What's new... is old again
July 1999 News
With the advent of overclocking programs, it is now possible to set all cards to the same exact memory and core speeds so that statements can be made about the overall efficiency of a design. If such a comparison were done with the Rage 128, it would show a design that is more efficient per clock cycle (more efficient architecture), and more feature laden (DVD hardware, DDR support, TV-out, broadcast TV ready) and more reliable (ramp consistent with memory ramp) realized by targetting mainsteam memory speeds. From a number of posts around the web of users that have tried all three cards, I think that quality control issues are rampant with any video card clocked near the 180 Mhz region. Below are two posts that indicate that many of the preview/review boards may have been specially tested units not representative of what you will find at your local electronics store. This post is from Lightspeed
I for one am disappointed more with the TNT2 than I ever have been with my Rage Fury. For starters, the drivers still suck, the image quality still suffers from the same quirks as the original TNT, and considering the TNT2 is running a core that is 50% faster than my Rage Fury and Memory which is almost twice as fast (103/103 to 150/200) I would say that the TNT2 is pathetic in comparison. If it is an even playing ground with equal core and memory speeds, the TNT is toast! I guess those super-duper high core and memory speeds don't do as much as they should... I never got 4FPS in Q3A with my Fury...and the DVD and TV software ice the cake, feel free to flame away, there is more on this topic coming in the future.
It is almost like comparing a Rage 128 Pro to a first-gen TNT, makes no sense to me.
This second post is from the 3D Gaming Forum:
Date: Thu Jun 24'99 - 3:20pm Author: AS/400 Subject: Fury vs ultra TNT2 Fury vs ultra TNT2 Ok people, I've owned a Fury for the past 3 or so months..Pretty acceptable performance and I'm a massive gamer when I get home from work. No major headaches, just dithering issues which I fixed with the reg hack, and missing fog features which USED to work with the CD divers..and yet unsurpassed DVD to my TV. I haven't tried the latest rewritten directx drivers from ATI yet - which may fix the fog issue. I've just bought a Diamond ultra v770 (whatever) TNT2..and yes from the box it worked with the CD drivers ....but, it has it's share of problems...graphic glitches here and there that doesn't exist with the fury..I downloaded the latest detonator drivers...same deal, even worse - random crashes here and there in quake3 and piss poor performance with that particular game. Other than that, with 16 bit 3d games, the TNT2 is better overall...faster speed, better dithering...With 32 bit games - forget about it, almost a tie..with a slight edge going in favour with the TNT2, which is rediculous considering the speed of the memory and the core is 'supposed to be twice the speed of the fury. DVD on the Diamond? - yuck..I've tried all players...if I hadn't used the Fury's dvd features before I probably wouldn't be so critical, but I wouldn't watch DVD much on the Diamond either. Also, another issue, I tried a regular TNT2, and he graphic glitches were'nt there... I exchanged the ultra 3 times, even got a creative labs one, on my second trial - same glitches, on another machine with a fresh install of everything - same glitches.. Conclusion? Ultra TNT2s from diamond (and creative) are clocked too high for perfect imagery....the speed is slightly noticeable over the fury especially in 16 bit mode, although not as drastic as you may believe before actually buying and comparing yourself...i.e don't believe the hipe that's caused by benchmarks you see on various 'biased' webpages. If ATI would get it's act together with those drivers (c'mon guys you're nearly there), and if you already have a fury and contemplating "upgrading" to an ultra TNT2 - In my unbiased err tottaly objective opinion, I would say, DON'T bother - if you MUST have the best soon, I'd say wait for the G400 from matrix - that will eat both the TNT2 variants and Fury for breakfast...or if you can wait another few months... then WAIT and see what else is pending out there in 3d tech land... Date: Thu Jun 24'99 - 3:38pm Author: Gurm (gurm42@hotmail.com) Subject: Re: Fury vs ultra TNT2, in response to Fury vs ultra TNT2, posted by AS/400 on Thu Jun 24'99 - 3:20pm Re: Fury vs ultra TNT2 I agree wholeheartedly. I've tried 5 different TNT2's now. The standard Maxi Gamer Xentor from Guillemot had lockups when you changed screen resolutions... randomly. I exchanged it, and another was the same way. I tried the Maxi Gamer Xentor 32. Very nice. Very fast. Piss-poor DVD and Video playback. Not enough faster in 32-bit to warrant $250... although it DID have fog, which is kinda nice. Hehe. Then, the fan died... 2 days in. Lockups galore. These things are tweaked right to the edge... so I took it back. Another one, it overheated even with the fan working. Argh! Tried a Hercules. Exactly the same, only more expensive, and not as good TV-out. I ditched it. The Viper V770 features Diamond's Out-of-Control 99. Sucks. Slow. Oh, well. Back it goes. The Creative card is slow, but features "unified". Which sucks. Oh, well. I'll keep my Rage128. It has better 2D quality than ANY TNT/TNT2/TNT2Ultra. And the DVD and video acceleration RULE. - Gurm
It has been a while since I came across a relevant article on the state of computing hardware and sound advice on upgrading and it came as no surprise that it was by C't. They were the orignal seeds and technical sources behind the pioneering hardware sites includings Tom' Hardware Page. Their opinion on framerate seems to back up what a number of people are pointing out on the web:
Resolution Equivalent Rage 128 Q3 TNT2 Q2
640 framerate framerate framerate Rage Fury
640x480 20 46.2 48.4 63
800x600 31 na na
1024x768 51 na na 48
1280x1024 80 18.1 21.7
1600x1200 125 na na 23
My relationship with Tom has been much longer than many sitcoms but I am taking the advice of people that if a site does not meet your approval then do not "tune in". I will replacing Tom's Link with the neutral and authrorative C't site that now has many of the articles translated into english. You can also click on the following link to get machine translation of the more numerous German sampling of articles. Cheers to you, C't
The ID Boys have doing their own compatability testing and the following are some rough translations of some of the problems found with Super 7 boards, AMD processors and differente motherboard builds originally highlighted in C't. Old news from last year but a fuller description has been translated here :
"test indicate that there are serious stability
problems with Super Socket 7 mainboards with 100MHz
FSB, if you are using an AMD K6-2 CPU and an AGP
board with the nVidia Riva TNT. In our tests the
demo loop of Quake II (patch level 3.19) proved to
be very sensitive for this. This game creates a lot
of graphics load, especially vsync disabled (timedemo 1),
and uses multitexturing. Tested were 4 mainboards with
ALi-Aladdin-V chipset (Asus P5A, Micro-Star MS5169)
and VIA-MVP3 chipset (A-Trend ATC-5200, AOpen AX59Pro),
two K6-2/300 CPU's (production weeks A9821 and A9814)
and three TNT boards (V3400TNT, Viper550, ErazorII),
using Win98."
The article goes through BIOS and driver versions and
patch requirements, install details and other potential
pitfalls, PC BIOS configuration settings tried etc.
Various combinations "locked the PC within 10 minutes
of the Quake II demo loop". At that time, the article
stated that "the problem [seems] to be TNT specific,
as tests with 3Dfx' Banshee (Monster Fusion) and
Matrox G200 (Millenium) did not show any problems over
an hour's [loop]". Disabling Vsync makes the problems
more prevalent. They could not find any effect of heat,
but they did find a dependency on which CPU was used - the
same PC was a lot less stable with the A9821 CPU. The
MS5169 boards was less stable than the P5A. They could
not find similar problems with VIA based boards. Their
conclusion: "vendors obviously had varying success in
building upon the Aladdin-V chipset. In addition, the
system's behavior is affected by [production] deviations in
the AMD CPU's."
....frankly I am getting flooded with email about this and basically have to say "patience"....it is going to happen and will happen and it will happen it a bigger fashion than most of you could ever dream about. That said, any Rage 128 card will be a good Linux investment once the Xfree86 driver is coded but in the interim you can try this link to run a 1024x768 window at 60Hz (ouch).
My personal opinion of patents (especially software) is similar to that of Donald Peterson who was a professor instrumental in openly developing the SPICE circuit simulation program openly. Commercial development and support was allowed to develop in parallel with the ideas presented by SPICE. The details of SPICE can be understood by very few people and could only be picked up by lerning by doing/coding. If Peterson had decdided to be more protective, the IC industry would be nowhere close to where it is today. He offered constructs such as "tires" so that air filled ones could be developed. Unfortunately, many of patents today are what you might call obvious "tire" patents. Suppose you were to put 10 people in 10 different rooms with a general description of a well known problem, six out of 10 of them would come out of the room with the same solution. From these six, one comes out after two hours, 3 come out after three hours, and two hand in the same solution after a day. To me the solution is "obvious". I think that many patents are handed out to the first person out of the their room today. The concept of a patent was significant in Edison's mechanical engineering dominated day and electric power generation was in its infancy, but the fast and furious development of engineering tools allows for insight and knowledge and a spread of technology far beyond what patents were origninally designed to do in the first place. Ironically, patents now serve to slow the progress and spread of technology. The above thougths are purely my own and it is my hope that the patent law be reviewed and put more in the hands of a review board consisting of academia more concerned with the long term survival of the engineering profession (with an added element of the Baseball Hall of Fame thinking thrown in). I motion the IEEE get involved. Perhaps one way to have patents validated would be a requirement to have a Engineering paper published on the details and implementation of filed invention submiited within one year of filing. This submitted paper would then be marked up and return to both the inventor and the patent board. This would serve to put the people who grant the patents in contact with experts from the academia and increase their knowledge skills that are so critical to appraising a patent application. The above opiinions are purely my own and reflect my current study of the patent law.
Toyota once made a car called the 4Runner that was available as a two wheel drive truck. Sadly, the last few incarnations come only in 4WD versions today. It has been found that tooling car production for common combinations of often requested features (automatic transmission and air conditioning) make for less cost and more profit. I was pricing the Rav4 with air conditioning but found that it could only be had with package "B" that gives you power windows (frill), power door locks (frill #2), matching door handles (yeech!...almost as bad as colour matched bummpers), power remote mirrors (adjusted once at purchase time), cruise control (a pacifier to put you to sleep on long drive ?), colour co-ordinated mirrors (aheemm...), CD (OK...if you force me to...), moulded spare tire cover(aheemmm...), floor mats and of course air conditioning. The cost for air alone is $1,600 and for the value package is $2,200. I could purchase air from GM for $945 for the Tracker (Vitara clone) which is installed for them by Suzuki. Suzuki charges you $1,600. Go figure.
If Toyota is listening, I like to see a basic Rav4 with only two wheel drive in the 4 door hardtop version retailing for about $2,000 dollars less thaan the price of the 4WD version. This version is available in the States with a $1,400 US ($2,0000 CAN) difference in suggested list prices. Over the years, I have adapted to the lack of air conditioning by vacationing in late spring and fall. Toronto only needs air condititioning for 2 month out of the year but when you have it...it sure is nice. I hope my sister-in-law reads this because it is for her that air conditioning is only list of options to have. As my grade seven IA teacher, Mr Reeves, preached "Simpler is Better".
The Design Automtion Conference is coming up. It is a industry event put on to showcase developments in CAD/CAM from applications to high-end professional workstations. A funny thing was noted by a friend of mine several years ago, animations had gotten so good that it was difficult to tell the difference between something real and something synthetically generated using 3D visualization. It seems as if the dream that Evans and Sutherland pioneered 2 decades had come to fruition. Advertisers using these tools where the ones who wanted more of the cartoonish qualities traditionally found in 3D generated images. Today, Marilyn Monroe can be resurrected from the dead, a dog can seemingly wink on cue, a two year old toddler can be made to speak like a 40 year english professor... all through the wonders of animation. 3D games (I think) will continue to be in more enticing in the world of 'toon world where colors are super saturated and the images out of cartoon books. The focus on visual processing will shift more and more to that of image processing, compression and decompression to support areas that demand lifelike video...namely DVD, MPEG-1 and MPEG-2. 3D has reach a level of maturity at the rasterization level such that the traditional CPU tasks of lighting and transformation now present bottolenecks. Migrating these transforms might be and effective strategy but the increasing CPU power and data specific instructions of Katmai and 3DNOW promise to keep this architectural decision up in the air. The Rage 128 in my opinion is one of the best balance graphic/video/mutlimedia chipsets on the market today. Anyone hat has sat close to a good quality TFT panel cannot help but comment on the sharpness and clearness of the image offered through a TMDS interface with data coming through digially. The newest panels come in 1280x1024 and can be viewed from a wide viewing angle with little loss in contrast. The cost for 21" units goes well over $5,000 dollars while the 15" 1024x768 TFT panels hover around $1,000 and are predicted to hit the soft spot in the market. Next time you are in a store, take a look at the extended character set character consiting of a happy face inclosed within a cirle...it is absolutely still and sharp. Say goodbye to headaches. Support for this interface by card manufacturers and display adpater companies is becoming rampant. It is especially the case in countries such as Asia with cramped desk spaces and smaller than the 8 foot a side cubicles found in North America.