S3 heeft vandaag bekend gemaakt dat z'n Savage4 chip geoptimiliseerd zal worden voor de Streaming SIMD Extenstions van de PIII. Hier de info van de The Register:

According to John Brothers, S3's VP for architecture and software, optimising Savage4's ICD and Direct3D drivers for the SSE allows the host PC to "feed transformed, clipped, lit triangles to or hardware at a rate closer to what the Savage4 architecture is capable of rendering".

Sony heeft z'n eigen alternatief gebouwd voor MP3. Het heet 'MagicGate' en legt een zware nadruk op de beveiliging:

Sony's entry into the SDMI race is codenamed MagicGate. It's designed for consumer electronics kit, including Internet-enabled hi-fi separates and systems, and portable players like the Rio. Sony is also working on a version of MagicGate, called OpenMG, to protect tracks downloaded into a PC.

At the heart of both systems is an authentication system hard-wired into silicon. The device checks that downloaded files are correctly encrypted and were received from a MagicGate-compliant source. The MagicGate chip could easily be added to music players, said Sony. For PCs, it would be built onto an add-in card which would automatically encoded MagicGate files stored on the computer's hard disk or other storage media.

Inmiddels zitten de leden van het Secure Digital Music Initiative (SDMI) in LA gezellig te bakkeleien over welke technologie de standaard moet gaan worden. Meer info bij The Register.

Be gaat gratis BeOS licenties verstrekken aan OEMs, die het (naast Windoos) kunnen installeren op hun systeempjes:

n an article published on the Be Web site, Gassee grumbles that, for all Microsoft's claims at the DoJ trial that there are viable alternatives to Windows and thus the company doesn't have a monopoly, PC OEMs prevented from taking these alternatives -- specifically Linux and the BeOS -- on board because of the way they often interpret Microsoft's Windows licensing terms.

That's not an unreasonable point, and Gassee's solution, in the short term at least, is to encourage PC vendors to install the BeOS gratis and ensure users can see it's there as an option when they boot their computers.

Gassee's goal, he claims, is "put a crack in the wall" of the Microsoft monopoly, the idea being that if the Great Satan of Software kicks up a stink about what an OEM has done, its arcane licensing terms will be revealed to the world for what they are.

Meer info vind je bij The Register. Zie ook dit artikel op de site van Be Inc.

The duitse Magazine für Computer Technik (C'T) is een actie begonnen tegen de plannen van de Europese Unie om spam te legaliseren. Hier wat info van The Register:

The editor-in-chief of c't, Christian Persson, is calling on all Net users to sign a petition demanding EU officials take a closer look at the proposals. The petition will go live on Monday 1 March.

While the petition doesn't call for commercial spam to be outlawed completely, it does demand that users should not be bombarded with junk e-mail unless they specifically request it.

Moreover, it says that users should not be discriminated against simply because they have chosen not to accept spam.

Die petitie kun je hier ondertekeken.

Adrian heeft een review van de IBM Deskstar 14GXP in z'n Rojak Pot gestopt. Effetjes roeren op 7200rpm en je kunt in alle stilte van je UDMA soep genieten:

If there's something one would notice immediately after moving from the Medalist Pro to the Deskstar 14GXP is its silent operation. Now, any drive running at 7200rpm is sure to produce a certain amount of noise. The Medalist Pro was silent while reading but it was absolutely noisy while seeking. However, the Deskstar 14GXP managed to do everything quietly.

Perhaps it's just my system but I can't differentiate the sound of the Deskstar 14GXP from that of the CPU and hard disk fans. Now, that's real quiet. If I close the casing, hmm... total silence.

Review-Zone heeft wat lopen spelen met een AMD K6-III 400 en 450. Hier wat info over 3DNow vs SSE:

he K6-III has 8 64-bit (= 2 x 32-bit) wide single precision packed CPU registers, which allow to do two single precision FPU operations per pipeline simultaneously. The Pentium III, meanwhile, uses 8 new 128-bit (= 4 x 32-bit) wide single precision packed CPU registers to compute four single precision FPU operations per pipeline simultaneously. 

This is not exactly as lopsided as it may sound. The Pentium III is only able to issue one SSE computation for every two clock cycles, because Intel did not see fit to widen the entire data pipeline of the Pentium III to 128 bits to fill the registers. So, the fact that the Pentium III has 8 new 128-bit registers and 70 new SIMD instructions Vs the 21 SIMD instructions of the K6-III does not necessarily mean that the PIII’s SSE implementation provides twice the performance provided by the K6-III’s 3DNow! implementation. What’s more, the Pentium III is helped mostly by the memory streaming architecture. Unlike the K6-III’s 3DNow! implementation, however, the Pentium III’s SSE enables it to use MMX and FPU instructions in parallel.

Meer stuff vind je in de review van Review-Zone.

GA-Source heeft een artikel in elkaar gepleurd over het volgens hun steeds maar toenemend aantal bugs in computergames. Persoonlijk vind ik dat wel meevallen. In de goeie ouwe DOS tijd was ik 2 uur bezig om m'n pc zondanig te configgen dat-ie een spelletje onder DOS4/GW op m'n krappe 4Meg 486 doos aan de praat kreeg met ondersteuning voor m'n Gravis Ultrasound. Startte het spul dan eindelijk toch op, dan liep 't geheidt ergens vast tijdens de intro of ging-ie plat tijdens het laden van een game... Mocht Windhoos ergens een voordeel hebben dan is het wel dat 't met DirectX tegenwoordig vrij simpel aan de praat te helpen is.

Anyway, geen gezeur meer, hier wat info uit het artikel van Gamer's Alliance:

The wonderful thing about games in general is that they continue to become more and more technologically, visually, and thematically marvelous. Games now have depth, and feel, and immersive sound, graphics, and interaction with the environment. In many ways, they push the technological envelope like no other software application does, and with this cutting-edge technology, there is a price. That price is usually the cost of hardware upgrades, but it's now being said that buggy, sometimes broken, games are the cost as well. Is this too high a price?

In addition, developers often say that they're under pressure from the gaming public to continually increase the complexity of games. Gamers want better sound, more and better visuals, better AI (artificial intelligence) in their opposition, more involved and intricate plots and adventures, faster and smoother online play... and that the cost of all these advancements is -- you guessed it -- more unstable and quirky games in general. Is this too high a price?

Chris van Voodoo Extreme heeft een review van de Hercules Terminator BEAST gebakken. Hier wat stuff over de standaard BEAST vs de BEAST Supercharged:

he original BEAST has 8MB of SDRAM clocked at 100MHz, and a 250MHz RAMDAC. It supports resolutions up to 1600x1200x16bit@85Hz and 1280x1024x32bit@75Hz. The Savage3D’s claim to fame is S3TC, a texture compression scheme that allows huge textures to be rendered at insane speeds with little to no loss in image quality - 200MB of textures in one scene are possible...let’s see AGP do that. It also provides hardware motion-compensation for DVD playback, but since I don’t have a DVD drive, and that doesn’t relate to gaming, I’ll let you take that at face value. The same goes for TV-out.

The SuperCharged is effectively identical to its older sibling, except for the major fact that it sports fast SGRAM instead of slower SDRAM. This allowed Hercules to safely increase the clock speed from 100MHz to 120MHz - much closer to the Savage3D’s rated specification of 125MHz. How much of a difference does this make? Well, the benchmarks tell it better than I ever could. Just to make things interesting, I got a pre-production mini-Lasagna fan from Tennmax and stuck it on the SuperCharged, to find out just how well this card suits us overclockers (as we all know, it’s one hot little bitch with its stock heatsink). So without further ado:

Daikatana Nation heeft Bobby "Xcalibur" Pavlock, leveldesigner bij Ion Storm, zitten interviewen. Hier een stuk:

When creating a level, do you draw on personal experience? Like do you see architecture in real life and try and recreate it in your levels? Or do you design totally from your imagination?

Some of both actually.  Everywhere I go and everything I do, I find myself thinking,"Hey, that would look cool in a level!", whether it be real life architecture, or pictures, or movies, or books, or whatever.  There are cool ideas to be found everywhere, you just gotta be looking for them.

Brian Hook heeft z'n .plan weer eens zitten updaten, dit keer met info over hoe ze bij id Software bepalen welke videochips ze gaan supporten:

So there are basically "chips we have to support", "chips we like to support", "chips that aren't terribly relevant but we'd like to support anyway", and "chips that we could give a damn about". Chips we HAVE to support include the Voodoo, Voodoo2, and Rage Pro because they are out there in HUGE numbers. Chips we like to support are  NVidia RivaTNT and ATI Rage128. Chips that aren't terribly relevant but that we'd like to support well include Rendition V2x00, Savage S3, and Permedia2. And chips we could give a damn about are things like TriTech Pyramid3D, Virge, and Number Nine Ticket To Ride.

We can't comment on Permedia3 or Voodoo3 since we are not in possession of any of these devices. And for those of you still with PowerVR PCX2, you're probably out of luck since I don't believe NEC/VL/PowerVR are actively updating those drivers (although the hardware itself should work fine with our vertex lighting model).

On a somewhat related note, several Matrox G200 users have asked me about the support for that board. The G200 is a very good accelerator with good performance and great image quality, however they haven't gotten on the ball with the OpenGL thing. When they have a high quality, high performance OpenGL driver then I'll be able to recommend their hardware enthusiastically. Their reliance on the D3D wrapper for so long did not endear them to us.

Fullon3D heeft een interview gepost met Paul Crossley (S3 ventje) over de S3 Savage4. Hier alvast een vraag:

Savage 4 is supposed to be playing on the same field as TNT2, PVR250, Rage 128 GL, Permedia 3 and to a lesser degree, Voodoo 3.

Where in this pecking order would you place your product and for what reasons? Who should be buying a Savage 4 board?

Answer: We're better than all of them combined =)

Seriously, it depends on what you're looking for and how much you want to spend. Take Voodoo3 for example, it will cost more than a Savage4 board, but it won't deliver features such as texture compression, AGP texturing, true 32-bit color, support for textures larger than 256x256, support for 32MB of memory, hardware DVD acceleration, etc. In pure frame rates, it may be faster in some benchmarks and a few "twitch" games, but does that make up for all its other shortcomings? We don't think so.

Meer Savage4 info vind je bij FullOn3D.

Op RIVA 3D staat dat er geen 3DNow! ondersteuning in de NVidia Detonator drivers voor de TNT zit. Dat is voor mensen met een K6-2 en een TNT natuurlijk heel vervelend. Hier ff het stukje :

I had the opportunity to speak with Kevin Wagner from AMD today, and here's his fill in for all you AMD/TNT fans out there. The "predator" drivers as they have been referred to in rumors may or may not be the actual name for them, but AMD and NVIDIA are vigorously engaged in optimizing the TNT drivers for both D3D and OpenGL. This should result in much improved performance in all D3D and OpenGL games. The initial timetable for release will be in two to three months, but of course, unforeseen development delays may move that back a bit.

Thresh's Firingsquad heeft een preview van RollerCoaster Tycoon geplaatst. Hier een knip en plakwerkje :

How many times have you gone to a theme park like Disneyland or Magic Mountain and told yourself "I can run this place better than they can!" Or, "they should have put that drink stand over there." Perhaps maybe, "this ride would be PERFECT with just one more loop." Well now's your chance to make those tough decisions and a whole lot more, with Hasbro/Microprose's Rollercoaster Tycoon. You'll build your park from the ground up, and hire staff members, set prices, build and design rides, and arrange concession stands all around you park.

It sounds like a really unique idea but Bullfrog actually pioneered the amusement park simulation with "Theme Park." I was hopelessly addicted to Theme Park, because it had many of the same basic principles as Rollercoaster Tycoon. You had to micromanage almost every little detail of your park. If people became hungry, build more stands. Pricing for food and park admission were the source of your income, while your money was drained away on staff salaries, park expenses, etc. Rollercoaster Tycoon takes this premise to the next level, with even more features, micromanagement possibilities, and of course, better looking graphics!

De review staat hier.

Corsair heeft nu PC133 SDRAM. Hier een stukje uit de press release.

FREMONT, Calif., Feb. 25 /PRNewswire/ -- Corsair Memory Inc., the industry leader in high density DRAM Dual In-Line Memory Modules (DIMMs) is launching its family of PC133 SDRAM modules. The initial module densities offered will be 64MByte, and 128MByte non-ECC modules. These modules will use the latest PC133 RAMs from only the top tier OEM manufacturers like Micron and Samsung. These modules will be available in early March 1999.

Many of the major Chipset companies such as Ali, VIA, OPTi, and SiS are currently launching products that support the 133MHz SDRAM. The PC133 SDRAM modules are likely to be used in motherboards supporting Intel's (INTC) Pentium III and AMD's (AMD) K7.

``Corsair was one of the first module manufacturers to design PC100 SDRAM modules. Today, most of our high density PC100 modules have been qualified on the major motherboards from Intel, Tyan, Asus, Abit and Supermicro. Once again, as a leader in the SDRAM module industry, Corsair intends on being one of the first third party manufacturers to offer PC133,'' said Richard Hashim, Director of Memory Products at Corsair.

Sharky Extreme heeft een "first glimpse" van het spel Redline. Hier de geripte text :

Accolade looks to be on the right course for 1999. With Slave Zero already in the pipeline and Redline nearing completion, they've firmly pitched their intentions on the PC front. The second of today's look at Accolade's involves Redline, their 3D first person shooter. After having played the recently released Redline demo, the game is perhaps best described as being 3D combat simulation due to the 3D engine's versatility allowing for the driving of vehicles or the more conventional on-foot style combat. Indeed, we will be taking a closer look at the full game later on this month- in particular the multiplayer capabilities that allow for up to 12 players across a LAN or the Internet (via M-player support). With over ten different vehicles to choose from and some 50 weapons to boot, it's already clear that mincemeat is certainly on the menu as far as Redline is concerned. Stay tuned for a preview and in the meantime take a look at the shots below.

En die screenshots staan hier

Er zijn weer nieuwe biossen voor de Abit PII moederplankjes. Ze zorgen ervoor dat de PIII goed herkend en ondersteund wordt.

To assure potential Pentium III purchasers, Abit issued a press release announcing complete support for Intel's new processor in their existing lineup of BX mainboards. BIOS flashes for their popular BH6 and BX6 2.0 are available from www.abit.com.tw/html/download.htm. BH6 owners will want to grab update JJ while those with a BX6 2.0 should download JH.

Shary Extreme heeft nieuwe Voodoo3 3500 benchmarks. De eeste kaart die ze hadden was niet stabiel dus ze moeten hem onderklokken.

Earlier this week Sharky Extreme was unfortunate enough to receive a Voodoo3 that could not operate at the 183MHz clock speed (a bad apple) and failed to survive rigorous testing during our in-house preview. It keeled over and died on us raising rumors on whether or not 183Mhz memory yields were all that they were cranked up to be. Although the board in question is currently undergoing a post-mortem at 3Dfx Interactive, we've been given another Voodoo3 3500 to test and thankfully the initial signs are good. It's alive and well after two days of non-stop testing/benchmarking. This new board is different from its predecessor in that it uses EtronTech 183MHz SGRAM instead of the more expensive SEIMENS (a case in point where you don't always get what you pay for) that came with the last one. This version looks to be a little closer to what the final retail Voodoo3 3500 will look like as it does not have an on board fan but merely a heat-sink. So with the board alive and well we've updated our benchmarks for you at the correct 183Mhz speed. Obviously there wasn't much of a difference from 179Mhz (in some cases none at all). At no time did we have to turn the clock speed down to 179Mhz or less and the board is currently purring in our number 5 test machine. Certainly a good sign.

Voor de grafiekjes moet je hier zijn.

Voodoo Extreme heeft weer een stukje van Grandmaster B. Hij heeft een theorietje over waarom 3Dfx nog steeds geen fatsoenlijke OpenGL drivers voor de VoodooII heeft :

Call me a genius, but the only reason that I can see for 3Dfx not releasing "good" OpenGL drivers is because they think it's for their own good. If they were to release good OpenGL drivers for their Voodoo line of 3D accelerators they would be hindering their own API (Glide). So in order to keep Glide alive and its performance on par with OpenGL performance, they only made the 3Dfx OpenGL drivers "only good enough" to run OpenGL games on their own cards. They obviously don't want OpenGL overshadowing Glide.

If a good 3Dfx OpenGL driver were to be released, then it seems that it would clearly outperform their Glide API and therefore lose developers who code specific to 3Dfx (Glide). Said developers would switch to OpenGL and 3Dfx would lose the only advantage they still think they have, a proprietary API (Glide) that is exclusively less supported everyday. Maybe they don't realize it yet, but they're not the only show in town now. So why doesn't 3Dfx let go and put out an excellent OpenGL driver so owners can get better performance out of their 3Dfx hardware?

En hij heeft nog gelijk ook denk ik...

Ace's Hardware heeft een verhaaltje waarin wordt beschreven hoe je een CPU geschikt krijgt voor een hoog aantal MHz's. Hier ff een introotje :

Have you ever wondered why an Alpha CPU is able to reach such high speeds? Why the P6-Core of intel will be able to reach 800 MHz and more and why the K6 core (most probably) never will? Why does AMD claim that the K7 is designed for high clockspeeds?

Well, it is time to find out. This article is the beginning of a series of articles that will explain you what makes a fast processor. We'll take a good look under the hood of all those number crunching beasts. Let me tell you, there is a lot to discover. And don't worry, the purpose of this article is to explain, not to bombard you with technical terms. Let us get on with it!

Lees hier de rest...

Hiero kan je een patch halen voor het spel Viper Racing. Deze dingen zijn er gefixed :

Patch notes:

• Fixed: Replay Controls Inaccessible in Higher Resolution Modes
• Force feedback effects boosted
• Mirror now scaled to be appropriately sized for the video mode.
• Disable the intro movie by using the "-nointro" flag.
• Memory leaks fixed
• Bug caused when cars were still "teleporting" when the race ended, has been fixed.
• "Garage lock up" fixed
• Many more bug fixes and enhancements

Hier vindt je trouwens een review van de demo van dat spel.