Got Apex? heeft z'n i820 moederplank review van een update voorzien. Dankzij een nieuw drivertje blijkt de i820 (met PIII op 533MHz) nu opeens wél in staat om z'n oude BX vriend (op 550MHz) bij te houden:

Talk about big increase. The new drivers push the frames per second way past the old scores and even past the BX at 550Mhz scores at most places. We're talking an increase of up to 53.0% (Demo1, 640x480, High Quality) over the old scores! Things suddenly aren't looking quite as grim.

PC Paradox heeft een behoorlijk dikke review van de Matrox G400 neer gemikt. Naast de gebruikelijke EMBM screenshots en benchmarks worden we verblijdt met info over de overklokbaarheid van de G400:

This card is clocked by default at 125 Mhz Core and 166 Mhz Memory. The MAX version is clocked at 150 Mhz Core and 200 Mhz Memory. Remember, 6ns memory maxes out at around 180-190 MHz 5ns memory on the other hand, maximizes at around 215-230 MHz So you now can approximate some numbers for the G400 vanilla ( running with 6ns SGRAM) and G400MAX (running with 5ns SGRAM). Well my testing gave me an approximate speed for maximum overclocking with the G400 Vanilla of about 138 to 145 for the core and 183 to 195 MHz for the Memory. For achieving a maximum of 145-195 you will almost always need extra cooling (remember the G400 MAX has active cooling which is one of the reasons it can achieve the 150 MHz setting). It is also recommended that you turn your CAS latency down one notch before any overclocking. You can achieve this with the "MGA Tweak" which can be found at www.matroxusers.com. You can also you can find a lot of tweaking programs and information on new drivers there also.

Verder heeft ook 3D Spotlight het net verrijkt me een G400 review.

WinMag.com heeft een lap tekst bij elkaar geraapt waarin ze zich afvragen of wij (die Windoos altijd kopen ) afgetroggeld worden door meester Bill. Hier een stukje uit dat artikel:

First there was Windows 98, which bore the same $109 price tag for those who already owned Windows. (The company quietly discontinued the ability for MS-DOS-only owners to upgrade to Windows.) About a year later came Windows 98 Second Edition. Next year Microsoft plans to release Millennium, yet another upgrade to the Win9x line. From the late 80s through early 90s, Microsoft's mainstream operating system received an upgrade roughly every two to three years. Now we're on a yearly cycle. Is Microsoft building value into each successive version of its operating systems twice as fast as it used to? Actually, no. They're just giving us less, and asking us to pay for it anyway.

Why is this happening? Well, like most public companies, Microsoft must keep growing in order to please its shareholders. But it goes deeper than that. Microsoft is fighting a multi-front war: America Online, Linux and Open Source, the Department of Justice, up-again-down-again PC sales, and corporate America's disillusionment with seemingly endless upgrades. Add to these problems the bald fact that, while consumer demand is there, the Windows PC still isn't anywhere near to being an "it just works" consumer environment. Think about it: People complain about programming VCRs.

PC Week heeft een wedstrijd uitgeschreven waarbij het de bedoeling is om de twee WinNT en Linux servers van www.hackpcweek.com te hacken:

In a major test of the security of Linux and Windows NT, PC Week Labs today threw down the gauntlet to Internet hackers, challenging them to break into a Web site, http://www.hackpcweek.com, to try to crack each or both of the operating systems. The site goes live today for a one-month trial.

The site contains near-identical systems, one running Windows NT with Internet Information Server (IIS) and the other running Red Hat Linux 6.0 with Apache as its web server. PC Week Labs created similar classified-ads engine applications running on each system. The challenge is to break into the site, mark up the home page and steal user information from the classified-ads engine.

"Security is extremely important in the Internet environment and both Microsoft and the Linux community, via Red Hat, boast that their operating systems are secure," noted PC Week Labs Director John Taschek.

PC Week Labs' will track the number of attempts for each operating system and report the results in an upcoming issue of PC Week. Additionally, PC Week will issue a prize to whomever hacks into the site first. The challenge terminates when the first hacker accomplishes any of the test challenges. Winners will receive computer-equipment gift certificates of up to $1,000.

Zie GA-Source voor het volledige artikel.

Nieuwe zet van Bill: span de consument voor je karretje en mobiliseer publieke steun tegen de rechtszaak die de amerikaanse overheid tegen Microsoft heeft aangespannen. Computable heeft er een artikel over:

Microsoft heeft gekozen voor een rechtstreekse PR-aanval als beste verdediging in de door de Amerikaanse overheid aangespannen antitrustzaak. In de hoop Microsoft-gebruikers te mobiliseren voor een ondersteuningscampagne, heeft de software-onderneming het 'Freedom to Innovate Network' (FIN) in het leven geroepen. Op de website heeft het concern hiervoor het adres http://www.microsoft.com/freedomtoinnovate/default.htm aangemaakt.

[...] Met het Freedom to Innovate Network slaat de onderneming uit Redmond duidelijk een andere weg in. Er zijn weliswaar eerder campagnes geïnitieerd die waren gericht op het verkrijgen van steun van gebruikers, maar die deden allemaal hun best om de link met Microsoft zoveel mogelijk te verdoezelen. Met het FIN windt Bill Gates' onderneming er geen doekjes om: dit is een Microsoft-campagne. Het logo bestaat uit een Amerikaanse vlag waarbij de 'stars' zijn vervangen door een PC.

Ook de boodschap van het FIN laat aan duidelijkheid niets te wensen over: de overheid bemoeit zich met zaken waarmee ze zich niet moet bemoeien en de handelswijze van Microsoft is ingegeven door legitieme zorg voor de belangen van de consument. Als je Microsoft aanvalt, open je de aanval op de fundamenten van een moderne, geïndustrialiseerde democratie. Door criticasters van Microsoft wordt op deze 'nieuw

Om spontaan -neigingen van te krijgen...

Volgens Variety.com heeft Sony plannen om een film te bouwen rondom de Redneck Rampage games van Xatrix:

Sony and Tall Trees said the picture will be a comedy, and won’t necessarily be as violent as the game itself (in which the characters wield shotguns and crowbars) — a sensitivity born of the recent Columbine, Colo., tragedy.

“It won’t be gory, grizzly or graphic,” Topping said. “We’re sensitive to the current climate, but we’re also interested in making it a commentary on violence and how destructive it can be.”

Indeed, the comedic potential is high: The trouble begins when aliens, seeking out the most powerful brain on the planet, discover it belongs to the redneck’s prized porcine, which they abduct on its way to the country fair.

Speedy 3D heeft een review gepost van het Acer Labs Ergo 61, een ergonomisch gespleten toetsenplankje met ingebouwd touch pad:

Now I will cover the most important issue on a keyboard, its comfort. I found this keyboard to be incredibly comfortable. It far outpaces any other keyboard on the market in this area. I have not ever used a keyboard that has been more comfortable than this one. Never. The ergonomic design was perfect for me. Some people might find the keys to be a little big, but since I have big hands, the keys were great for me. That’s another thing, if you do have big hands then I would particularly recommend this keyboard to you! I Love this Keyboard. No Acer didn’t pay me to make any of those statements, they just made a great product. Obviously they care about their customers, unlike some companies, which I will leave unnamed. So far we have an ergonomic keyboard that is comfortable and easy to install.

Ik kwam via AMD Zone bij deze wazige japanse site, waar men een Athlon-600 op 800MHz heeft draaien door de multiplier te verhogen naar 4 en de L2 cache divider te verlagen naar 1/3. Hier een plaatje van het systeem (vraag me niet waar de Athlon gebleven is )

Slecht nieuws voor AMD, want volgens the Wall Street Journal zal Gateway stoppen met het bouwen van AMD pc's. News.com heeft er een artikeltje over:

Citing "people close to the matter," the paper said Gateway, which in February unveiled its first personal computers using AMD chips, will not continue with purchases after its current models are replaced. The paper said sources told it AMD believes it lost the Gateway business because Intel began offering Gateway more favorable terms than it had previously. The paper said the move is other sign of rival Intel's renewed gains in the price-competitive, consumer PC market.

But the Journal said a Gateway spokesman declined comment other than to say the company continues to sell existing PCs using AMD chips and won't preclude AMD from future bids. [break] The Register trekt hierop de eerdere geruchten over een deal tussen Gateway en Intel maar weer eens uit de kast: [/break] Earlier this year, Gateway declined to use AMD's Athlon K7 in its products. That is believed to have happened after Intel offered Gateway more favourable deals on its microprocessors, estimated to be worth a cool $20 million.

Buy.com heeft de Elsa Erazor X voor 238 dollar in de prijslijst staan. Deze prijs valt op zich mee (de G400 Max en Viper V770 Ultra kosten 250 dollar), alleen vraag ik me af of dit inclusief rebates is (waar wij hollanders weinig aan hebben). De normale retail prijs is $289. Buy.com doet overigens niet aan buitenlandse orders.

PVR-Net heeft een review gepost van de Videologic Neon 250. We hebben al eerder reviews van dit kaartje gezien en telkens ontbrak een direkte benchmark vergelijking met andere videokaarten. Ook in het artikel van PVR-Net is dit weer het geval, dus het blijft vooralsnog moeilijk te zeggen hoe de Neon 250 zich exact verhoudt in vergelijking met een TNT2 of V3:

In general, performance is very good. With a peak of nearly 3,800 at 640x480 and 800x600, it's up there with the TNT2. However, the card definitely sees a performance drop for 32-bit colour and higher resolutions. Something which isn't shown here is the standard deviation for frame rates. I should certainly point out that significant slowdown (from ~40fps to 17-20fps) is evident in certain "in-game" scenes of 3DMark99 MAX. When larger, open areas are encountered. Whether this is a limitation of the hardware, or of the drivers remains to be seen. Image quality in 3DMark was very good as all DirectX blending modes are fully supported. Also, Neon250's filtering for bilinear, trilinear and anisotropic tests, seems to be very clean. However, there was further noticeable slowdown for the anisotropic tests compared with standard n-linear processes.

So in terms of benchmarked Direct3D performance, Neon250 delivers exactly what VideoLogic promised a few months ago -- speed up there with today's leading 2D/3D accelerators.

The Reverend Pulpit vond het nodig om onze vriend Tim Sweeney (Epic / Unreal) lastig te vallen met een vraagje over de T-Buffer en FXT1 technologie van 3dfx:

Supporting T-Buffer for antialiasing will be either zero or one lines of code, depending on how 3dfx exposes that in Glide, so that will be easy. We're not planning on implementing any funky motion blur effects, just the standard antialiasing once the Voodoo4 is available.

FXT1 is definitely superior to DXTn, and I'd love to see the rest of the industry support it, since it's a nice clean superset of DXT1. I think it's excessively cool that 3dfx released the spec freely to the world. There's so much crap going on with hardware makers suing each other that it's nice and refreshing to see that level of openness -- not only that, it's also good for business, because a hardware feature is worthless if only one 3D hardware maker is able to implement it.

Sharky Extreme heeft een review van de Turtle Beach Montego II Quadzilla op het net gepleurd. Het betreft hier een geluidskaartje met de welbekende Aureal Vortex II chipset, zoals die bijv. ook in de Diamond MX300 en Xitel Storm Platinum gevonden kan worden:

We listened to the Quadzilla on a Cambridge SoundWorks $99 four speaker + subwoofer setup under a variety of audio circumstances along with a Diamond MX300.

We feel in our particular acoustic environment there was a slight difference between the MX300 and the Quadzilla with the Quadzilla producing clearer high notes in various situations (with MP3s particularly) while the MX300 was more effective in reproducing sounds in the middle to low range of the frequency scale.

Both offered excellent playback of several popular audio formats including the soundtracks of several DVD discs when used in conjunction with a DVD-ROM drive.

So as not to create any confusion, the MX300 and Quadzilla were more alike in their sound reproduction than they were different over the days that we compared them. To illustrate this, when given "blind" tests, personnel who were asked to identify which sound card they were listening to could not separate the Quadzilla from the MX300 on a regular basis. The cards are that close.

3D Spotlight heeft info los weten te peuteren waaruit zou kunnen blijken dat de GeForce 256 licht (enkele weken) gedelayed is:

I have learned from a manufacturer that NVIDIA has extended the GeForce256 NDA to the end of the month, that means we won't be seeing any GeForce products until early next month, if you remember some manufacturers had planned to get their products out to stores before the end of the month, not a big deal really =).

Got Apex? vroeg Xitel of het al een beetje wil lukken met de reeds lang beloofde EAX ondersteuning voor de Aureal Vortex II:

Wondering when Aureal is going to release EAX drivers for their A3D2 cards? Well, here's the response from Xitel:

"EAX is coming. Aureal has finished it and it will be released soon. Biggest change in the drivers is the inclusion of the EQ tab which allows you to adjust the imput to which the EQ is applied. Also, CPU usage is a lot lower. MIDI is happier. 4 speaker support functions perfectly. Basically, a slew of fixes. This set includes all of the changes incorporated into the previous Aureal reference driver releases and a couple of others."

BDW: een paar dagen geleden berichtte i-XBT dat EAX support eind dit jaar zou komen.

Over bij Site Sieve staat een klein artikel over Anti-aliasing, een feature waar 3dfx nogal op geilt. Hier heb je een stuk uit het artikel.

3dfx is aiming for 1024x768x32bit @ 60fps with full screen anti-aliasing. 2000x2000x32bit color will probably take roughly the same amount of fill-rate as 1024x768x32bit color with AA (Depending on the implementation). The reason 3dfx (among others) is aiming for 1024x768x32bit color with AA is because 1024x768x32bit color w/AA will run on every system with a monitor made during the last millenium. If 99% of user's monitors supported resolutions of 2000x2000, 3dfx would be aiming for 2000x2000x32bit without any form off antialiasing. Since this resolution (movie quality) is not available because output hardware cannot deal with it; 3dfx is taking a hollywood quality rendering and formatting it to fit the average users screen. In order to best replicate the 2000x2000x32bit color frame; Anti-aliasing is required. Therefore, 1024x768x32bit is NOT enough...unless AA is implemented.

FPS3D.com heeft een artikeltje gebakken over de huidige situatie bij Intel, waar men de ene naar de andere wazige stap zet (Rambus dat trager is dan PC133, i820 chipset die de oude BX niet bij kan houden, i810 chipset die niet met PIII's overweg kan, PIII-600 die dankzij voltage tweaking stabiel wordt gehouden enz.). Hier een hap uit dat artikel:

Probably the #1 biggest of Intel's problems right now. They successfully convinced nearly every memory manufacturer that Rambus RAM (known more commonly as RIMMs or RDRAM) was the future of memory. Now this would be a great thing, because RDRAM is great on paper. However, everyone who has tested RDRAM so far has not only found it to be up to 6 times as expensive as PC133 RAM, but also far slower. We're talking 15-30% slower with the only variables being RAM and Rambus-supporting Intel Camino motherboard; and the Camino motherboard came direct from Intel!

After repeatedly denying that they were even remotely considering or testing PC133 in-house, Intel finally came out of the closet just over a month ago saying that they were considering PC133 support in their Camino chipset, as well as Rambus support. The catch? If it does end up supporting PC133 it will probably be through an interpretor chip, which means a nasty performance hit.

MacOS Rumors heeft de specs van de Powerbook G4 op een rijtje geplaatst:

- 400-550MHz PowerPC 7400 processors. 1MB backside L2 caches standard at 200-275MHz.

- 100MHz main bus, 64 bits wide.

- Standard 14.1" flat panel display with improved clarity, brightness, and wide-angle viewing. Rumors of an available 15.3" display with 1280x1024 resolution remain unconfirmed.

- Approximately 25% thinner than Lombard, and 1-1.5 pounds lighter. This will bring it firmly into the Ultrathin class. Horizontal dimensions of the computer are limited by the flat panel display, but could be reduced slightly. The new enclosure is expected to continue to be black, but could include some translucent components in Dark Grey and "Ice."

- RAM type remains unclear; 100MHz-compliant SO-DIMMs (the Small Outline SDRAM modules used in current iMacs, iBooks, and Powerbooks) are expensive and hard to find at present, but this could easily change between now and WWDC. Another possibility is that these machines could use traditional DIMMs -- unlikely, but it could help reduce costs and improve performance for the Powerbook line, and is by no means an impossibility inside the Powerbook enclosure, which is expected to contain enough room for at least two full-size DIMM slots if Apple considers it a useful feature.

- 2X (133MHz) AGP on-board with a second-generation RAGE Mobility-based graphics chip and 8-12MB of SGRAM. Support for a second independent analog display is expected.

- UltraDMA/66 storage bus with high-performance hard disks starting at 8GB and ranging up to 18GB, 7200RPM versions. [break] Verder wil Steve er FireWire in stoppen, eventueel ter vervanging van de SCSI controller. Klinkt allemaal erg leuk. Tevens bij MacOS Rumors info over MacOS 9 en Altivec performance verbeteringen: [/break] Also, the first reliable report we've received of Mac OS 9 running on a dual-processor G4 machine (a Yosemite) suggests that running against 8.6, performance of common system tasks is already accelerated dramatically by Altivec (OpenGL-based games and rendering applications simply screamed with even a single 450MHz G4, between 4 and 10 times faster than the same apps running on a 450MHz G3) -- and with two 450MHz PPC 7400's, we saw OS 9 outrun 8.6 on Quake 3: Arena by over 40%; in fact, it was likely higher than that, but we seemed to hit a frame rate ceiling at around 120fps. The dual G4 under OS 9 stayed near that rate even in the most complicated environments (a dozen-plus netgame players, weapons fire and explosions virtually covering the field of view, all high-quality options on at 1024x768 pixels). Even the most optimistic of us were shocked by how high these numbers were -- something that could bode very well indeed for the next generation of Mac gaming.

Het volk van Vision4D (semi-wazige japanners) hebben een plaatje van de Intel Coppermine op hun site gezet:

Verder vinden wij rond dit plaatje nog een inspirerend stukje tekst:

ÁÕ€ÇÒÁà»ç¹ä»äŽéÊÙ§·ÕàŽÕÂÇ€ÃѺ·Õè·Ò§ Intel ÍÒššÐà»ÔŽµÑÇ Coppermine ¢Í§µ¹àͧã¹àŽ×͹˹éÒ·ÕèšÐ¶Ö§¹Õé šÒ¡áËÅ觢èÒÇǧ㹢ͧ·Ò§àÁ×ͧ¹Í¡ášé§ÇèÒ Intel Çҧጹ·ÕèšÐà»ÔŽµÑÇ Pentium !!! Coppermine ã¹àŽ×͹µØÅÒ€Á¹Õé âŽÂ Pentium !!! ÃØè¹ãËÁèÅèÒÊØŽ¹Õé·Ò§ Intel šÐÁÕ¡ÒûÃѺවâ¹âÅšҡ 0.25 à»ç¹ 0.18 äÁ€Ã͹ á¶ÁÂѧÁÕ L2 Cache ¢¹ÒŽ 256K on-die áÅÐÂѧ·Ó§Ò¹ ·Õè€ÇÒÁàÃçÇàŽÕÂǡѺ CPU áÅзÕèàÃÒÃÙé¡Ñ¹ÍÂÙèáÅéÇ¡ç€×ÍšÐãªé FSB ·Õè 133MHz ÍÕ¡ ŽéÇÂ

Proost...

Ik kwam bij Slashdot bij dit interessante artikeltje van Linux Journal, over Billies strategie om Unix / Linux bij de wortels aan te pakken, te beginnen bij de universiteiten:

Want to murder a tree? You can hack away at the branches and leaves, but you probably won't do much damage. It's much easier to simply take out your Bowie knife and cut a ring all the way around the tree's base--just deep enough to cut off the flow of sap. Most of it's concentrated near the surface, just beneath the bark. No sap? Dead tree.

Now, don't go out and start slaughtering our tall leafy friends. I'm making a point here. If Microsoft's plans for colleges and universities pan out, the tree-death scenario is precisely what's going to happen to UNIX-like operating systems in general (and Linux in particular).

Here's the argument in a nutshell. As you'll see, UNIX's long reign in the halls of academia nourishes the Open Source phenomenon. But cash-strapped college and university computing administrators are cutting deals with Microsoft--and not surprisingly, the Redmondians are pressuring schools to move their servers to NT. The likely result? UNIX (and Linux) could well become an endangered species on campus. Such a development could cut off the nourishment that's helping to keep the Open Source movement alive.

The Adrenaline Vault heeft een update over z'n super handige release date list gekieperd. Hier als voorbeeld alle releases van deze maand:

20 - After Dark Hoyle Bundle - HAVAS INTERACTIVE

20 - AMA Superbike - MOTORSPORTS SIMULATIONS

20 - Close Combat Trilogy - MICROSOFT

20 - Creatures Adventures - MINDSCAPE

20 - Disciples - GT INTERACTIVE

20 - Gabriel Knight III - SIERRA

20 - Poker Night with David Sklansky - INTERPLAY ENTERTAINMENT

20 - Sega Rally Championship - SEGA ENTERTAINMENT

20 - Virtual Chess Academy - INTERPLAY ENTERTAINMENT

20 - West Front: Sea Lion - TALONSOFT

21 - Chutes and Ladders - HASBRO INTERACTIVE

21 - Expert Pool - PSYGNOSIS

21 - Rage of Mages 2 - MONOLITH

21 - Rogue Spear - RED STORM ENTERTAINMENT

21 - Scrabble II - HASBRO INTERACTIVE

21 - Tanktics - INTERPLAY ENTERTAINMENT

21 - Tetris - HASBRO INTERACTIVE

21 - Tiger Woods PGA Tour Collection - ELECTRONIC ARTS

22 - Homeworld - SIERRA

22 - Warcraft Battle.net Edition - BLIZZARD

22 - Lego Land - LEGO MEDIA

23 - Trivial Pursuit II - HASBRO INTERACTIVE

23 - Furby - HASBRO INTERACTIVE

24 - Chestmaster 7000 - MINDSCAPE

24 - Splinter - ELECTRONIC ARTS

24 - Catz 4 - MINDSCAPE

24 - Dogz 4 - MINDSCAPE

28 - Armored Fist 3 - ELECTRONIC ARTS

28 - Driver - GT INTERACTIVE

28 - Grandmaster Chess - INTERPLAY ENTERTAINMENT

28 - Interplay's 15th Anthology - INTERPLAY ENTERTAINMENT

28 - NHL Championship 2000 - FOX INTERACTIVE

28 - Panzer General 3D Assault - MINDSCAPE

28 - Silver - INFOGRAMES

28 - Space Invaders - ACTIVISION

28 - Spec Ops: Ranger Gold - RIPCORD

29 - Dragon Magazine Archives - TSR

29 - Forgotten Realms: Atlas - TSR

29 - Heroes of Might & Magic III Expansion - 3DO/NEW WORLD COMPUTING

30 - Abomination - EIDOS INTERACTIVE

30 - Freespace 2 - INTERPLAY ENTERTAINMENT

30 - Hoyle's Board Games - SIERRA

30 - Hoyle's Casino - SIERRA

30 - TRIBES Action Pack - SIERRA

30 - Ultimate Collector for Sports Cards - MINDSCAPE

Bij de weg: Duke Nukem Forever staat op de derde van januari - moge dat waar zijn. Check AVault voor het volledige overzicht.

Paint Shop Pro lijkt een redelijk populair alternatief voor Photoshop te zijn (in ieder geval bij onze lokale held Danny ). Webmaster Resources heeft een review van PSP 6, dat onlangs gereleased werd. Laat ik maar meteen richting conclusie fietsen:

Paint Shop Pro 6 is without a doubt the best program in the low-end graphics market. However, Jasc software should work on the usability of the program, as the learning-curve is rather steep and comparable to that of Fireworks or PhotoShop. The number of steps required to do certain tasks isn't as low as it should be. If you have an extra $90 I would recommend the more-professional Fireworks 2 tool. But if you're on a budget, for the money, you won't find anything better.

Planet Quake heeft een interessant artikeltje gepost over het verschil tussen pixels/sec en texels/sec. In feite zegt de schrijver dat getallen in texels/sec onzin zijn en dat alleen pixels/sec gebruikt zouden moeten worden. Uiteindelijk gaat het immers om de framerate, dus het aantal pixels dat per seonde naar de frame buffer kan worden weggeschreven. Het aantal texels (texture elements) die per seconde op een pixel geplakt kunnen worden heeft niets met de pixel fillrate te maken. In het artikeltje wordt 3dfx dan ook hevig afgezeken, want zij gebruiken getallen in texels/sec en dat getal komt bij de Voodoo2/3 niet overeen met de pixel fillrate, want de Voodoo doet aan dual texturing, maar heeft in tegenstelling tot de TNT en GeForce (die meerdere pixels per kloktik kunnen renderen) slechts 1 pipeline.

And one final point, the most important statistic is not how fast you can fill the trucks (megapixels) or how many stickers you can use (megatexels). The statistic that your customers would really want to know is how many truckloads can you deliver per hour (i.e. your frame rate). You can try to fool people with all that hyperactive megatexels but the frame rate is the one statistic that matters more and tougher to abuse. The title for "king of speed" is based on frame rate, not fill-rate. 3dfx can keep the title of "king of fill-rate", the "king of speed" is the more important title. The state of the drivers can greatly affect the real performance of the card and we measure that in frames per second, not megatexels. As an old Quaker said many times, we have all been down this road before. A poor driver can really kill an awesome card. If GeForce256 ends up being a flop (and it might), chances are you can blame it on the drivers.

Verder vraagt de schrijver zich af of 3dfx e.a. de hoge fillrate claims voor hun next-generation kaarten waar kunnen maken, aangezien memory bandwidth op dit moment al een erg belangrijke limitatie blijkt te zijn. Dit zal alleen maar erger worden naarmate de (theoretische) fillrate toeneemt. Check Planet Quake voor het complete verhaaltje.

Volgens The Adrenaline Vault verloopt de amerikaanse introductie van de Sega Dreamcast voorspoedig, desondanks zijn er een paar kleine probleempjes. Zo blijkt er een vrij groot percentage foute CD's op de markt geplant te zijn:

Unfortunately though, not all titles are fairing as well. There are numerous reports, including several here on staff, regarding defective Dreamcast games and even some systems. Sega has admonished some of the problems and setup hotlines and updated their website with information regarding the troubled titles.

There appear to be a number of titles affected, especially those manufactured by Midway, Activision and Sega themselves. Specifically, consumers should be aware that up to 10 percent of retail versions of Midways' Ready 2 Rumble, Activision's Blue Stinger and Sega's NFL 2000, Sonic Adventure and Web Browser discs have been identified as having possible defects. Reports from Sega blame the problem on the duplication process associated with the discs themselves. Symptoms from the problematic games are frequent crashes, stuttering sounds and defective video.

Sega remains adamant that none of the problems are hardware related, but only time will tell. If you are having trouble with your system or any of its software, you should call Sega at 1-800-USA-SEGA, Activision at 1-888-402-89511 or Midway at 1-877-213-0579 and report the problems. While the most effective method for replacing your defective discs is to return them to the retailer from which you purchased them, Activision has even setup a no-hassle exchange routine for the defective copies of Blue Stinger.

De PowerMac G4 is natuurlijk een fantastisch snelle computers, alleen schijnt er een nogal wazig geurtje omheen te hangen. Salon Magazine heeft er een artikeltje over:

A recent review of the G4 on the MacInTouch site for Apple news commended the new machine for being "pleasantly quiet," but also noticed that the computers had an unpleasant odor when turned on. The reviewer wasn't the only person to notice the G4 stink: This little bug has also been documented by the Apple Tech-info library, and filed under "Odors May Be Present Short-Term."

According to Apple, a plasticky smell will occasionally waft up when the computer is turned on and allowed to warm up to its "operating temperature." As Apple placates concerned owners, "the odor is detected when the product is new, similar to odors generated from new carpeting or a new car. In most cases the odor will dissipate over a short period of time." But the only quick solution, the site says, is to turn your computer on and leave it running for 24 hours or more, effectively burning off the smell -- and filling your room with a wallop of a stench that could last several days.

Ah, the smell of hot plastic in the morning ...

Hmm, let's burn some plastic ...

Bij AMD Zone toch weer een klein beetje hoop op een Abit Athlon bord:

We may come out with one in the future, but for now there are not enough K7s in the channel market. They are mostly all going to the OEM market and ABIT does very little OEM. This is why we are sitting out the first wave.

Thanks for your interest,

Nic
ABIT

Matrox heeft een aantal met EMBM (Environment Mapped Bump Mapping) opgeheerlijkte Dungeon Keeper 2 screenshots op z'n site geparkeerd. Twee voorbeeldjes:

Uiteraard is dat allemaal beter te zien wanneer de plaatjes groter zijn. Fiets richting Matrox voor meer detail...

Gamer's Alliance heeft een overzichtje gebouwd van de nieuwe features waarmee SinterBill ons met de release van DirectX 7 gaat verblijden:

- Visual Basic type library for DirectX.
- Visual Basic DirectX samples.
- DirectDraw support for stereo display devices.
- Direct3D support for hardware transform and lighting engines and cubic
- environment mapping.
- D3DX library providing increased ease of use for Direct3D.
- DirectSound support for hardware voice management and selectable 3D algorithms.
- DirectMusic support for DLS-2 hardware.
- DirectMusic Producer support for hybrid notation and editing of sequence tracks in segments.
- DirectInput support for WDM drivers on Windows 98.
- Enhanced game controllers control panel.
- DirectPlay performance enhancements.
- New sample applications that illustrate the use of the new features.
- New DirectInput force editing tool.
- Improved documentation, including. dynamic filtering for language-specific information

Volgens GA-Source kan de officiële release ieder moment plaatsvinden. Hopelijk gaat dat snel gebeuren, want Epic zit er op te wachten in verband met de release van de Unreal Tournament D3D demo.

Voor de bezitters van een super hete 12.000rpm Hitachi SCSI kookplaat heeft Speedy 3D een review van de 3DfxCool hddHO harddisk cooler neergemikt. Verwen je PC met een heerlijk gezonde stroom frisse lucht...

Besides that it comes with 2 20mm thick fans. That provide us with some serious horse-power to cool your entire hard drive effectively. After getting all that for $30 some people would call it a “bargain.” Now I get to talk about how this cooler performed, which is what really matters. By using a heatsink to take the heat off the top of the drive and 2 fans to blow the hard drive’s hot air out of the case, it has put it self in a good position. This cooler cools the whole hard drive effectively. The first that I have come across that can lay claim to that. Reducing the entire temperature of the whole hard drive, that is. With no cooling at all the hard drive was hot to the touch after I ran several benchmarks on it for 30 minutes. When I installed the 3dfx hddHO cooler it was almost “cool” if you will, to the touch. When I say this I am talking about the whole hard drive, top, bottom, and sides. That is definitely a solid reduction in temperature, but as always no actual measurements will be released until the HDD Cooler Roundup.

FIC heeft een nieuw BOS voor z'n SD11 Athlon plank gebakken, waarmee het mogelijk wordt om busspeeds van 100, 120 en 133MHz te gebruiken. Een AMD Zone lezer heeft het ff uitgeprobeerd, hier z'n ervaringen:

I just downloaded the new SD11 Bios that you posted. I then went to FIC's website and found a copy of AMI's latest flash Utility "flash818". I put it in a folder named "Bios" in the root of C: and then rebooted my machine while hitting "Crtl F5" so no TSR's or anything would load. I ran the flash util and all seemed to go well. It accepted the new program and said all was a success.

So I exited the program and rebooted again while I cringed at the thing hoping it was a legit bios copy. It posted just fine and I immediately noticed that my pc speaker now makes noise while it goes through the memory count. After that It prompted me about a checksum error... no big deal Ive seen that on tons of new Bios's. So I hit F1 and went into the Advanced chipset settings or whatever and there it was, a new selection for FSB!! The settings were 100, 120 and 133MHz which coinsides with what FIC says their new bios will have. I immediately selected the 120 setting then saved it and exited. To my extreme dissapointment the thing would not even post =(. So I cleared my bios via the jumber and tried again but this time with the 133 setting hoping maybe that 120 possibly had a PCI or AGP timing problem.

Again I got nothing... no post or anything but the sound of fans and Harddrives spinning. Once again I cleared the bios and tried some more settings. This time I selected 120MHz FSB and fooled around with the PCI clocks, memory timing. Memory timing shouldnt be much of a factor in my case because I have a stick of PC133 memory with Micron 7.5ns chips. After that I got the same results as before so I went back and restored to the settings that I know would work.

After the bios was working to normal specs I let it boot into Windows which went just fine but It was obvious that something was different because Windows had to make hardware driver additions which were a PCI bus and a parrallel port. (The parallel port cant be counted in this case because I had it previously disabled and forgot to re disable it in this bios.) So far everything appears to be performing normally.

My guess about the overclock problem is that either this CPU (Athlon 500) wont go to 600MHz without a voltage tweak or at all, or maybe the bios still needs some work. I plan on trying 2 other Athlon 500's that I can get my hands on and ill keep you posted on how it goes. Thanx

Matthew [break] Jammer. Nog even wachten dus op de Athlon plankjes met meer busspeed settings tussen 100 en 133MHz.

En nu we het toch over de Athlon hebben: Micron zal geen Athlon systemen gaan bouwen. De reden hierachter (tenminste, wat ze er zelf van zeggen, wie weet hangt Intel hier ergens boven...) is nogal errug lame. Het op een juiste manier spellen van het woord Athlon blijkt trouwens ook al een hele kunst te zijn... [/break] No we do not intend to offer the Athalon because of many important reasons: most of todays software is written for Intel architecture. There is only a noticable difference in high end games. the motherboard selection for the Athalon is almost nil. The top three direct computer manufacturers Dell, Gateway and Micron, are not going with Athalon-- that should say something.

Carl Dufford Hays III
Micron Web Sales