Slashdot Mirror

No, but it pretty much made them saleable. If you remember back to the dawn of the MP3 era, everyone either had a small (like 128 or 256mb, tops) flash-based mp3 player, maybe with a few memory cards (I went threw several Rios before my first iPod). Almost nobody actually bought HDD MP3 players.

Not really surprising when you think that the iPod was up against something like this Archos. It was a long time ago but I remember the Archos UI being really poor and you had to manually copy all your music across.

It was no surprise that the iPod kicked the MP3 HD market up the arse. The competition was woeful.

That trend has been happening for the past 40 years. If you looked at a 1990's graphics accelerator card (Hercules Graphics Station Card or a Voodoo 5000/6000, you would see that all the different components (RAMDAC, graphics processor, memory controllers) were all on different parts of the circuit board. Now, most of that logic is within a single chip Geforce 9800GTX

Memory chips keep changing as rapidly as the CPU's do. Assuming that a CPU manufacturer wanted to enter the memory chip market, by the time they had caught up with current state-of-the-art in memory technology, bus communication and got the product onto market, the memory chip manufacturers would already be designing, producing and marketing the next generation.

I'm not a troll, and I'm not an audiophile. I'm someone who likes good-to-decent STEREO audio, does not care about EAX or surround sound, and likes *rock solid drivers* with a very small footprint. I would say you quickly using Google to find fanboy reviews of Turtle Beach products would classify as trolling. I guess you didn't bother to read the defined review from Sudhian back in 2004, talking about how Turtle Beach had gone downhill since the release of their Santa Cruz card:

http://www.sudhian.com/index.php?/articles/show/512

I recommend you read this article IN FULL, because it was written by folks who love the Santa Cruz. The days of Santa Cruz are over; no more nice driver changelogs, no more high-end support. Just generic reference design garbage.

And what I say of onboard audio is quite true: I hope you like bus noise. The only onboard audio IC which doesn't exhibit that problem is Intel, and it's predominantly used on their own boards (read: majority of consumers do not buy Intel mainboards). Most generic HD audio or AC97 crap is Realtek, who has a history of making absolute atrocious garbage (audio ICs, network ICs, with quite possibly the worst drivers in the industry.)

And don't even get me started on VIA.

I was on an SFF kick a few years back, built several SFF systems for friends. I thought they were the greatest thing. Then I came across an SK43G that wouldn't be stable.

Here's a long thread on Sudhian discussing it:
http://www.sudhian.com/index.php?/forums/viewthrea d/50166/

I've built lots of computers. I know there are quirks and problems. What really frustrated me is that shuttle did *nothing* to help an obvious design or manufacturing error. So, I was stuck with bad hardware. That leaves a bad taste in my mouth, and makes me VERY leary to try another shuttle, or to recommened one to a friend.

For this friend, I ended up taking the system back, swapping out parts and getting him a different box. He LOVED the shuttle design, so it was a similar system. I have the SK43G around here somewhere, still not sure if I trust it.

I think the newer nForce 590 has more PCIe lanes. IIRC, the one motherboard I have with nForce 590 is a 16+16+4+1 for a total of 37 lanes (another note says 46 lanes)

http://www.sudhian.com/

nForce 550 - 20 lanes
nForce 570 - 20 lanes
nForce 570 SLI - 28 lanes
nForce 590 SLI - 46 lanes

You're describing this disaster scenario as if it's a new problem introduced by this technology. I've already had plenty of demos I was doing for important people blow up because of a hard drive error (I'm just lucky that way). Which is more likely: that you'll hit the flash write limit, or that the mechnical part of the drive will crash and burn? My experience with hard drives suggests they're none too reliable right now, and anything that can reduce the amount of time they spend moving around has a potential to improve the mechanical failure rate.

With wear leveling and some spare capacity to replace early bad bits, 100,000 write cycles for each bit works out to quite a bit of data with a decent size chunk of flash. An analysis I liked at http://www.sudhian.com/index.php?/articles/show/68 6/3 suggests a typical office worker will get 33 years out of a hybrid drive, while even someone who writes 6GB of data every day should get 4 years out of the drive. There are plenty of drives out there that aren't even warrantied for 3 years right now.

As for too much writing induced disasters, you just have the flash part of the drive start throwing SMART errors when it gets low on useful bits, the same way drives right now can report when they're running out of spare sectors to relocate bad drive sections into. That should catch the issue well in advance of a flash crash.

When you run the numbers it doesn't sound that difficult to create a flash-based design that would likely outlast the mechanical parts of a standard hard drive.

The two most likely successors are the Geode [...]

Samsung's OneNAND design uses some interesting implementation details to run faster than is normally expected with flash. Their specs at http://www.samsung.com/Products/Semiconductor/OneN AND/index.htm suggest writes at 9.3MB/s and reads at 108MB/s; plenty fast for many applications even without running multiples in parallel. It's certainly much slower than writing directly to the drive if the drive is active. However, if the drive isn't spun up at the moment and the amount of data to be written fits in the NAND flash cache, I could see this being a net performance boost over the spinup/seek time combination of the hard drive.

They also spec 100,000 erase cycles before it's worn out. As was noted by an underrated poster the last time this came up, intelligent flash controller designs like this can cope with bad bits and assuring level usage of the memory much better than what you normally see in random hunk of flash.
An analysis at http://www.sudhian.com/printdocs.cfm?aid=686 suggests 33 years of usage for a typical worker. When you run the numbers it doesn't sound that difficult to create a design that would likely outlast the mechanical parts of a standard hard drive.

This upcoming box from Shuttle was shown at CES and has a CableCARD slot. The links go to some pictures that were taken surreptitiously by an attendee.

Yeah, the Geode GX is pretty slow, even compared to the Via C3. It's based on the Cyrix MediaGX chip which was based on their 5x86 CPU (how's that for some retro computing?). More info here.

yeah you have some fixation that cpu determines tools....

I run a dual-core AMD64 [yeah I know that's like the 10th time I've said that on Slashdot, what you wanna fight about it?] using Gentoo Linux.

My work uses Fedora Core on a Pentium4.

Guess what.

Just guess.

Oh come on, guess!

WE USE THE SAME TOOLS!

There is no reason why a ARM based linux box couldn't run GNU tools like gdb, gcc, as, make, etc... and GPL/OSS software like Gnome, xorg-x11, firefox, etc, etc, etc.

It's only idiot windows fanboys and people running very OS specific [e.g. EDA tools] software that think processor dictates OS platform.

Sure I can't use the same x86_64 binaries on an ARM but the software will recompile there just fine.

And you say "low power" as some form of disparraging remark. A 500Mhz ARM is more than powerful enough to run firefox or OpenOffice without being crippling slow.

The reason why those 624Mhz ARM based Axim's can seem slow is that they run wince and usually the cache has been crippled [or reduced].

A 500Mhz ARM with proper 64/64 cache would blow any VIA C3 or oldschool Geode out of the water in terms of MIPS and still take less power.

They're also way more scalable. Take that 500Mhz clock down to a dozen Mhz or so and basically take the same amount of power as a LED [2mA * 9V == 18mW @ 0.25mW/Mhz => 72Mhz ARM]. Granted the power draw doesn't scale linear and is dependent on usage but you can probably pack a good steady 50Mhz into 18mW with a 130nm process at a stable voltage.

Sure if your a developer or artist or performing some simulation you need a very high MIPS level that ARM can't provide.

My point though is for the hundreds of millions of people with [usually multiple] desktops out there a much lower power ARM [or heck even MIPS, but I like ARM better] based computer would be better.

If it means not using windows all the better. Let them use a real operating system like a BSD or Linux distro.

I personally don't really care if people can't use the latest Windows bloatware. It's not a fucking right and it's an eventuallity anyways [e.g. as soon as we hit a good energy crisis or two].

Personally if I owned a computer fab plant I'd edge some new tech in there so I'd be on the ground floor when something else hits big. Right now Dell banks solely on pushing Intel desktops. Look how good sole products work for car dealerships. In north america they're selling cars for less than MRSP because the public simply doesn't want what they have to sell anymore.

Similarly we'll get to a point were a 450Wh desktop is too much for simply checking a recipe online or reading an email and people will start trying to find alternatives.

Now you can be like Dell and HP and Gateway and bank your entire multi-billion dollar business solely on wintel machines. Or you can simultaneously invest into more efficient desktops/laptops that in the end cost the users less [smaller, less battery for laptops, no windows tax, etc...].

I'm also certain a decent ARM [e.g. ARM9 or similar with 64/64 cache and v6 ISA] is a hell of a lot cheaper to fab then a Pentium{4|M}.

P4 Prescott = 109.0mm^2
ARM922T = 3.2mm^2 [with 16/16 cache]

On the same 300mm wafer with ~580 P4s you could pack ~21,700 ARM922T cores [according to here]. And the ARM922T is 130nm not 90nm like the prescott.

So clearly the ability to produce a bazillion of these things is there. The OS and associated OSS tools already exist. All that lacks is people smart enough to capitalize on it.

Tom

Intel does its best to make sure you don't get a choice.

While Intel has the occassional good product, their server chips are completely outclassed by AMD's Opteron, especially in multiprocessor setups. Buying, say, a 4-way Xeon instead of a 4-way dual-core Opteron server these days ought to be a firing offense. If you look at performance-per-watt the difference is even more extreme.

It sure looks like Intel has been unimpressed by the JFTC case against them.

I found this intersting quote at http://forums.sudhian.com/messageview.cfm?catid=18 &threadid=78260

"Also on Monday, Jobs said the next version of OS X, called Leopard, will be released in late 2006 or early 2007, which he said was the same timeframe as Microsoft's next Windows update, dubbed Longhorn. Microsoft has said Longhorn will be released by late 2006. After Jobs' presentation, Apple Senior Vice President Phil Schiller addressed the issue of running Windows on Macs, saying there are no plans to sell or support Windows on an Intel-based Mac. "That doesn't preclude someone from running it on a Mac. They probably will," he said. "We won't do anything to preclude that."

Basically, at this point, I am really confused. Agreed, this guy is a marketing chump... but what is he suggesting? What does this suggest about BIOS/Open Firmware Issues?

In the end, if OS X is available for non Apple hardware, I will be completely floored.

LOL. I think that Toshiba and Panasonic are in better position to improve the battery market than anyone. There's even been quite a bit of discussion by one guy over here and here for a while now about this being a long standing problem and possible ways to solve it. As for battery solutions, nuclear batteries are a good idea, but given the number of stupid people in the world, I wonder if it's all that safe. Then again, it may be a good idea to have them anyways as a way to clean up the gene pool a bit? Maybe? ;) Well, all joking aside, nuclear decay batteries are a good idea, but only for low voltage use. Anything above 12v and 20 amps is going to go beyond the current capabilities of nuclear decay technology. So for small devices, they're perfect. Larger devices such as electric cars are going to present a problem...UNLESS, for example, someone finds a way to make them safe, durrable, and able to take being abused like lead acid can. They'd also have to find ways to protect them in event of car fire or accident. If they can do that then they would also be a great solution for electric cars.

Everyone should read about apple's shitty customer service.

I just found this editorial link about what happens to apple's normal customers (ie, not the guys getting sent review machines at tom's hardware) http://www.sudhian.com/showdocs.cfm?aid=665

It's shocking and I can't believe the apple fanboys put up with shit like this.

AMDZone.com Tech Report Sudhian Hexus Hot Hardware Anandtech xbit xbit PCWorld Trusted Reviews

The Inquirer is the greatest IT web site in the world, no question about that. But they rarely if ever do any reviews themselves. They just tell you trends and like to reviews done by other sites.

It is so rare that I find a review on a hardware web site that does not stink of bribery and/or incompetence, in fact, that I wrote down the URL for the one case that I noticed: http://www.sudhian.com/showdocs.cfm?aid=556

I salute journalists who value integrity higher than bribery. I'm not sure I could do that, if I were in their place. You have to feed the family, right? At least, here in Germany, we have c't, which is one of the last bastions of integrity in PC print journalism, and even they have their down sides.

The Intel is obviously CPU-bound in the gaming benchmarks. Anyone with half an interest in games would see that one, right off the bat.

That is wrong. The Intel system is graphics bound, not CPU bound. In a clock-for-clock comparison Pentium M performs very similar to Athlon 64 in gaming performance when a high end video card is used as Tech Report, Anandtech, Sudhian and others have shown.

The problem with Intel's gaming performance in this review falls squarely on the low performance of Intel's integrated graphics. It is a problem of Intel's making. Intel will not let vendors use the Centrino name unless an Intel chipset is used. Being frozen out of the Centrino notebook market discourages others like ATI from creating better integrated graphics solutions for Pentium M.

Somewhat to my amazement, this actually worked quite well despite my dramatic ineptness with electronics.

Pictures are in this thread.

I'd post a pic of the setup.. But I'm ashamed of it. Sellotaping wires to a battery just isn't .. y'know.. valid somehow.

Given the topics in the audio section (it has an audio section!), the site seems to lean more towards audiophiles.

I don't agree with the dismissal of lossy algorithms either, but I think it makes sense given the context.

Have you ever heard of 'stealthing' the optical drive? It basically amounts to mounting the drive slightly further back than usual and attaching the Shuttle frontplate. It's a pretty basic mod, and lots of information is readily available.

The four GPU one was the Voodoo 5 6000, a twelve and a half inch long monster that originally needed a special 3dfx made power supply to run correctly.

Multiple cores _can_ work, but there are limits to how far you can push it before it becomes silly. I'm sure nVidia and Gigabyte aren't about to go down the road 3dfx took.

Review with benchmarks can be found here

Did anybody else's butthole start to hurt when they saw the Crystal Silicon Ingot?

The link works. Just a browser fart. Never mind.

http://www.hothardware.com/viewarticle.cfm?article id=592&cid=1
http://techreport.com/reviews/2004q4/athlon64-fx55 /index.x?pg=1
http://www.bit-tech.net/review/364/
http://www.short-media.com/review.php?r=266
http://www.pcstats.com/articleview.cfm?articleID=1 666
http://www.tbreak.com/reviews/article.php?id=331
http://www.amdreview.com/reviews.php?rev=fx-55-400 0
http://www.techwarelabs.com/reviews/processors/amd 4000_fx55/
http://www.hardocp.com/article.html?art=Njc1
http://www.xbitlabs.com/articles/cpu/display/athlo n64-fx55.html
http://www.sudhian.com/showdocs.cfm?aid=614

SilentPC Review is always THE place to go when you are concerned about PC noise.

And Sudhian had a review of the new Shuttle XPC a few weeks ago (hey, that how far ahead slashdot is :-( ). There's a noise comparison in all of their SFF tests (direct link for convinience. IMHO Sudhian is THE place to go when you're searching for SFF infos. The forums are very valuable resources on both sites.

Bye egghat.

SilentPC Review is always THE place to go when you are concerned about PC noise.

And Sudhian had a review of the new Shuttle XPC a few weeks ago (hey, that how far ahead slashdot is :-( ). There's a noise comparison in all of their SFF tests (direct link for convinience. IMHO Sudhian is THE place to go when you're searching for SFF infos. The forums are very valuable resources on both sites.

Bye egghat.

I've had the SN41G2 running for about 15 months non-stop as my media center. With a little Modding you can make this a silent machine. The design these guys come up with is remarkable; while cramming that much functionality into a "toaster" is in and of itself an accomplishment, the forethought in the design of these machines make installation less tedious than one would imagine. My only regret is how much horsepower I'm wasting by only only using it as a home theater pc.

Mod? yeahhhhhh

I thought it was pretty cool at that time.
Personally, I am waiting for the athlon 64 Shuttle XPC to come out WITH the nvidia 250 chipset. The current chipsets just arent worth it;)

Yummmmm......

Here are some more review links for those who are interested:
Tom's Hardware
Bit-Tech
Driver Heaven
AMD Zone
Hard Tecs 4U
PC Perspective
Ace's Hardware
Sudhian

The FPU is a little better now. It runs at full CPU speed instead of half like the earlier C3's. It's still underpowered though. Sudhian has a review of the last generation MII 12000 here. The 1.2Ghz w/o hardware MPEG4 acceleration can't play 720x540 DIVX file smoothly. If the hardware MPEG4 works, 1.3Ghz should be fast enough for any home theater PC job except video encoding like recording TV. You'll need a TV tuner card with a hardware MPEG encoder.

Well, if you would ever do something like upgrade your memory or change a hard drive, and don't consider it "building one yourself", a Shuttle XPC would be my bet. Yes, it's a barebone PC. Yes, you do have to install the processor, memory and hard drives (plus the obligatory DVD-+RW - who needs floppies anymore). But if as you said you don't do gaming, everything basic you'll need from a PC is already there. NIC (on some models two), sound, USB/Firewire and graphics; some models even feature an in-built GF4 MX with two VGA-outs, very nice for an integrated chip.

On the downside, there's only one free PCI slot and one AGP slot should you wan't to install a better graphics card. But on the upside (and surprisingly related to the discussion), all Shuttle XPC models should be shipping with Mandrake - here 's one review of the combination, there are many others to be found, just google for "shuttle mandrake".

As covered by arstechnica, there are also reviews at [H]ardOCP, Hexus, HotHardware.com, Sudhian, and The Tech Report. AMD's official announcement is here.

Details are available on their forum.

Link here to the Sudhian Forum.

If the current state of their server is any indication of this thing's performance, its sad :)

There are numerous "fan-sites" about mini-ITX boards, especially those from VIA. Maybe some of the folks on there have already managed to obtain some docs.

For example: forums.sudhian.com have sections about linux and mini-itx boards.

Or how about the kernel mailing lists? I think that Alan Cox himself mentioned in his diary (before he started writing it in Welsh, that is) that he's a fan of these machines and had done some driver development.

I'm referring to their crappy chipsets, which made me chuck a motherboard+CPU to switch to Intel. Check out this FAQ that I basically know by heart, and even that monster list of bugs and incompatibilities couldn't save my machine from instability. I'm stuck on a P3 now because of the money I wasted on that thing.

I can't wait for Prescott because it'll push P4 prices down in to my range.

Sudhian has a forum dedicated to HTPCs only.

HTPC forum.

There are some good HiFi like cases like the CoolerMaster ATC 600,610, 620 and 630.
The LianLi 9100, 9300, 9320 and 9400 are fine as well.

The other forums at Sudhian are interesting as well if you into XPC and similiar stuff.

Bye egghat.

I have a KT7 too and put a AthlonXP in it. No need for devices like the XP-TMC. It's a XP running at 1700MHZ with a 100MHz bus, sounds like a bit of a mismatch, but for ~$60, it's a good boost. You can see how to do it here: here

HEXUS reckons a 200MHz front side bus can't hurt. here. There's a picture of a brown semiconductor, also known as the "brains of a computer".

TOM'S HARDWARE has a controversial conclusion about the 3200+ and describes it as a "spineless paper tiger". It thinks the 3200+ is "much too aggressive" and it should be an XP2800+.


SUDHIAN Some crazy looking geek at Sudhian (hi Joel), reckons that AMD is being a little coy with clock speeds while its PR speeds have rocketed skywards.


FIRING SQUAD says AMD's odyssey for the performance crown has been a little more treacherous than Her Indoors, but welcomes the introduction of the 3200+ and the 400MHz bus.


TECH REPORT says there's not much new to report about the 2.2GHz chip apart from the fact that it runs on a 400MHz front side bus. But it reckons that the release is timely. There's a picture of a brown semiconductor which appears to resemble the brains of a computer.


LOST CIRCUITS contrasts the real brown brains of a computer with the hypothetical 3200+ brains of a computer it previewed a month or two ago.


BIT-TECH reckons that AMD's finally released the processor that the 3000+ should have been, denies the site's too pro-Intel, and puts it through its paces. There's a picture of a brown chip which appears to be the brains of a computer.

I stop whoring now, more to be found at amdzone

It looks so cool. As a SFF nut any time you can shove a screen inside your box, like this xenarc, which supports 800x480, makes me drool. xenarc also sells a single DIN carputer. With built in pcmcia slots perhaps you could VNC into your remote servers from the parking lot using 802.ll. Try swinging that past your purchasing officer.

Great write-up.

Hers's the [H]'s take on it, , and here's Sudhian's.

The first poster isn't really that far off. While each card has access to all IRQ lines, in practice, most only use 1 IRQ (off of Pin 1). Here's a chart of how VIA south bridges map 'em: http://images.sudhian.com/faqs/kt7/irqs-figs.gif

More info can be found here: http://www.sudhian.com/showfaqs.cfm?fid=2&fcid=26# 55

The first poster isn't really that far off. While each card has access to all IRQ lines, in practice, most only use 1 IRQ (off of Pin 1). Here's a chart of how VIA south bridges map 'em: http://images.sudhian.com/faqs/kt7/irqs-figs.gif

More info can be found here: http://www.sudhian.com/showfaqs.cfm?fid=2&fcid=26# 55

Although the extra cooling on the OC'd R9700 or Ti4x00 might be a lil loud ;oP

Running Debian Linux 2.4.19 on a Shuttle SS40g. There are a couple of pages with instructions on how to set up Linux on these machines (here and here). The biggest prob on this machine regarding Linux seems to be the SIS 740 chip. You have to specify 'pci=bios,biosirq' as a kernel parameter even if the PCI bus seems to initialize fine without. You'll have probs with the modules otherwise. You also need a custom graphic driver from Thomas Winischhofers page and you have to configure X-Windows by hand.

I have graphics, USB, sound, network and firewire running fine for me. Note though that I did not even try to get 3D accel (DRI? DRM?) support configured.

As for Linux performance I have to confess and to display my ignorance towards performance freaks in public by saying: Yes, the performance is fine. For me. ;-D

regards,

-rolf

go to sudhian forums to see some freeBSD and Red Hat users discuss how clean it installed, but that there may be some driver issues.

----
Beware of women who pay their rent with one dollar bills

VIA's end-user Support site

VIA's Web site

VIA is onto a really good thing with the Eden series motherboards using their CPUs. The upcoming EPIA M-series motherboards are based on VIA's CLE266 chipset, which offers DDR RAM and on-chipset support for offloading DVD display operations from the CPU. If VIA delivers those features, the EPIA M will make an excellent living-room multimedia box.

Sudhian's Small Form Factor Forum is very active, and a good place to watch for more on EPIA hardware.

Look at the last page

It has a picture of the box under the guys arm