Sharky Extreme heeft een artikel gepost waarin ze dieper ingaan op de features van de Matrox G400. Hier een stukje over mip-banding op de G400, TNT2 en Savage4:

G400 employs 'per pixel mip-mapping' which gives rise to horizontal 'mip-banding' as shown below. 'Mip-banding' is a set of artifacts that occur when there is an abrupt transition from one mip-map level to another. Trilinear filtering is a technique to remove 'mip-banding' artifacts introduced by mip-mapping. With trilinear filter switched on, 'mip-banding' is effectively removed.

TNT2 uses 'trilinear dithering' which actually introduces dithering artifacts. In addition, the 'mip-band' is sloped, resembling 'per triangle mip-mapping'. Compared to 'per pixel mip-mapping', 'per triangle mip-mapping' requires less computation and is therefore easier to implement. However, the results are not quite as visually appealing as 'per-pixel mip-mapping'.

Savage4 produces a horizontal 'mip-band' that is characteristic of 'per pixel mip-mapping'. Of the three, Savage4's mip-map implementation has the least 'blurring'. Notice the clearly defined tiles at the far end of the narrow path.

Review Zone heeft een review van de 3dfx Voodoo3-3500TV neer gekwakt. Hier een stukje over die leuke TV-in/out friemels van deze kaart:

The Voodoo3 3500TV’s PCB is a very tightly packed affair, with six 2MB SDRAM chips on the front of the PCB and another 2 on the back which brings the total RAM count -as with the other Voodoo3 offerings- to 16MB. The board has two voltage regulators, one near the AGP connector, again like all the other Voodoo3s, and another for the TV tuner. Just below the second voltage regulator is a Micronas MSP3430G sound processor for converting the FM/TV input sound signals, which also drives the two audio interface ports at the top of the card for interfacing with your sound card.

The TV tuner in the 3500TV is Philips-made and takes up a lot of the space on the PCB, it features both coax FM and TV inputs that connect directly to the shielded Philips tuner. Even though the card is definitely more packed with capacitors than any other card I have seen so far, the surprising fact is that the length of the card remains the more or less the same to that of its slower siblings.

This card sports no VGA output, rather a 30-pin Plug & Display output port which connects to the 3500's AV break-out-box which contains a split cable for an analog VGA input. The P&D port is powered by the Xilinx 9572XL controller that also functions as a sort of connection between the video capture and the Voodoo3 chip itself, therefore allowing 3dfx to leave the Voodoo3's current design unmodified while adding support for a FM/TV tuner and video capture, which saves them the hassle of refabricating the chips.

En zo ziet deze jongen eruit:

Onze Singaporese (Singapoerse whatever ) vrienden van Hardware One hebben een review van de Elsa Erazor III met Revelator 3D online gezet. Hier een stukje over die 3D glaasjes:

Well, how does it perform? In general, I must say I’m impressed. I never pictured that I could actually see Lara in her full 3D glory (er..hum..) and get excessive motion sickness playing Descent 3 today!! The glasses does a great job in simulating (or stimulating.. =P) 3D imagery and typically projects the 3D images “into” the screen / monitor.

To give you a clearer picture, just think Descent 3 and its HUD in the ship. The HUD will seem to be overlayed on the monitor screen with the rest of the 3D imagery projected beyond it (ie. Inside the monitor).

Meneer Gonzo van Ars Technica heeft een diepzinnig artikel in elkaar geknutseld waarin hij antwoord probeert te vinden op de vraag wat er zo leuk is aan het spelen van computer games...

Well, the title says it all... gameplay is where it's at. But what exactly is gameplay? Gameplay is a number of things, the way the controls work with your character onscreen, the way your character interacts with the environment, and the way this interaction makes the living breathing person behind the monitor feel. These feelings can range anywhere from pleasure or satisfaction to frustration and anger. A really good game will make you sympathize with your character. Innovations in gameplay often spawn cultish followings for certain games and/or whole new genres. A game with poor gameplay is relegated to the bargain bin a few short months after its release.

Prima voer om te lezen => Ars Technica.

Vince van Extreme Hardware heeft een review van de Videlogic SonicVortex2 op z'n site geparkeerd. Het betreft hier een Vortex2 kaartje dat gebaseerd is op Aureal's Vortex2 SuperQuad reference design, met 4 speaker support en een digitale S/PDIF uitgang. Een probleem met de Vortex 2 kaarten was altijd de hoge CPU utilization. De nieuwe Aureal v.2030 drivers blijken wonderen te verrichten: bij gebruik van 32 DirectSound3D voices ging de CPU utilization terug van 25% naar 2,5%!:

This scenario has been a recurring story in our Vortex2 reviews, with the 32 voice DirectSound 3D benchmark being the Achilles Heel of these otherwise fine soundcards. Since Videologic has updated their drivers from the original 4.06.2016 Aureal reference drivers to a new set based on the 4.06.2030 revision, I decided to install the driver upgrade and see if Aureal had fulfilled their promise to fix this problem.

Simply put, the results were downright amazing. I have not posted DirectSound comparison benchmarks because the scores with the new drivers are so incredibly low as to be inconsequential. I’m talking about scores like a .0627 % CPU Utilization at 32 voices. The Vortex2 problems were always the DirectSound 3D scores anyway, and the new drivers give a whole new picture of the SonicVortex2. We’re talking about scores that decrease by over 90% at the 32-voice level and still bring out some impressive 60-70% decreases at the 8-16 voice levels as well. If you’re currently using the release version of your drivers, I highly recommend upgrading to the newest ones available. As a bonus, the new drivers also included some desktop enhancements, with many functions and settings available by right-clicking on the SonicVortex2 tray icon.

HardOCP heeft een artikeltje gepost over het bezoek dat zij tijdens de Computex beurs aan Abit brachten. Best wel leuk om eens te zien hoe die mammaplanken nou eigenlijk geborden worden...

Voor meer van die leuke plaatjes moet je ff richting dat artikel van HardOCP fietsen.

Storage Review heeft een review gepost van de Quantum Vuurbal Plux KX, een schijf die in veel opzichten gelijk is aan de Fireball KA (7200rpm, 512Kb cache), maar dan met een hogere platter size van 6,8Gb. De performance is errug goed, hij gaat zelfs de IBM / Western Digital schijven voorbij:

When stacked against the Summer 1999 ATA Drive Roundup 7200rpm Editor's Choice, the Fireball Plus KX performs admirably. The KX inches past the Western Digital Expert AC418000 by a margin of 1% in the Business Disk WinMark 99 test run under Windows 95. Results are similar in the High-End Disk WinMark, with the KX sliding ahead of the Expert by 3%. The gap increases in tests run under Windows NT 4.0. Here the Quantum triumphs by 7% in the Business Disk WinMark and a substantial 19% in High-End tests.

Interestingly, in low level tests, the KX is not quite the pioneer that its predecessor was. Though sustained transfer rates hit 23 MB/sec (an all-time high for ATA drives), access time has actually regressed by a couple milliseconds. Even so, the KX still manages to beat its advertised seek time- something that doesn't happen often.

Adaptec ThreadMark 2.0 results place the Fireball Plus KX significantly ahead of the Expert. Here the Quantum drive flies by the WD by about 27% in tests under both Windows 95 and Windows NT.

Check Storage Review voor vergelijkende benchmarks.

FutureMark Software beschuldigt AMD ervan dat ze mbv een voor de Athlon geoptimaliseerd .dll'tje de 3DMark99 scores van de Athlon onrechtmatig hebben gemanipuleerd, zo meldt Semiconductor Business News:

FutureMark Software Corp. Ltd. alleges that AMD is now in violation of the trademark surrounding its 3DMark 99 MAX test suite and related licensing agreement. While FutureMark has not taken legal action against AMD, the company said AMD's actions were "inappropriate."

According to FutureMark, AMD modified FutureMark's 3DMark 99 MAX test by optimizing the DLL test code for its recently released Athlon chip. AMD's own test results demonstrated that the 600-MHz Athlon outperformed the 600-MHz Pentium III by over 30% in one of the two 3DMark tests. Of the 13 test results accompanying the Athlon press release, the 3DMark figures were among the highest in favor of the Athlon chip.

FutureMark's claims are potentially the latest example of "specmanship," the complicated practice of manipulating the equally convoluted melange of testing applications and benchmarks. Benchmarks are a series of tests performed to reduce a chip's performance in a variety of applications and situations into a single simplified number for marketing purposes. The process is further complicated because the mere choice of tests is somewhat subjective.

Ik kwam bij Voodoo Extreme nog wat info tegen over de ontwikkelingen op het gebied van de Linux Unreal/Unreal Tournament server:

Brandon has the server up and running great; there will probably be a public beta soon to help us bash out any bugs that aren't occuring in our little tests here.

The client is fully playable now, but silent...Brandon is porting the sound code now.

Brandon is a madman! The dude hadn't done serious Linux development before, but he had the server up and running after a couple of days (with help from Mike Danylchuk from MPlayer), and the client up and running after he spent a few more days on it.

-Tim

ANP berichtje afkomstig van het forum (thanx Pepijn). Ik weet niet in hoeverre dit al in kranten enzo gestaan heeft, maar Free Record Shop gaat z'n deuren op 26 augustus om middennacht open gooien ivm de release van Tiberian Sun:

AMSTERDAM (ANP) - Veel winkels van Free Record Shop gaan in de nacht van 25 op 26 augustus open in verband met de introductie van het videospel Command & Conquer Tiberian Sun. In 24 zaken van de keten kan de consument terecht om het spel direct aan te schaffen.

Tachtig winkels zijn verhuurd aan de Vereniging Free Raiders, die aan haar leden de behendigheidscursus Tiberian Sun aanbiedt tegen de prijs van het spel. Free Record Shop was een jaar geleden de belangrijkste keten die een groot deel van haar zaken om middernacht opende voor de verkoop van de cd De Bestemming van Narco Borsato. Het openbaar ministerie maakte toen een hele rits processen-verbaal op wegens overtreding van de Winkeltijdenwet.

R. Hermans van Free Record Shop verwacht nu geen problemen. De keten heeft in 160 gemeenten ontheffing van de Winkeltijdenwet aangevraagd, Daarvan hebben er tot nu toe 24 positief beschikt. In de tachtig andere filialen is geen sprake van verkoop. De vereniging Free Raiders huurt de winkels om er een cursus te geven. ,,Dat is onderwijs, geen verkoop'', aldus Hermans.

De wereldintroductie van Tiberian Sun is al enkele keren uitgesteld. Het spel moet de opvolger worden van kassuccessen als Command & Conquer en Red Alert.

Red Orb heeft een 22meg trailer in mekaar geflanst om ons alvast een voorproefje te geven van Prince of Persia 3D (waarvan de demo versie als alles goed gaat de 27e van deze maand uit gaat komen)
De echte die-hard fans kunnen het hele filmpje natuurlijk ophalen... Er zijn ook nog andere filmpjes die je kunt bekijken. Die kan je hierdownloaden.

Ik kwam bij JC dit tabelletje tegen met basic info over toekomstige Slot-A chipsets van AMD:

Chip Version	Tech & Voltage	Max Core Speed	Die Size (pad limited)	# pins
SDRAM, 1P 2X AGP	0.35 micron 3.3V	100 MHz	107 mm^2	492
SDRAM, 2P 2X AGP	0.35 micron 3.3V	100 MHz	130 mm^2	656
DDR, 1P 4X AGP	0.25 micron 2.5V	133 MHz	133 mm^2	553
DDR, 2P 4X AGP	0.25 micron 2.5V	133 MHz
RDRAM, 1P 4X AGP	0.25 micron 2.5V	133 MHz	107 mm^2	492

The first one on the table looks like the northbridge that is currently being used in all the Athlon motherboards. Seeing that there is a .35 micron chip that supports 2 processors, makes me think that dual Athlon motherboards won't be too far away. Also, if they already have die sizes for their DDR and rambus chips means that they probably already taped out prototypes at least for these.

Happy Cat (wazige japanse site) heeft benchmarks gepost van een viertal Athlon borden (MSI-6167, Gigabyte GA-7IX, Biostar M7MKA en FIC SD11). Hoewel de tekst in onleesbaar japans is, zijn de benchmarks wel begrijpbaar voor ons simpele polderbewoners. Check ze hier.

Microsoft gaat een stap verder in de instant messenging oorlog en gaat de code van het MSN Messenger protocol vrijgeven. Blijkbaar verwachten ze dat er gekken op deze aardkloot rondsurfen die voor noppes MS code gaan kloppen...

Microsoft today launched its most powerful weapon yet in the messaging war with America Online.

In a move reminiscent of its struggle with Netscape Communications for browser dominance, Microsoft said it will publish the protocol for its MSN Messenger service.

Publishing its messaging service code means Microsoft could rely on third parties to help bolster its position in the messaging wars. Microsoft also hopes to gain allies among Internet service providers and the Internet Engineering Task Force (IETF), a group that is partly responsible for setting up standards for how messaging software communicates.

Bloomberg today reported that three online companies, including Prodigy Communications, back Microsoft's messenger service. Another instant messaging company backing Microsoft is Tribal Voice, which today said its 5 million PowWow users will have access to MSN Messenger by the fourth quarter.

"Microsoft tends to do two things at once. Not only is the first part to work the standards issue, to try to make their own technology the accepted standard, but the second issue is opening up their protocol to gain eyeball shares," said Joe Ferlazzo, an analyst with Technology Business Research.

Meer info vind je in dit nieuwsartikel van C|Net News.com.

News.com heeft een artikeltje over de afwachtende houding van veel bedrijven ten opzichte van de release van Windows 2000. Tegelijkertijd wordt Linux steeds vaker toegepast in bedrijfssystemen:

Companies are increasing their use of the upstart operating system while delaying deployment of Microsoft's latest operating system for businesses, according to International Data Corporation (IDC).

[...]The survey also indicated that responding companies have no plans to immediately implement the Windows 2000 operating system, Microsoft's latest and greatest update to the former Windows NT. Microsoft is putting the finishing touches on that software.

"Organizations of all types and sizes indicated they plan to wait anywhere from six to 18 months before beginning wide-scale implementation of the new Microsoft [corporate] enterprise operating system," the report said.

"The No. 1 reason given by respondents for the delay in adopting Windows 2000 was technical stabilization of the operating system. A full 50 percent of respondents said the technical stabilization of Windows 2000 would cause them to delay implementation," the report added.

"Past issues with first-release operating systems from Microsoft have caused organizations to rein in their Windows 2000 deployment plans," said William Peterson, research manager for IDC's Client Infrastructure Software programs. "This is not to say Windows 2000 will not be a success...IDC believes Windows 2000 will succeed--over time."

Okee, de titel wekt misschien te hoge verwachtingen, maar volgens info van AMD Zone zijn ze bij Abit aan het evalueren of ze toch een Athlon plank gaan bouwen:

Thanks to Roger Cormier for letting me know that he shot an e-mail off to Abit asking about them making an Athlon board. He said the reply he got back was that they are evaluating making a board,

May the force deliver een Abit Athlon bord .

Afgelopen tijd waren er geruchten dat Apple de firmware van de nieuwe B&W G3's dusdanig had gemankeerd dat deze een upgrade naar de snellere G4 onmogelijk zouden maken. Mac OS Rumors weet ons te vertellen dat er inmiddels een third-party G4 upgrade kit is die een work around biedt voor dit probleempje. G3 gebruikers hoeven dus niet bang te zijn voor een doodlopend upgrade pad :

Earlier this week, a Rev. B Yosemite was upgraded with a prototype third-party G4 ZIF card right in front of our eyes. After some initial problems setting the proper bus speed timing (the card came set for a 66MHz bus clock, rather than the Yose's 100), the card booted perfectly at 550MHz. There was a small amount of instability in using the machine -- but no less than was experienced when using the same card in a Beige (Gossamer) G3 minitower.

The point here is simple: whether or not Apple has created some kind of firmware hack in the hopes of restricting the viability of G4 upgrades before it has released its own G4-based Macs, at least one third party -- who remains anonymous at their request -- appears to have worked around it. MOSR has also been told that the workaround is relatively simple, and will most likely be discovered by other third parties soon....if they haven't already.

So, Blue G3 buyers can rest easy; your machine will not be a dead-end, and in fact is one of the most upgradable Macs ever made. Enjoy it!

Het door Micro$oft ontwikkelde secure audio format (WMA) is koud twee dagen nadat het officieel geintroduceerd was alweer gekraakt. Muziek in WMA formaat kan dus net zo makkelijk doorgekopieerd worden als in MP3.

Hier een paar snippers uit het originele bericht van Wired:

The cracking file intercepts the audio data stream as the file is being sent to an output device -- such as a speaker -- according to Kevin Unangst, lead product manager in the streaming media division at Microsoft. The utility functions similarly to cracking programs Audiojacker and Totalrecorder, intercepting the data and re-saving it without the restrictions.

Riva Extreme heeft een artikeltje gepost over het Oxygen Project van de faculteit Computer Science en het Artificial Intelligence Lab van de M.I.T. Klinkt best wel kewl wat die slimbo's daaro in elkaar aan het dokteren zijn:

This ability to expose the raw hardware on a chip - including the wires - to the software compiler will transform the world of computing. By using the free logic gates to direct and store the signals that run through the chip's wires, the compiler basically customizes the wiring for each application. This is a radical departure from existing architectures, in which software controls the chip's logic operations - basic functions such as "ADD" and "SUBTRACT" - but not the chip's wiring. In contrast, Raw allows the software to program the wires, the microchip's most valuable resource.

Tom Jensen van 3D Spotlight heeft een review van de Elsa 3D Revelator glasses op z'n site gekickt. Tommie heeft het brilletje maar meteen uitprobeerd op z'n pa en zusje, en die schijnen het nogal heftige gevonden te hebben...

I didn't even look at the other screenshots and I just went straight to my "Games" folder to open up Halflife. I started up a new game and watched in amazement, just like Disney World.... I quickly ran upstairs to tell my family about my awesome discovery, but nobody was too enthusiastic, I grabbed my little sister and dragged her into my room and shoved the glasses on her face.

I pointed at the screen and she sat there and drooled. She said, "This is just like a movie" and "What is this test that everybody says that I am late for?" After causing an enomilee in the testing chamber we journeyed through the giant underground laboratory for a good hour with her screaming as "face suckers" leapt at the screen. Blood guts and other body parts looked so much groovier when flying at you.

My dad came downstairs to see what all the screaming was about and I just slapped the 3D glasses on his face... He watched for a good ten to fifteen minutes saying he just came down to say good night which would have taken about thirty seconds under any other circumstance. His reaction was, "Now this looks real!", and "What was that thing that just jumped at your face?" These glasses made a big impression on a computer illiterate girl and a penny-pinching father.

Guru3D heeft een artikeltje in elkaar gedraaid over de T-Buffer technologie, die kort geleden door 3dfx werd aangekondigd. De T-Buffer technologie verzorgd hardware acceleratie van vijf (vanuit Glide ondersteunde) special effects: depth of field blur, motion blur, soft shadows, soft reflections en full-scene anti-aliasing. Guru3D vraagt zich af of het hier om écht vernieuwende technieken gaat, of dat het een marketingmanouvre is die tot doel heeft de aandacht van nVidia's NV10 weg te leiden:

I was pretty shocked to read about the NV10's power and 3dfx's T-Buffer announcement in so few days' time. Nvida's next card will make ultra-high polygon models more likely, so the in-game objects will get closer to the quality of prerendered cutscenes. 3dfx's response is to give the cutscenes nicer looking camera-like effects, but apparently not to improve the models themselves. Unfortunately for 3dfx, high-poly models make the difference between an arm with a bulging bicep and a texture-mapped rectangle. Without improvements in the polygon count and speed, the T-Buffer looks like an "all icing and no cake" approach to 3D graphics.

[...] That's why the timing of the T-Buffer's press release is so suspect. In my corporate-conspiracy-filled mind, 3dfx obviously wanted to steal some of Nvida's thunder with a press release of their own, and it's a good short-term strategy. Even with a comment of minimal consequence, any announcement from an industry leader like 3dfx is bound to be repeated more often than the less-famous competition.

Check Guru3D voor het complete artikel (best wel interessant IMO).

Overclockers' Workbench heeft een review gepost van de Logitech QuickCam Home, een simpele webcam die op de USB poort aangesloten dient te worden:

It's small and compact and it is really useful. Everything that I need to capture pics and video is all here and even though they are simple to use, they provided me with lots of fun and the "easy to use" and "fun" factor, combined with "stylish" makes a good web cam to buy! The only drawback for such a fun and useful web cam is the software that is being provided. I spoke with the support team and they told me that there would not be a software revision in the near future. Well, since the software plays an integral part of the package, it prevents me from giving the full accolades that it deserves so...

Ik kwam bij The Register dit best wel geinige plaatje tegen :

Toevallig liep ik aan aantal weken geleden rond op Manchester Airport en daar hing ook zo'n tv'tje met een Windoos (3.1!) error .

Lost Circuits heeft een artikeltje neer gemikt met tips over het tweaken van de Athlon BIOS settings. Verderop in het artikel staat ook een interessant stukje over de busspeeds/overklokbaarheid van de Athlon:

Unlike SS7 or Slot1 platform, the Athlon chipsets do not employ one master clock from which the different other clocks (PCI, AGP etc.) are derived but more than 10 individual high speed clocks that are interfacing the system logic and the CPU. Changing one of the clocks, e.g. to raise the FSB, therefore will not allow to run the system stable because of asynchronism between the different clocks. In order to address higher bus speeds, therefore all clocks will need to be reprogrammed. While this is technically possible, according to sources within AMD it would also involve independent reprogramming of both the CPU and the system logic to achieve stability (We cannot guarantee the accuracy of this statement but it sounds plausible enough to post it).

Epic's GreenMarine heeft op het GT Interactive forum een posting geplaatst over de status van Unreal Tournament:

Heh. Actually that 1% is just nailing the last stuff down. Netcode has been really great for a couple weeks now. We are just going through the GT bug reports fixing what they report and playing a bunch every night and fixing bugs that pop up.

Specifically, for example, tonight we had a bug where I started a server with InstaGib and no one had a weapon. (Making the game unplayable.) The reason this bug popped up is because we recently rewrote the mutator ordering code to make it a little nicer. Because its possible to select mututally exclusive mutators, problems were occuring. The rewrite fixed some of the problems, but caused others.

Tim is expanding the installer to handle a second CD (which we will need). Steve has been fixing misc bugs, I've been working on misc bugs and the

Linux port, and Jack has been working with Yang to implement a HTML based server administration system.

There isn't any one thing in particular holding us back, just a bunch of minor issues.

Check deze GT forum thread voor meer info.