Article View: comp.sys.ibm.pc.games.action
Article #6381Re: Upgrade Recommendations Needed For Gaming
From: Thomas Russell Hong
Date: Sun, 01 Oct 1995 00:00
Date: Sun, 01 Oct 1995 00:00
51 lines
3141 bytes
3141 bytes
On Thu, 28 Sep 1995, Nearly Normal Norman wrote: > Give it six months and VRAM based graphis cards will be a thing of the > past. the price/performance trade off isn't really worth it, compared to > the speed boosts you can get from graphics accelerators using 128 bit > buses and EDO DRAM. Hi Norm, I think your assessment is just a tad too optimistic. I don't think we will be seeing 128-bit buses nor EDO RAM on graphics cards! =) PCI buses are 32-bit. The dual PCI standard to be released with P6s are 64-bit. We are still a long way from 128-bit buses. :-) If you are referring to the Number Nine Imagine 128, it doesn't have an 128-bit bus, but rather it processes video data 128 bits at a time. It still fetches data 32 bits at a time over the PCI bus. VRAM will still be around for a while. Its current incarnation is WRAM, or Window RAM, which is also dual-ported like VRAM, and is being used on upper-end graphics cards such as the Matrox Millenia. The advantage of dual-ported VRAM or WRAM voer single-ported DRAM is not apparent at lower colors such as 640x480x256 since a computer does not need much video RAM at this resolution/color. At such color levels and resolutions, an Imagine 128 would not outperform a humble w32p-based card noticibly (except maybe in Windows when the expanded 2D acceleration capabilities of the I128 would come into play). However, the advantage of VRAM/WRAM becomes more apparent when one needs 16-bit, 24-bit or 32-bit color for stuff like 3DStudio renderings at say, 1024x768x16.7million resolutions, where the amount/speed of the RAM onboard the video card becomes important. The improvement in graphics cards in the next six months will neither be in bus width nor RAM on-board, but rather with the coprocessor chips which can take massive computational loads off the CPU by handling 3D computations such as all the operations for calculating, shading, texture-mapping, and even video-mapping onto polygons, and for full-motion video acceleration such as a hardware decoder that supports various formats such as MPEG or AVI. Regards and clear skies, TRH Sends ________________________________________________________________ Thomas Russell Hong Columbia University in the City of New York Department of Biological Sciences email: th...@columbia.edu The W3 USNF Graphical Landing Tutorial on my WWW homepage: http://www.columbia.edu/~th34
Technical Details
Message-ID:
<Pine.SUN.3.91.951001012822.15646A-100000@aloha.cc.columbia.edu>Path:
mbox-mo1.pugleaf.net!unknown!not-for-mailReferences:
<9508188114.AA811457094@hqafsoc.afsoc.af.mil> <43ks0v$2a8@newsbf02.news.aol.com> <43v2as$en0@clyde.open.ac.uk> <O88SFBAtnXawEAL3@jrca.demon.co.uk> <norm-2809950914440001@orbit.demon.co.uk>Show Full Headers
Complete Article Headers
From: Thomas Russell Hong <th...@columbia.edu> Subject: Re: Upgrade Recommendations Needed For Gaming Date: 1995/10/01 Message-ID: <Pine.SUN.3.91.951001012822.15646A-100000@aloha.cc.columbia.edu> X-Deja-AN: 116850697 references: <9508188114.AA811457094@hqafsoc.afsoc.af.mil> <43ks0v$2a8@newsbf02.news.aol.com> <43v2as$en0@clyde.open.ac.uk> <O88SFBAtnXawEAL3@jrca.demon.co.uk> <norm-2809950914440001@orbit.demon.co.uk> x-sender: th...@aloha.cc.columbia.edu organization: Columbia University mime-version: 1.0 newsgroups: comp.sys.ibm.pc.games.action,comp.sys.ibm.pc.games.flight-sim Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit