Talk:TMS34010

"The first true real-time 3D"
It's already flagged as citation needed, but for me "In 1987 TI provided the the first demonstration of true real-time 3D games with stereo sound effects on a personal computer (PC)" is trivially false. "true real-time 3D games" are a lot older than 1987, e.g. Elite, Mercenary, and many of the home conversions of Star Wars, Battlezone, etc. Some of the platforms on which they appeared, such as the Amstrad CPC came wired for stereo out of the box — in that case it's not even opt-in, it's simply the case that if stereo speakers are plugged in then the different audio channels have fixed assignments to the left and right speakers. So if I had stereo speakers plugged into my Amstrad CPC and I loaded Elite, in what sense am I not looking at "true real-time 3D" with "stereo sound effects", before 1987? I think a leap in quality is being oversold as a first. — 64.48.93.0 (talk) 20:57, 8 November 2017 (UTC)

-- "but for me... true real-time 3D games are a lot older..."

Your comparison is to wire-frame vector graphics which very few people, even at that time, considered "true 3D". That class of graphics was used because real-time 3D modeling of solids and their use in 3-space, even if limited to flat shading, was too difficult for personal computers. Even in the game instances cited, the graphics are limited to operations on a plane, substituting 2D map collisions for bounding box collisions in 3-space, extensive references to look-up tables instead of real-time calculations, etc. These are all limitations that don't support the claim for pre-existing real-time 3D on PCs. In 1987, Texas Instruments demonstrated a flight simulator prototype that had none of those limitations. I don't have a video of that, but look at the rolling, rotating dodecahedron in the article's linked TI promotional video (made in 1986). That is real-time true 3D, far beyond anything such as Battlezone. Carrellk (talk) 07:51, 3 September 2020 (UTC)

"Such boards were common in the early 1990s."
Regarding the above disputed claim, currently present in the history section: While the cited sources certainly are existence proof (if you weren't already aware, which many were not, both back then and now), these sources do not really do much to support the claim these boards had been common. These boards were very high-end back in the day, and arguably aspirational products to some, but they didn't ship in volume and drop in price in a way that would justify calling them "common". Actually, they were relatively uncommon, and even many IT experts back in the day did not know about them, that's how niche they were, and they didn't really escape that high-end niche, despite their manufacturers' best attempts. The mainstream moved from cheap VGA clones to SVGA, to Diamond Stealth and S3s, and onwards and upwards to Voodoo and the like, passing by what could have been the next big thing. Granted, this comment isn't well-sourced either, but neither is the above article claim, and I don't think it can be.

Second thought, I should probably clarify/concede that the claim "Such boards were common in the early 1990s" becomes justifiable if it's only generally applied to "video accelerator cards for IBM PC compatibles", however even then it's not right to add "of which the TMS34010 was central". It's more like "of which the TMS34010 was one." The TMS34010 certainly was an early video accelerator card chip. You might even call it trailblazing if you want to big it up, but in and of itself, it did not have the good life (in the mass market) that maybe you might feel it deserved. Video accelerator cards became very common a little later, but the TMS34010 itself? In PC-compatible video accelerators? Not really.

I'd even say that in a way, the currently existing paragraph contradicts itself, because the "professional-level" qualifier is pretty much the opposite of "common" – so there's already a concession in there that "common" isn't true. —ReadOnlyAccount (talk) 07:09, 9 January 2024 (UTC)

The preceding is inaccurate, incomplete, and misleading. The mainstream did not go from VGA to SVGA to Diamond Stealth to Voodoo etc. The latter two are names for particular brands of PC graphics cards and not graphics standards. Including clones and variants but leaving out the Micro Channel specific graphics which were not adopted by the mainstream ("ISA" architecture PCs) the chronological progression of PC graphics is CPU 2D from 1981-1990 (CGA/MDA & clones through SVGA), Windows/GUI accelerators 1990-1995 (TMS34010 and 8514 clones), and 3D starting in 1996. Neither VGA/SVGA or GUI accelerators have died. VGA/SVGA is the default display mode until the operating system determines a GUI/Windows accelerator capability is present, and all 3D chips have both SVGA and GUI acceleration integrated.

Why "starting in 1996"? What was desired in PC graphics for gaming in a windowed environment wasn't well defined until Microsoft started defining the DX standard in 1995. As the DX standards developed they gave nV, S3, ATi, and 3Dfx targets for what to put into their 3D capable graphics chips. So there is a long period from VGA in 1987 to 1996 where the market moved from VGA clones to 3D capable graphics, and during that period windows or GUI accelerators displaced CPU driven graphics rendering. That period started roughly in 1990 with cards based upon TI's TMS34010 and expanded in 1991 with 8514/A clones combined with SVGA. In 1991 the most popular Windows/GUI accelerator cards were TMS34010 based - somewhere over a million units were sold. By 1992, combined 8514/SVGA clones by companies such as ATI and S3 dominated the Windows/GUI accelerator market and it remained that way until 1996. For 3D, gaming really, the first NV1 based graphics cards weren't released until late 1995 and 3Dfx's first VooDoo card wasn't released (by Orchid) until October 1996. If you bought a cheap computer from 1990 to 1995 and didn't use Windows, it used some form of SVGA. If you bought a better computer and used Windows from 1990 to 1995, you had a Windows/GUI accelerator. After 1995, the majority of PCs had Windows/GUI acceleration. What started in 1996 was demand for 3D capability for gaming.

Virtually all current 3D graphics technology has its origins from Silicon Graphics Incorporated (SGI). 3Dfx was founded by ex-SGI engineers. nVidia was founded by ex-SGI engineers. AMDs (ATI originally) current graphics architecture came from acquiring ArtX which was founded by ex-SGI engineers. Carrellk (talk) 19:28, 3 April 2024 (UTC)