Dienstag • 22.08.2017 • 15:04 Uhr
 Neuigkeiten
 Artikel
 Downloadbereich
 Bildergalerie
 Forum
 PowerVR FAQ
 Suchmachine
 Mitarbeiter
 Impressum
 SGX544MP angekündig...
 Der aktuelle Stand
 zweiter SGX Artikel ...
 Die Videofähigkeite...
 neue Treiber für Po...
 neuer Treiber mit Ha...
 SGX - Aufbau, Techn...
 neuer Serie5XT Grafi...
 Apple und Intel inve...
 Mehrere weitere Lize...
 SGX - PowerVRs Shade...
 Interview mit ImgTec...
 Quack - das andere D...
 Hercules4500 AGP4x-M...
 Optionen des KYRO Tr...
 FSAA auf KYRO
 Nummer 5 lebt!
 Startpartition von W...
 MBX - Mobile Grafik
 Die Grafikfähigkeit...
 Vivid! - PowerVR Gra...
 Videologic Neon 250
 Geforce 6800 - the n...
 Beyond3d
 BluesNews
 Chrome-Center
 Hartware
 Imagination T.
 ParaKnowYa
 Planet 3dNow!
 PowerVR
 PowerVR-Ext.
 PVR Generations
 Tom's Hardware
 Videologic
 VoodooExtreme

3D Center


PowerVR

Imagination Technologies

Interview mit ImgTec 2003

37. As mentioned, applications seem to be becoming more shader than bandwidth limited as time goes by. Do you feel that TBDRs in their future incarnations will require just as much raw bandwidth as competing solutions, or do you think that you will still sustain an advantage in that department?

John: see above (Question 36 - Previous page).


38. We've seen IMRs through the years drift slowly away from strict brute force rendering approaches, while incorporating more and more advanced bandwidth saving techniques and essentially closing the gap in some of the departments where deferred renderers had clear advantages. Have you managed in the meantime to increase the efficiency of your own approach even more and if yes is there anything you can share as of yet?

John: IMRs have improved their bandwidth use over time, but they still are considerably less efficient than a deferred renderer, and we have also been making significant strides in improving PowerVR technology though I wouldn't want to share too many details with you yet ; )


39. TBDRs can have differing render states per pixel inside a tile and must be able to switch them quickly, because they process primitives somewhat out of order. This problem is amplified for pixel shaders, where a renderstate means there is suddenly a lot of data. On top of that comes programs and constants, fixed functions like texture bindings, etc. Is this a theoretical problem only, or is it something that actually requires special attention?

John: Yes, it requires special attention.


40. I´ve seen recently NV3x´s pipeline being described as “thin and deep”, while R3xx´s pipeline rather “shallow and wide”. How would you describe the Series5 pixel pipeline in that sense?

David: That's too specific to an unannounced product to answer.


41. Up to now we've seen deferred rendering only with pixel processing. Have there been thoughts - or even better - attempts to use DR with vertex processing? If yes, are the results encouraging?

John: It's something that has been researched.


42. With more and more complex shaders appearing and polygon amounts constantly increasing, the data volume of a whole frame after vertex processing can become quite large. On an IMR, said data volume is only streamed through internal data busses. On the contrary, a TBDR has to store a large amount of that data in the memory. Is that a concern for PowerVR, or have you found a solution in the meantime?

John: we've developed a range of techniques that continue to mean that PowerVR has lower bandwidth requirements than competitive solutions


43. PowerVR hasn’t had its own OpenGL extensions in the past even though quite a few additional functionalities can be unlocked with those. Do you plan to continue that policy and if yes is it possible to clarify the reasons for such a decision?

John: No. Extensions will be used where appropriate - for example to enable developers to access all of PowerVR MBX's features under OpenGL ES.


44. If you have the chance to expand DX9.0 features today, which feature in particular would you add?

David: That's too specific to an unannounced product to answer.


45. In October 2002 John stated that Series5 will make its debut in 2003. Does that mean that we will most likely see an announcement of the product before this year runs out?

David: That's too specific to an unannounced product to answer.


46. What would the responsibilities of a potential future partner for Series5 be?

John: The responsibilities for licensees of our IP would be largely the same as in the past. I can't comment on "Series5" specifically.


47. Will we find out by then who the potential (or even existing) licensee of Series5 will be?

David: That's too specific to an unannounced product to answer.


48. Finally....(no we can't resist this time...) what are the Series5 specifications, what will it cost, who will be your future partner, when it will be available for sale, are you going to sit there and say nothing? (*grin*)

David: ...



Deferred Power dankt an dieser Stelle John Metcalfe und David Harold für die Teilnahme am Interview.

Das Interview kann und sollte in unserem Forum diskutiert werden!

Die folgende Seite enthält eine Bildergalerie mit Bildern von PowerVR MBX Demos.

Seite 1 2 3 4 5
 Interview mit ImgTec 2003
 5 Seiten
 verfasst von loewe
 Montag - 07.11.2005 - 19:11 Uhr
[1]  Seite 1
[2]  Seite 2
[3]  Seite 3
[4]  Seite 4
[5]  Seite 5