, I'd be much happier with faster memory (and CPU/GPU ;)) over more. 4GB is much more than the PS3 with it's strangely split 512mb.

Or I will put it this way, we're not seeing more than 4GB of DDR5 as it'll give near another ?500 PS3, which I hardly see Sony wanting this time around. I think it'll launch at ?349/399 across two SKUs.

PS3 OS currently uses just 50Mb, I think theres a bit of breathing room with 512Mb :p

The RAM itself isn't split, Sony learnt their lesson with PS3, it will have one big 4Gb Unified chunk but 512Mb will be reserved for OS function.

There are unsubstantiated rumblings that Sony could free up RAM like a desktop PC does, it could unload everything unnecessary and just run whatever functions of the OS you would need while running a game, which could bring the OS limit down to 256Mb or possibly less.

That's very troubling. I really hope Microsoft doesn't drop the ball with Durango. I assumed the GPU would be similar to the PS4's GPU (based on AMD's GCN architecture). It just doesn't make any sense for them to use something older like AMD's VLIW4 architecture (e.g. Radeon HD 6000 series).

I hate saying it, but the current rumours are that both consoles will be using GCN, Orbis will have 18GCN Compute Units, and Durango will have 12GCN Compute Units. 1.8TF compared to 1.2.

PS3 OS currently uses just 50Mb, I think theres a bit of breathing room with 512Mb :p

The RAM itself isn't split, Sony learnt their lesson with PS3, it will have one big 4Gb Unified chunk but 512Mb will be reserved for OS function.

There are unsubstantiated rumblings that Sony could free up RAM like a desktop PC does, it could unload everything unnecessary and just run whatever functions of the OS you would need while running a game, which could bring the OS limit down to 256Mb or possibly less.

It would be stupid for Sony to use non-unified memory. They couldn't add support for cross-game chat to the PS3 for that reason. Most of the RAM was already pre-allocated to games and couldn't be re-allocated to the OS.

I hate saying it, but the current rumours are that both consoles will be using GCN, Orbis will have 18GCN Compute Units, and Durango will have 12GCN Compute Units. 1.8TF compared to 1.2.

Okay, that's good. I was afraid Durango would use a non-GCN GPU. A difference of 600 GFLOPS coupled with slower RAM means Durango is clearly inferior. In terms of raw processing power, Durango's GPU is close to the Radeon HD 7770 (1.28 TFLOPS). And Orbis' GPU is a little faster than the Radeon HD 7850 (1.76 TFLOPS).

Sony looks like they're going to have an edge over Microsoft with exclusives. Cross-platform games will obviously cater to the lowest common denominator (Durango).

PS3 OS currently uses just 50Mb, I think theres a bit of breathing room with 512Mb :p

The RAM itself isn't split, Sony learnt their lesson with PS3, it will have one big 4Gb Unified chunk but 512Mb will be reserved for OS function.

There are unsubstantiated rumblings that Sony could free up RAM like a desktop PC does, it could unload everything unnecessary and just run whatever functions of the OS you would need while running a game, which could bring the OS limit down to 256Mb or possibly less.

Yeah I know what I was trying to convey was I think 4GB of RAM will be fine for Sony to fit the PS4 OS into with how well they've done on the PS3 considering most of it's actual OS features were backported in, and the OS shrunk a lot in size. That would be a good way to work, but when it comes to in-game menus/soundtracks and other stuff, Sony need it to be mandatory out of the door for each developer. The PS3 ended up a mess with a lot of games as things weren't ready from the get go.

Maybe MS need more RAM to put 1080p adverts on the next Xbox dashboard :p

Yeah, the Orbis is meant to have 7970M 'based' GPU, which is a desktop 7870 at 7850 clocks.

It should be a little faster than the Radeon HD 7850.

Radeon HD 7870 (1000 MHz) = 20 compute units = 1280 shader processors

Radeon HD 7970M (850 MHz) = 20 compute units = 1280 shader processors

Radeon HD 7850 (850 MHz) = 16 compute units = 1024 shader processors

Radeon HD 7770 (1000 Mhz) = 10 compute units = 640 shader processors

Orbis (800 MHz) = 18 compute units = 1152 shader processors (based on this)

Durango (? Mhz) = 12 compute units = 768 shader processors (see above link)

They lowered the GPU's clock speed to 800 Mhz (from 850 MHz) and removed 2 compute units. Regardless, it's still much better than the Durango's measly 768 shader processors. A difference of 384 shader processors is a lot to developers.

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • lots of people us facebook for stuff, threads though no
    • Can you read? I've said I'm willing to pay more for a notchless (no notch) 3:2 screen.
    • Not even an OLED display on the laptops. Also it seems that the laptop design isn't the same as the Surface Ultra model. Looks like bargain bin at high prices.
    • make your own notch - it's not that hard
    • VirtualBox 7.2.10 by Razvan Serea VirtualBox is a powerful x86 and AMD64/Intel64 virtualization product for enterprise as well as home use. Targeted at server, desktop and embedded use, it is now the only professional-quality virtualization solution that is also Open Source Software. Presently, VirtualBox runs on Windows, Linux, macOS, and Solaris hosts and supports a large number of guest operating systems including but not limited to Windows (NT 4.0, 2000, XP, Server 2003, Vista, 7, 8, Windows 10 and Windows 11), DOS/Windows 3.x, Linux (2.4, 2.6, 3.x, 4.x, 5.x and 6.x), Solaris and OpenSolaris, OS/2, OpenBSD, NetBSD and FreeBSD. Some of the features of VirtualBox are: Modularity. VirtualBox has an extremely modular design with well-defined internal programming interfaces and a client/server design. This makes it easy to control it from several interfaces at once: for example, you can start a virtual machine in a typical virtual machine GUI and then control that machine from the command line, or possibly remotely. VirtualBox also comes with a full Software Development Kit: even though it is Open Source Software, you don't have to hack the source to write a new interface for VirtualBox. Virtual machine descriptions in XML. The configuration settings of virtual machines are stored entirely in XML and are independent of the local machines. Virtual machine definitions can therefore easily be ported to other computers. VirtualBox 7.2.10 changelog: VMM: Fixed issue when CentOS 10 VM was not booting due to the message "Fatal glibc error: CPU does not support x86-64-v3" (​github:gh-642) Devices/EFI: Fixed booting issue when ARM VM had less than 1024 MiB of RAM assigned (​github:gh-679) USB: Fixed issue when it was not possible to attach USB device to headless VM on Apple Silicon/macOS 26.4.1 (​github:gh-631) Storage: Fixed issue when VIRTIO-SCSI device was not recognized as SSD device by guest system (​github:gh-634) Network: Fixed issue in E1000 emulation code which triggered debug log creation (​github:gh-645) Network: Fixed issue in E1000 emulation code which prevented OS/2 guest from booting (​github:gh-683) Linux Host: Fixed issue when VMs could not be started due to kernel oops (​github:gh-639) Linux Host and Guest: Fixed issue when kernel modules were failing to build with openSUSE 16.0 kernel Linux Host and Guest: Added initial support for kernel 7.1 Linux Host and Guest: Added extra fixes for RHEL 9.8 kernel (​github:gh-676) Linux Host and Guest: Added possibility to build source code using NASM instead of YASM as the assembler (​github:gh-520) Linux Guest Additions: Added initial support for Extended Data Control Protocol for clipboard sharing with Plasma on Wayland guests (​github:gh-33) Linux Guest Additions: Added extra fixes for preventing vboxvideo kernel module build with kernel version 7.0 and newer (​github:gh-655) OS/2 Guest Additions: Fixed issue when Shared Folders automount and clipboard sharing stopped working (​github:gh-551) Download: VirtualBox 7.2.10 | 170.0 MB (Open Source) Download: VirtualBox 7.2.10 Extension Pack | 19.1 MB View: VirtualBox Home Page | Screenshot Get alerted to all of our Software updates on Twitter at @NeowinSoftware
  • Recent Achievements

    • Week One Done
      suprememobiles48 earned a badge
      Week One Done
    • One Month Later
      Windows Guy earned a badge
      One Month Later
    • One Month Later
      Prasann earned a badge
      One Month Later
    • Week One Done
      Prasann earned a badge
      Week One Done
    • First Post
      Dys Topia earned a badge
      First Post
  • Popular Contributors

    1. 1
      +primortal
      510
    2. 2
      +Edouard
      174
    3. 3
      PsYcHoKiLLa
      100
    4. 4
      Steven P.
      87
    5. 5
      ATLien_0
      70
  • Tell a friend

    Love Neowin? Tell a friend!