AVL schrijft dat er bij Thresh's FiringSquad een verslag te vinden is van de Consumer Electronics Show onlangs gehouden in Las Vegas. In het eerste deel wordt vooral aandacht besteed aan Creative met haar nieuwe MP3 producten, zoals de Nomad II en de Nomad Jukebox, waarvan die laatste een niet misselijke 6GB aan opslagcapiciteit biedt (zie pic onderaan).

The NOMAD Jukebox has 6GB of storage built-in thanks to a 2.5 inch notebook hard drive. The Jukebox is shaped like a portable CD player; when you first see it, you're tempted to look for a non-existent CD eject button. The unit weighs in at just under a pound, 14 ounces with four AA batteries. The batteries only support 3-6 hours of continuous playback depending on battery type, but the Jukebox does come with an AC adapter. Like the NOMAD II and NOMAD II MG, the Jukebox uses a USB interface for PC connectivity.

The Jukebox also features an onboard digital signal processor (DSP) and has support for 4 channel surround sound. You will be able to download new DSP algorithms, security features, and playlist generators from www.nomadworld.com. The unit also has a five minute shock protection buffer. The Jukebox will be available this spring with a sub-$600 retail price.

Creative plans on adding several more Jukebox features and options. One feature will allow you to use the Jukebox as an external hard drive. You'll be limited by the speed of the USB port, but having 6GB of portable disk space could come in handy. And yes, Creative also plans on making a jukebox car power adapter. "Um… no officer, that isn't a radar detector. I swear. Want to hear some rap music I downloaded from the internet? It's NWA -good stuff."

AVL schrijft: "Het Australische AussiGamer heeft een review gepost van de 3dfx voodoo3 3500, met TV tuner en video in/out, concurrentje voor de Matrox Marvel G400. Goed te gebruiken volgens deze Aussies.":

The Voodoo3 3500TV is one impressive piece of silicon. Even though it was publicly released some time ago, the card is still a very valid choice for gamers, and with added extras such as DVD, TV Tuning and heaps of other multimedia enhancements, you can't go wrong. 3dfx have some of the most solid drivers around, and most any recent 3D game is sure to run fantastically with the 3500 installed. I was very impressed with the card, and would recommend it to anyone looking to upgrade their old 3D accelerators, or even if your a semi serious gamer dabbling with video editing.

Meer van deze review vind je hier.

insane meldt dat de 800MHz Pentium III en Athlon inmiddels verkrijgbaar zijn in Japan: "De PIII-800 kost $912 (voordat die prijs haal je hier nog niet eens een PIII-700) en de PIII-750 heeft een prijs van 856 dollar. Een 128Mb PC700 Rambus RIMM is daar te koop voor $757, nu nog maar 4x duurder dan SDRAM":

The Japanese Pricewatch site that reports on Tokyo's busy goings on in Akihabara, is reporting that the first Pentium III 800s Coppermine processors are appearing for sale, at the same time that AMD's Athlon 800MHz chips are also going on sale.

According to Japanese reader Battlax, chips are available in two stores. The prices for the 800MHz and 800MHz EB Pentium IIIs, both available in SECC2 flavours, are $912. Intel changed its branding of Pentium IIIs last autumn and is set, in the near future, to offer a somewhat bewildering display of Pentium IIIs in different packaging

Check de The Register voor het volledige nieuwsartikel.

aFreak@beer.com" rel="external">DaFreak schrijft: "Eindelijk is het raadsel van de CSC keyspace, die inmiddels nabij een onmogelijke 125% is geraakt, opgelost: er werden verkeerde blokjes op onze koetjes losgelaten. Ongeveer 25% van de keyspace moet overgedaan worden, da's genoeg voor een week. Dus gassen met die koetjes! DPC 4ever!".

David McNett heeft een verklaring van de CSC problemen gepost:

Behind the scenes, we've been viewing the rapid approach of 100% CSC keyspace completion with mixed feelings. While it's exciting to see how fast we are all tearing through the keyspace, a bit of pre-100 nervousness is to be expected.

When we passed 90% on the statsbox, we decided to explain the impact of work duplication (as detailed in dbaker's plan the other day) because it looked like we might hit an apparent 100% keyspace completion. Anticipating the chance that our work re-verification system might eventually lead to an "over 100%" state on statsbox, we made the announcement explaining its impact.

At the same time, we stepped up our investigation into computer errors and malicious vandalism. Throughout the project, we have kept watch for anything out of the ordinary that might lead to a problem in our key processing. As the big 100 drew near, our suspicions began to rise...

So did the suspicions of many of our users, but until now we've had nothing concrete to tell you. This has changed. During the the latter part of our search, a clue smacked us in the face and eventually led us to a serious problem, one which affected a section of CSC work performed to date. We received a success packet on the master but were unable to use the reported key to decrypt the ciphertext. By backtracking the reported success, we uncovered an issue which affected the keyspace assignments done by the keymaster. Soon we had isolated it to a specific version of the keymaster code. The bad news: for a period of time, the keymaster was generating blocks with corrupted contest information, making those blocks invalid. Thankfully, the error only affected the CSC contest and not the RC5-64 contest data.

However, the minor error had a major impact on CSC. Roughly 25% of the CSC keyspace was affected by the block corruption. Fortunately, we were able to flag the invalid blocks as "untested", and start dumping them back to clients immediately.

What this means is that a quarter of the CSC work done to date was invalid and needs to be re-tested. Rather than 100% complete, in actuality we're only 75% through the keyspace -- about a week away from true completion, at current rates. It's very fortunate that our user base is so powerful that we can recoup the losses well before this contest comes to a close. It should also be noted that this error actually occured a few weeks ago and was in-advertantly corrected before we knew there was a problem. Because of this, there is no need to dump your buffers; the bad blocks are no longer in the pipeline.

We will be amending the percentage complete reported by the stats server to reflect our true keyspace completion. Participants will still receive full credit for their completed blocks, regardless of their validity. In the near future we will be supplying the stats server with the information it needs to properly discount the effects of any reverification work in the reported percentage.

We realize that a setback of just over a week is quite annoying in a short contest like CSC, but it's entirely possible that the winning key is working its way back to the keymaster right now.

To have passed this problem into the production network is seriously embarrassing. Simply put, we have failed in our responsibility to the users running our client, and we will do everything we can to ensure this never happens again. Despite coordinating some of the most powerful networks on earth, distributed.net is still run by those fallible human being types. We obviously need better testing procedures.

We already have a reliable system in place for client and core coders, to test the proper function of new and beta client code revisions. However, it's now quite apparent that we're lacking an equivalent sandbox (or process) to test changes in the keymaster code and network structure. With recent major revisions to the keymaster code (for OGR, CSC, and others) it's clear that this situation cannot continue.

As a result, distributed.net has committed to building a keyserver testing environment, complimenting our existing client-tests area. For a very long time the keymaster code was static enough that such a provision wasn't really necessary, but as we undertake more projects this will become less and less true.

We'll also take this time to step back and carefully chart a more reliable route through our development cycle, ensuring that errors and mistakes can be caught long before they would enter production.

On behalf of distributed.net, I'd like to apologize to our users -- and also thank you. The magnitude of your support has given us all a good chance at recovery, despite enormous keyspace fallout. We have learned our lesson, and we're very, very sorry. Now, with your help... let's finish this thing.

David McNett
nugget@distributed.net

Techweb bericht dat nVidia werkt aan eigen audio core. Of het ontwerp wordt toegepast in een standalone chip of geïntegreerde audio/video chipset is nog niet zeker, maar de laatste optie lijkt waarschijnlijk. nVidia zou daarmee terugkeren naar haar oorsprong, toen het bekend werd met de NV1. Deze geïntegreerde audio/video chipset werd gebruikt in de weinig succesvolle Diamond Edge 3D (thanks eenTurk voor de tip):

By designing audio products, the company can continue expanding beyond the PC and follow a path it struck through a recent development deal with workstation maker Silicon Graphics. Executives this week said they plan to bring an audio core to market late this year or early in 2001, but have not yet committed to either a discrete audio chip or an integrated graphics/audio IC.

[...] "An integrated multimedia subsystem is going to happen, combining things like 2-D, 3-D, and audio," Hara said. "The key is what form factor we bring it out in."

[...] According to analysts and industry executives, an audio core now requires only about 40,000 to 80,000 gates, less than a third of a low-end graphics chip. Analyst Jon Peddie of Jon Peddie Associates, Tiburon, Calif., said on a leading-edge 3-D chip, the audio core would probably fit under the copyright symbol in the logo. Moreover, audio standards, save for perhaps LucasFilm's THX, are used and can be duplicated in PC and set-top products.

Network World Fusion schrijft dat de Taiwanese Microsoft site afgelopen woensdag door een hackertje werd gede-faced. De hacker maakte waarschijnlijk gebruik van een bekende exploit in Remote Data Server, die niet dicht gepatched bleek te zijn op de server van Microsoft Taiwan (thanks wildhagen voor de tip):

On Wednesday, an unknown attacker broke into Microsoft's Web server for public information in Taiwan, leaving the message: "Hi, Bill, welcome to the real Y2K bug."

The hacker, who signed himself as "inferno.vr," apparently exploited a known Web server vulnerability in the Microsoft Internet Information Server that hadn't been fixed. The server has hosted at the Taiwanese company, AUNET, a Microsoft spokesman says.

[...] Still investigating the incident, Microsoft believes at this point that the hacker exploited a hole in Microsoft's Remote Data Server software used to let Web developers update information to the Microsoft Web server, IIS.

"We patched this back in July '98, and then again reminded customers in July '99 to install the patch," Sohn says. Ironically, the Web server on which Microsoft kept its own information in Taiwan hadn't been fixed.

RoEL komt vertellen dat er op de duitse site van MSI een nieuw BIOS voor de MSI-6167 Athlon moederplank neergehaald kan worden:

Version 1.5 ( 01/2000 ) selbstextrahierendes Archiv, 141kB incl.Flash-Tool.
- behebt Probleme mit USB Geräten

Graig Campanaro van Sharky Extreme heeft een DVD Player Buyer's guide gepost, met uitleg over het DVD gebeuren en met reviews van een aantal standalone DVD spelers. Hier een stukje over het verschil tussen Digital Theatre Systems en Dolby Digital 5.1:

DTS is a surround sound alternative to Dolby Digital 5.1. In order to replicate this sound, buyers will need either a separate DTS decoder or an audio receiver or DVD player with a built in DTS decoder. As with Dolby Digital 5.1, there are six discrete sound channels for playback through dedicated speakers--two front, two back, one center, and one subwoofer.

DTS uses a higher bit rate than Dolby Digital 5.1 does, and has a lower compression rate that requires more space on the DVD disc itself. It's not uncommon for the DTS version of a DVD movie to be offered on two separate DVDs due to the size of the audio versus the same movie needing only one disc in its Dolby Digital 5.1 version.

Because of the higher bit rate, many claim that the sound is clearer with DTS, allowing it to better replicate the sound of the original film soundtrack. However, because of the size limitations, DTS discs do not often have all of the extra features that Dolby Digital 5.1 discs may have such as interactive menus, a director's audio commentary track, added footage that was edited out of the theatrical release, etc.

Atomic Geek heeft een review rond de Yamaha 8424S SCSI Rewriter in elkaar geknutseld. Een kieskeurig apparaat dat ondanks goede specs (4Meg buffer, 8X write, 4X rewrite en 24X read) in praktijk erg gevoelig bleek voor de kwaliteit van het gebruikte type recordable:

All in all, this burner has been a very emotional up and down roller coaster ride for me. The hardware installation went fine, but the software installation concerns me. Adaptec Easy CD Creator is the one of the only software programs that can support the 8424S. This will change though in the next couple of months, but for now it's the ONLY one. The 8X burning is very sketchy. You need the buy 8X rated CDRs, and you have to burn from the hard drive in order to be successful. You cannot use a cdrom reader (even my Plextor 32 which is also SCSI and is considered to be one of the best readers out there, failed this test) to burn on the fly. No one tells you all of this though. I had to find it out the hard way.

Gert schrijft dat Alex van Kaam versie 4.13 van Motherboard Monitor online heeft gemikt. Hieronder de info:

Released MBM 4.13, it fixes a few small errors, including one with the shared memory.

Je kunt het proggie hier downloaden.

Casio heeft drie nieuwe horloges geïntroduceerd. De PC-UNITE BZX-20 is een PIM die via de infrarood poort data kan uitwisselen met PC's en PDA's. Ook leuk is de Digital Camera Watch, die bewegende beelden kan opnemen in een resolutie van 14.400 pixels. Helemaal cool is de MP3 Watch, een horloge met ingebouwde MP3 player. Het dingetje is uitgerust met 16, 32 of 64Mb flash memory en heeft de beschikking over een USB aansluiting:

Casio's MP3 watch is an ultracool water-resistant digital wristwatch that lets you download MP3 songs to it from your computer. By summer, the watch will ship in 16MB, 32MB, and 64MB configurations and will have both a USB jack for fast music downloads and a headphone jack for listening to your music in private. But even cooler than its supersmall size is that it comes with rechargeable batteries. If there's one thing we've learned with the current crop of portable MP3 players it's that they eat batteries faster than you can buy new ones. Of course the MP3 watch also keeps time, has a built-in stopwatch, and has an alarm. As of this writing, pricing has been set for the 16MB and 32MB versions at $199 and $249 respectively.

insane schrijft: "Gamewire meldt dat de Asus i820 mobo's eindelijk uit zijn. Er zijn 2 versies, de P3C-2000 met DIMM slots ($170), en de P3C-E met RDRAM support ($200).":

It looks like two of the Asus i820 motherboards are finally available (at least here in Vancouver). The P3C-2000 goes for around $250 CDN ($170 USD) and is configured with 4 DIMM sockets, 5 PCI, 1 ISA slots, 1 AGP 4X/PRO slot and support for 133MHz FSB. However, OEMs over here have stated that this board has 2 DIMM sockets which leads us to believe Asus may have changed the specs or that they have made an error. The other Asus board now available is the P3C-E which will cost about $300 CDN ($200 USD) and is the first i820 motherboard over here with RDRAM support. (i840 boards with RIMM sockets have been available for some time.) Anyway, this board is configured with 2 RIMM sockets while the rest of the specs are the same as the P3C-2000.

Hot Hardware heeft een review op z'n site gekegeld van de 3dfx Voodoo3-3000 in PCI smaak. Hieronder een gedeelte uit de test:

The Voodoo3 3000 PCI is an interesting combination of high end 3D Graphics Technology scaled to the mainstream market. Hard core gaming enthusiast may opt for AGP versions of the Voodoo3 line or other cards supporting 32 bit color. Never the less, there are a ton of PCI slots left out there and folks who have no other option for a graphics card, other than PCI. The Voodoo3 3000 does a nice job delivering good frame rates and image quality on low to mid range systems.

On a side note, with a liitle tweaking and some custom cooling on the card (think attach fan to heatsink), measurable gains in performance should be available.

All told, the Voodoo3 3000 PCI is yet another fine 3dfx product sure to please the masses. We give the Voodoo3 3000 PCI a Hot Hardware Heat Meter Rating of 8/10.

De DPC Hitparade Top 10 van 8 januari 2000

1. Maverick(93, Gisteren:4,030, Totaal:72,465)
2. Eddict(49, Gisteren:3,458, Totaal:133,683)
3. Robert Reuser(243, Gisteren:3,150, Totaal:18,765)
4. gambit(25, Gisteren:2,908, Totaal:177,524)
5. niels(62, Gisteren:2,878, Totaal:108,273)
6. sienema(2, Gisteren:2,800, Totaal:850,600)
7. rc5(65, Gisteren:2,784, Totaal:102,916)
8. blabla(114, Gisteren:2,590, Totaal:57,322)
9. Vodka(611, Gisteren:2,563, Totaal:92,035)
10. Berry(110, Gisteren:2,391, Totaal:59,418)

En de kanshebbers:

11. dennis.vanveen(264, Gisteren:2,328, Totaal:16,775)
12. witsflow(116, Gisteren:2,160, Totaal:56,283)
13. utct(175, Gisteren:2,115, Totaal:35,150)
14. [eNeRGy](47, Gisteren:2,038, Totaal:134,927)
15. HarryB(13, Gisteren:1,998, Totaal:302,602)
16. Lies(146, Gisteren:1,895, Totaal:45,464)
17. soes(135, Gisteren:1,786, Totaal:49,878)
18. Sub-Team Hack BV(22, Gisteren:1,755, Totaal:201,493)
19. j.arensman(80, Gisteren:1,650, Totaal:82,546)
20. rkeizer(60, Gisteren:1,324, Totaal:117,021)

De Top 5 teams van 8 januari 2000

1.( 53) Happy Wedding, Mr. SAITOO!! (Team RAUG) (1,495,993)
2.( ) AnandTech (482,369)
3.( ) Slashdot.org (314,318)
4.( 1) HackZone Team -- Combined power of Russia and exUSSR countries (277,200)
5.( 1) Team Warped (OS/2) (264,048)
17.( 1) Dutch power cows (121,512)
44.( 6) HIT2000 Linux, Hack & Crack event of 1999, the Netherlands (49,652)
53.( 8) The #koffie Team (45,684)

Teams om ons heen(total stats):
36.( ) Michigan State University (MSU Spartans) (Tot:30,848,888, Gis:96,432)
37.( ) Team HP (Tot:29,669,541)
38.( ) Japan Celeron Users Group (Tot:28,955,283, Gis:89,933)
39.( ) Texas Longhorns (Tot:27,915,645, Gis:41,749)
40.( ) Dutch power cows (Tot:27,430,364, Gis:121,512)
41.( ) Wonder Monkeys (Tot:26,598,143)
42.( ) Singapore RC5 Effort (Tot:25,790,840, Gis:78,166)
43.( ) Quakenet RC5 Team (Tot:25,664,124, Gis:42,866)
44.( ) Egham Hills (Tot:24,411,171, Gis:25,910)
(Teams waarvan de score van gisteren ontbreekt, hadden geen score in de top 100!)

Dutch power cows is vandaag 17ste en heeft een kans van een op 174 om de sleutel te vinden!

insane schrijft: "iXBT heeft details gepost van de chipsets die VIA in de loop van 2000 uit zal brengen":

VIA is planning some new chipset for 2000 - both integrated and non-integrated for Pentium and Athlon class processors. So, what can we expect this year?

Athlon (Slot A / Socket A) chipsets:

KX133 (VT8371 - Q1'00):
200MHz FSB, Slot A, PC133, AGP 4X

KM133 (VT8365 - Q3'00):
200MHz FSB, Socket A, integr. Savage4 2D/3D, AGP 4X, PC/VCM 133

KZ133 (VT8363 - Q1'00):
200MHz FSB, Socket A, AGP 4X, PC/VCM 133

Pentium (Slot 1 / Socket 370) chipsets:

PX266V (VT8653 - Q4'00):
133MHz FSB, (support up to 4 CPU ???), AGP 4X, PC266, Dual V-Links

Apollo Pro266 (VT8633 - Q2'00):
133MHz FSB (support up to 2 CPU ???), AGP 4X, PC266, V-Link

PM133 (VT8605 - Q1'00):
133MHz FSB, integr. Savage4 2D/3D, AGP 4X, PC/VCM 133

PM266 (VT8607 - Q3'00):
133MHz FSB, integr. GX4C 2D/3D, PC 133/266

Bij 3D Unlimited kun je vandaag een review vinden van het ergonomisch gevormde Microsoft Natural Keyboard Pro, dat net als vele andere modellen ook over een aantal Hot Keys beschikt. Hieronder wat meer info:

The first and foremost feature of this keyboard is that it has the implemented style of ergonomics in the main body of the keyboard. Lately everyone and their dog are making ergonomic keyboards for the home user market. Most of these are pretty standard as in the how the keys are laid out for each hand and the size of most of the buttons. The Natural Keyboard Pro is no different. The key style takes a couple of days to get use to and you won’t be able to type as fast compared to the old straight style. However, your wrists won’t hurt as much and will reduce the risk of carpal tunnel syndrome.

This board also incorporates 2 USB Ports, which are located at the back of the keyboard for devices like a mouse or other device that uses less then 100mA. However, they require taking up one of your USB ports on the motherboard or USB hub before you can use them. Also a USB Internet camera won’t work because they require more than 100mA’s. Now lets take a look at all those blue buttons above the function keys and 10-key pad. Well 16 out of the 19 are already pre-defined keys and so you can’t change them no matter what. Of those 16, 7 are dedicated to Internet browser functions such as: Back, Forward, Stop, Refresh, Search, Favorites and Web/Home. Another 8 are for media type functions such as: Mute, Volume up / down, Play/Pause (one button), Stop, Prev Track, Next Track and Media. And lastly 3 are user defined and 1 Sleep button. On default the three user defined buttons access My Computer, the Calculator and your default E-mail program.

Gamewire heeft een review van de Pine D'Music portable MP3 player online gezet. Dit spelertje wordt geleverd met 64Mb flash memory, 4-bands equalizer en ingebakken microfoon om onderweg geluid op te nemen:

Immediately when you see the player you notice its small size. It's no smaller then a regular MP3 player but it has a convenient button layout system with the main buttons IE stop, play, and skip buttons on the front as well as a 4-band equalizer button (set modes include, Pop, Classic, Jazz, Ex-Bass and Flat), mode button (to switch between a recorder and mp3 player), select button which acts like an enter button or used for a "repeat" feature, and the menu button which brings up available space for both internal external and together, has your serial number, and deletes files. On the left side there is a volume up, down, record button, and an a-b button, which allows you to play a part of a selection over and over in a repeated fashion. To the right we have a Date I/O port, the batteries, and a hold switch, which becomes very useful because the play buttons on the D'Music are just a little to easy too touch. The top has the headphone jack and a slot to but in your 32 MB flash card. On the back we have a lever to push out the flash card and the bottom we have nothing.

AGN Hardware heeft een review van de Alpha P612H35M Athlon cooler gepost. Om ruzie met ongelukkig geplaatste power connectors (bijv. op oude K7M revisies) te voorkomen heeft de heatsink enigzins bescheiden afmetingen:

I have used the single fan unit on a number of Athlon CPU's. The CPU speeds have varied between 500 MHz and 840 MHz , the core voltage between 1.6 and 1.8 volts. These heatsinks have always performed well and there have been no glitches that I can attribute to the heatsink performance.

These units are excellent heatsinks and will serve the majority of overclockers and non- overclockers well. Extreme overclockers may want to wait for the copper inlaid unit or devise another method like liquid cooling . These units are a solid aftermarket heatsink and I haven't seen a better Athlon heatsink available at this time. The price of the units is about $30 for a single fan unit with the dual fan about $36. They are available direct from Alpha or though distributors and retailers such as myself.

CTNews heeft een review online gedumpt van Abit's BE6-II moederplank, dat dankzij het Soft Menu III je de keuze geeft uit zo'n 121 bussspeeds. Hieronder wat meer over zijn overklokbaarheid:

The Overclocking Results: The fastest we were comfortable with was at 143 MHz FSB x 5.5 multiplier = 786.5 MHz ! Not bad for a 35E series Pentium III 500! We did need to bump up the voltage and that's why I've lowered the CPU speed back down to it's default for all the other testing now. I do want to keep my PIII running.

The Slot 1 Celeron 400 chip is known to be a little weaker than a PPGA or the PIII's are. It did O/C very well on the BE6-II as opposed to the BE6. We got in much higher FSB setting and loved the voltage adjustments. Cooling becomes a major problem very quickly with the Celeries. We needed to run a double fan with a huge heatsink. Next we removed teh case doors. Then we pointed a couple of oversized (4') fans CPU area of the motherboard to get stability at the 100 MHz FSB settings. The best overall settings, the one's that worked in gaming, and Benchamrking were the 83 MHz FSB settings. When at a multiplier of 6 and Voltage adjustments were bumped up to to ensuer the CPU would POST everything was fine. Results here: 83 FSB x 6.5 multiplier = 539 MHz!

Voltage Regulator Outputs are set automatically but can be changed in the BIOS. Playing with this setting can cause power timing problems. If you don't have the proper rated RAM don't play this game. Our Test Bed (PC133 SDRAM) had no problem with the CAS latency timing set on 2. Again, Wow!

The Register speculeert dat Apple binnenkort een op PalmOS gebaseerde PDA aan zal kondigen. Volgens bronnen in Taiwan worden er op dit moment al samples geproduceerd:

The results of Apple's co-development partnership with Palm Computing look set to emerge in the very near future, now that early samples of an Apple-branded PalmOS-based device have already started coming off Taiwanese production lines.

Apple CEO Steve Jobs himself confirmed the two companies' close ties this week, during his keynote at Macworld Expo, held in San Francisco. Introducing Palm CEO Carl Yankowski, Jobs told the audience Apple had been "doing a lot of work with these guys lately".

[...] Meanwhile, Taiwanese sources close to a leading OEM -- though none will say who this leader is -- claim contacts within the company have seen early Apple-branded Palm-style devices in limited production. Details of the device other than its existence are scant to say the least, but it does suggest Apple is getting much closer to announcing a PDA of its own.

UncaMity's Website heeft een review van de Abit VA6, een moederplank met VIA Apollo Pro 133 chipset, klaar staan. Het bord bleek prima te werken en overklokken vormde ook geen probleem:

Finally, I decided to try 770MHz, which involves setting the bus speed to 140MHz. I admit that I expected this to fail, since it was really pushing the CPU, to say nothing of my RAM. The system POSTed, and I was happy to see that the VA6 automatically kept the memory bus at 133MHz. But Windows crashed almost immediately after it began to boot. I set the voltage to 1.65v, and got the same result. I set the voltage to 1.70v, and the system booted into Windows. I was still skeptical, since I couldn't believe that a simple .05 bump in voltage would turn a CPU that crashed Windows immediately into a rock-solid performer. But so far, after 95 minutes of testing with CPU Test 1.5 the system hasn't had a single glitch. And trueSpace rendered a test scene without any problems as well.

Uit een eerdere review van Ars Technica bleek overigens wel dat de Apollo Pro 133 qua geheugen en AGP performance duidelijk achter blijft bij de BX chipset.