Via Unveils 1-Watt x86 CPU 276
DeviceGuru writes "Taiwanese chip and board vendor Via Technologies has introduced a new ultra-low voltage (ULV) processor aimed at industrial, commercial, and ultra-mobile applications. Touted as the world's most power-efficient x86-compatible CPU, the 500MHz 'Eden ULV 500' processor debuted at an Embedded Systems Conference in Taipei this week. Via says its chip draws a minimum of 0.1 Watts, when idle, and a maximum of 1 Watt, making it a great candidate for consumer electronics devices such as UMPCs, PVRs, and such."
laptop anyone (Score:5, Insightful)
How does it compare? (Score:5, Interesting)
The article doesn't say what socket and interface the chip uses. Are they still on Socket 370?
Re:How does it compare? (Score:5, Informative)
How does this chip compare in performance per watt against ARM, PowerPC and the like?
Pathetically badly. Most modern low power ARM variants are in the range 0.3-0.5mW/MHz. At 500MHz you'd see them chewing up about 150-250mW. Last I checked the Via x86 chips were single issue, so it's not too unfair to compare an ARM11 (or similar) against them. Quite frankly an ARM11 will outperform the Via chip and run lower power.
The idle power figure is a joke. I can't recall the last time I used an ARM chip that idled at 100mW. More like 1-10mW. Still, it's nice to see an x86 chip get into sub-watt territory.
Of course, ARM doesn't run native x86... and that's pretty much the only reason there's such a large market for these Via x86 chips. It's also the reason you never see them in deeply embedded systems where people don't really care so much about what ISA you're running.
Re:How does it compare? (Score:5, Interesting)
When i was building a linux based PVR, x86 compatibilty was not a deciding factor *. What i wanted was a cheap fanless board that could playback mpeg2 and divx, with a PCI slot for a tuner card, TV-Out, and SATA.
When i was looking there were hundreds of Via C3/C7 based boards from heaps of manufactures, with countless different options. There were one or two ARM and PPC boards, even one with a transmetta CPU, but they didn't have TV-Out, or they had TV-Out but no USB or PCI.
I would have loved to go with another architecture but the market for retail consumers just isn't there.
* Actually, now i've said that i imagine compatibilty of the tuner drivers with non-x86 could be an issue.
Re: (Score:2)
Re: (Score:2)
For putting a DIY media center I don't think this VIA or an ARM would be a good choice. Both would tend to lack the power to transcode video at anything like a reasonable speed. The one thing this VIA chip has over the other CPUs in the embedded space is that it uses the X86 isa. It is annoying but there are some Linux drivers that ONLY exist for the X86! they are nasty binary blob or wrapped windows drivers but if you want that hardware to wor
Re: (Score:2)
Re:How does it compare? (Score:5, Insightful)
Ex. Take an Barton core Athlon and compare it with a 1st Gen P4, running both at the same clock speed. That Barton will significantly outperform the P4, even with the same Mhz. Conversely, thake a Core2 Duo and an Athlon64 X2 of the same clock speed - the Core2 Duo will wipe the floor with the Athlon64 X2.
Mhz only means something when the processors are of the same line. Different lines in an arch can drastically modify the CPUs relative performance by Mhz, varying app to app, and changing the arch completely will destroy most comparisons.
Another example, would be to compare a 500Mhz EV6 Alpha to a 1Ghz Athlon - There are many tasks at which that Alpha will pretty much destroy the Athlon in terms of performance, even at half the clock speed.
So, what you want is power:performance-at-desired-tasks ratios, it's more complex, but it's not useless (and in some cases, counterproductive/counter intuitive)
Re: (Score:3, Informative)
Re:How does it compare? (Score:5, Informative)
Your premise is correct that it is an apples to oranges comparison, but not really for the reasons you describe.
Re:How does it compare? (Score:4, Insightful)
RISC is an instruction set thing, with the caveat that RISC instruction sets lend themselves to pipelined instruction execution as a by product.
Yes, modern x86 processors have RISC like microcode implemented using pipelined cores, but the x86 -> microcode converter is extra logic RISC processors just don't need.
There is no way you can implement an x86 chip in the same number of transistors as a RISC chip like ARM or MIPS, hence this VIA chip having considerably more power draw.
Re: (Score:2, Insightful)
(Score:4, Wrong)
Re:laptop anyone (Score:5, Interesting)
A really low-power Dothan or single-core Yonah will sure draw a few multiples of this beast, but they will do so while giving much better performance.
Re: (Score:2)
Not entirely sure why we specifically need x86 for embedded stuff like PVRs though... It's not like you're having to run something Windows on it, which is tied to specific architectures.
Re: (Score:2)
Re: (Score:3, Insightful)
For a user-oriented workstation, true. Even with the Via C7, a single HDD (at spin-up, anyway) could consume more than the CPU.
I think, though, these things mostly don't go into actual desktop machines. They go into car audio solutions (with a 4x20 non-backlit LCD or even VFD), or routers (headless and with CF or USB storage) or various low-demand servers (also headless
Re:laptop anyone (Score:5, Informative)
I have used every single Via CPU from the original Eden 533 up to 1.5GHz C7 and IMO the C3-C5 spec Edens are just about useful for a dedicated appliances, small firewalls, small specialised servers and such. They do not have enough grunt for a laptop. The fact that most of them have are shipped bundled with relatively weak video does not really help either. Even the mpeg support on some motherboards cannot really help. Xterm is probably the most you can do with them as far as clients are concerned. Still better than similarly clocked Crusoe though (now that is a drag of all drags).
C7 is a completely different beast. This is probably the best CPU for a corporate laptop out there at the moment. A laptop is worthless without a "link to the mothership". Intel Core and AMD have to use CPU resource to do all of the encryption and decryption. This may amount to 30-40% of your CPU on a 54G wireless lan. Compared to that Via C7 has hardware AES acceleration so you can actually protect your traffic properly while using less than 1% of your CPU. It also has enough grunt to run most common road warrior apps at acceptable speeds. It is a pity it is not available as a laptop choice anywhere outside the far east.
Re: (Score:2, Informative)
Re: (Score:2)
I think it was tested under OpenBSD.
Re:laptop anyone (Score:4, Informative)
The reason is that at least as of the versions present in major distros openssh does not for some reason support openssl engines. AES (and RSA in latest Via CPUs) is done using an openssl engine which has to be initialised and loaded. This can be done for OpenVPN, Apache, Pound and nearly any other piece of software using OpenSSL, but not OpenSSH. For some reason Theo's people in their infinite wisdom left that part out (it is trivial). There was a patch, dunno if it has made it into the main tree.
As far as non-OpenSSL software is concerned, the kernel itself can use the hardware AES for filesystems and IPSEC. I have run it for quite a while for both OpenVPN and IPSEC. It can run around a Dual Xeon in circles. I would expect it to have the same killer performance for encryption of filesystems and encrypted backups as well. In fact this is possibly the only CPU on the market at the moment where having all of your data encrypted is a realistic proposition. The rest will choke on it and crawl like a 486.
There is also further improvement from having true on-CPU hardware RNG which all programs in need of good random numbers can use as it is implemented at kernel level.
Probably the highest praise to it is the fact that most of these features have now started showing up on Intel roadmap documents for the future x86 CPUs destined for the embedded market. It is the Athlon history repeating. When someone else is doing something right Intel copies it, claims innovation and launches a marketing salvo trying to lie that "they have been doing it all along".
Re: (Score:2)
Re: (Score:3, Funny)
Sudoku?
Re: (Score:3, Funny)
Re: (Score:2)
Would you honestly want only 500Mhz in your laptop? But it would be great if handheld devices could have this much power though, preferably on something like a mini sized tablet (or an over-sized iPhone).
I'm not sure MHz are a very good absolute measurement of the processing capabilities of this thing...
But it could make a very decent laptop. I used to have a Vaio with a Pentium II 400 (PictureBook) that was quite nice. Not the kind of thing you'd run Vista on but with XFCE it ran like a charm despite having an abysmal battery life.
Re: (Score:3, Informative)
Centrino as well as any Core derived notebook under Winhoze uses voltage and frequency scaling. It will ramp up to its spec-ed frequency only when pushed really hard and in some laptops only when on AC power. If you want to actually have reasonable battery performance on Linux you end up doing the same using the cpufreq susbsystem. Example from a Core Duo on which I am typing this post:
model name :
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Would you honestly want only 500Mhz in your laptop?
Well, yes. Especially if it ran cool, which it sounds like is the point with low power.
The ancient junk Matsushita notebook I had in Japan was plenty fast enough and was slower than 500MHz. Ran a now ancient version of Turbolinux, and I never had to reboot it the entire time I had it - about a year.
I've gotten a Mac Powerbook Pro for my wife that runs much, much faster but it runs so hot I'm afraid both of it failing in the tropics and it is absolutely a Keep Out Of Reach Of Children thing. It gets HOT.
Re: (Score:2)
Re: (Score:3, Informative)
(1) Yes. Take a look at 2.5" drives used in laptops, for example. You could also use flash instead of an actual disk. Having done that myself, I must give you a word of warning: don't do flash+usb on Linux. It will hang because of I/O errors every few days. I believe this is due to there being a hardcoded limit on the number of writ
holy cow! and their 1.5GHz is only 7.5W (Score:5, Interesting)
20-30W. With HDD, DVD, encoder card etc, it draws 80W on start,
and somewhere between 30-60W when running.
Take 10-20W off my figures by using their 1.5GHz ULV
and you get potentially more processing power at less
than 50W!
I know that VIA chips are pretty feeble (i.e. their 1.5GHz
chip is probably closer to a 1GHz intel chip), but with an
encoder card (dual actually) I can be recording two
channels with the CPU at 10%. Given their mobos have
mpeg decoders on board, I can add watching a DVD or TV
for another 30-40% CPU time.
The only thing is ad-skipping and re-encoding are pretty
slow.
Re: (Score:3, Interesting)
The VIA epia platforms like the one you have weren't that great. I had their 600mhz chip and ITX board and on the meter it was still drawing about 40 watts idle at the plug. The power supply probably wasn't the greatest but still I had higher hopes. That was only the ITX board plus a normal 3.5" 7200rpm hard drive. The cpu was barely enough for most tasks and some tasks you didn't even want to do. It is probably much better with your cpu but you're still drawing more power than necessary.
As a comparison
Re: (Score:3, Informative)
Stop spreading bullshit FUD. The MPEG2 decoder hardware has been supported for years now in open source. My MythTV frontend, a Via EPIA M10000 running at 1 GHz uses the MPEG2 decoding hardware when playing back video saved from my backend's Hauppauge PVR 250 hardware mpeg2 encoders just fine with very little CPU usage. The only problems arise when you try playing DivX or MPEG4 streams.
Re: (Score:2)
Again, these things might be included into a new variant of the 128-processors-supercomputer in an ATX form factor. Power-wise, it should fit the bill, and it might even be more powerful than the 96-processors supercomputer under the desk Orion DS-96 (http://www.pcstats.com/articleview.cfm?articleID
Why not make 64 of these on a single chip? (Score:2, Interesting)
Re: (Score:2)
IOW, I'd take a Core2 quad core over 64 Vias anyday.
Re: (Score:2)
Concurrency is perhaps the biggest problem to modern day CS.
If we could figure out how to use all those cores effectively, it would be awesome. Until then, its of dubius as a archaeticture
Re: (Score:2)
Re: (Score:2)
Tom
Cool! (Score:3, Interesting)
Re: (Score:3, Informative)
Re: (Score:2)
Sounds like an overclocker's dream (Score:2)
Re: (Score:2)
Oh, I don't doubt performance will be crap. It's just curiosity about how far you could push something that would be just fine with a heatsink, if it needs it at all.
It seems that it takes liquid nitrogen to really squeeze out everything possible out of a normal Intel/AMD, but that can't be kept up for very long, and is very expensive to try.
Re: (Score:2)
PVRs? (Score:3, Insightful)
Re: (Score:2)
700Mhz Intel Pentium 3
512mb SDram
Nvidia 5200
Avermedia TV card
Windows Media Centre was barely able to run on the machine taking recordings was pointless using windows media centre or the avermedia PVR software. This 500Mhz chip would definitly not be enough, the 1Ghz and 1.5Ghz chips would do it though and even at 7.5watts the 1.5Ghz is redicoulously low power.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
For encoding - just don't bother. I thought that encoding H.264
Re: (Score:2)
Redundand? (Score:4, Insightful)
Linux and Windows CE (or whatever they call it today) run just fine on ARM and similar. Will a low-power x86 compete performance-wise with a low-power RISK architecture?
Re: (Score:3, Interesting)
I agree though, this
Can you provide more details? (Score:2)
Can you tell me which CPU and motherboard you are using? Is Window XP usable on that machine? (I intend to run a flavour of Linux on mine, but your feedback will be valuable anyway)
Re: (Score:2)
No, it does not compete with ARM, MIPS and others at the embebbed market, simply because it consumes a lot of power and is seriously unpowered.
If sucessfull, that chip will open a market on its own for legacy software to run on small power appliances.
Re: (Score:2)
Still, an hypothetical legacy-free x86-like CPU with a redesigned ISA to simplify instruction decoding could shed at least one pipeline stage along with ~10% of the core transistor count and ~10% of the power for
Re: (Score:2)
I wish mainstream CPUs / GPUs would focus on power (Score:5, Insightful)
Re:I wish mainstream CPUs / GPUs would focus on po (Score:2)
Re:I wish mainstream CPUs / GPUs would focus on po (Score:2)
Yes, some people want to buy a video card that requires some amps to play the current games. Some people want to buy a car that performs well. (I'm not talking about SUVs or huge waste hogs)
Everyone pays for the power they consume, be it gasoline or electricity. Who cares?
Re:I wish mainstream CPUs / GPUs would focus on po (Score:5, Insightful)
Exactly. Who cares? People are generally selfish and sometimes you must do things that benefit people as a whole instead of individuals. If slapping a tax on the most energy consuming devices in some category causes people to buy the more efficient ones, that is a benefit to every one. If you still want to buy that device despite the tax then nobody is stopping you. But I guarantee that for everyone who does than many more will choose one which doesn't.
It does not mean either that you're getting a crappier machine as a result. While there is a relationship between CPU / GPU performance and power, I doubt it is a 1:1 mapping. Some processors and GPUs are going to deliver more operations per watt than others. Companies and consumers should be encouraged to favour the more efficient designs over the less efficient designs and a tax for the worst offenders in any class is one way of going about that.
Re: (Score:2)
Re: (Score:2, Insightful)
In some EU countries economic cars have less yearly tax already, I think it's calculated from the CO2 emission pr. km.
And cars that can't perform 15km/l or more, have had their price tax raised, while longer running ones have had it reduced.
Re: (Score:3, Informative)
Re: (Score:2)
A 1.3L must be
Re: (Score:2)
Very nicely, actually. Bear in mind that these engines are in what you'd consider to be small cars; typically two- or four-door hatchbacks. I used to have a Ford Fiesta with a 1.4l engine (IIRC), and while admittedly I'm a conservative driver, I had no complaints about acceleration. (There is one hill near
Re: (Score:3, Interesting)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
I have a 'tiny' 1.9L diesel and shifting at 3000 RPM I can be at 80 or 90 by time I merge and that's carrying around a fat (by european standards) Jetta. But hell, in the states my car counts as compact and I can park in those compact spots.
Re: (Score:3, Interesting)
In Denmark, a sizeable chunk of the total car park are small or family cars with engines in the 1.3-2.0L range. Sporty cars (Alfa Romeo et al, not Ferrari) are probably in the 2-3L range, no more. Of course the SUV-style cars will have way bigger engines (but I suspect that's more to help push the ego rather than the car).
A relevant tidbit: we pay ~7$ per gallon of petrol.
I drive a VW station wagon. It's 4 cy
Re: (Score:2)
I can't see going from 40km/h to 100km/h without hitting at least 3K RPM, unless you're the type that merges on the highway at 20km/h under the limit (of which there are plenty around here). And I get about the same mileage as you do btw (well for highway driving).
Tom
Re: (Score:2)
I can't see going from 40km/h to 100km/h without hitting at least 3K RPM,unless you're the type that merges on the highway at 20km/h under the limit.
Ugh no, I rather dislike that, too. But I do shift as early as possible, and for casual accelleration (that is, NOT highway ramps) I often shift directly from 3rd to 5th (the ratio difference between 4th and 5th is not that great; when accellerating away from a red light I will usually spend only about 1/2 to 1 second in 4th).
... I'll make a note of my rpm at various intervals when I drive home fr
The only time I do high rpms is when I'm in 5th, driving fast. Or so I think! You know, now you got me curious
Re: (Score:2)
Though justice was handed
Re: (Score:3, Interesting)
Get a Diesel e
Re: (Score:2)
When I moved here, I bought myself a 2.0 litre Renault Megane and I've had a couple of people ask me why I didn't buy something with a smaller engine - to me, 2.0 litres IS small! I'
Re: (Score:2)
Power rating (Score:3, Interesting)
Re: (Score:3, Interesting)
With the relatively low cost and high availability of computing speed nowadays, the green5
It'd be good for tiny servers like the $99 decTOP (Score:2)
The 1 watt AMD Geode in the decTOP runs at 366 MHz and makes a fine light-duty server.
Re: (Score:2)
I'm so glad you pointed that out. I shall return my A rated washing machine immediately since clearly the only way it could have gotten that rating is if the engineers dangerously interfered with the power levels required by some of its components. Idiot.
On a somewhat related note (Score:3, Informative)
The release of Vista suggests that we need more and more powerful systems to do our work, but the irony, at least for me, is that I keep buying more of the little guys. Being able to use fanless cases and/or flash drives is a definite selling point, but there's a surprising amount of processing power available in such products and their uses are as limitless as your own imagination. Besides, hacking those ubiquitous blue boxes can never be as satisfying as building your own.
The VIA units I own could be described as underpowered, but having onboard MPEG decoders, for example, can make up for the shortcomings.
Re: (Score:2)
Can you mount the board in a regular ATX case?
Can it be hooked up to a regular ATX PSU?
Who is Soekris? Do you work for them? What's the warranty like?
Does it run Linux? I mean really. lspci output?
Is there video out? How do you 'interact'? COM port?
Seems interesting, but not enough to trawl the website!
Re: (Score:2)
Mount the board in a regular ATX case? I think it comes with a case.
Same with ATX PSU.
Power consumption is (I suppose) in the tens of watts maximum, so ATX PSU is overkill.
I know Soekris as they make boxes which can run OpenBSD and their PF firewalling solution.
If it runs OpenBSD, then you can find some Linux for it.
Once installed, you can access it by network (it has two or three LAN interfaces).
Re: (Score:2)
The new model can fill 100Mbit ethernet. Fairly useful, given that with 4 NICs, you might have more than one network (aside from your slow cable/ADSL link) through which to pass traffic.
Seems interesting, but not enough to trawl the website!
Judging from your laundry list of questions, you're unaware of both Soekris products and their widespread popularity. I'd suggest first visiting the link I provided for people in your position, and then do a Google search for 'Soekr
Re: (Score:2)
lol. Probably the least helpful response I've had on slashdot. Cheers.
Re: (Score:2)
Re: (Score:2)
http://www.atmel.com/dyn/products/tools_card.asp?t ool_id=4102 [atmel.com]
Here is a dev kit for the same chip:
http://www.atmel.com/dyn/products/tools_card.asp?t ool_id=3918 [atmel.com]
It's not an x86 platform, but there is gcc support and it runs linux just fine out of the box.
Check out avrfreaks.net [avrfreaks.net] for a great community based on atmel's AVR offerings (from 8bit to 32 bit). What really draws developers in is how great the community is for these devi
FIRST 1Watt device? (Score:2)
The FreeBSD kernel I run on them seems to think so too.
Kudos to Via for taking ULV to a whole new level and giving us P2-class performance in that watt range, but this is by no means revolutionary, just evolution that allows us to do more with a ULV box.
Re: (Score:2)
I seem to have had the impression that my Soekris firewalls, running a National (today AMD) Geode SC1100 at 266MHz, a P1-class CPU that, coupled with 3 100MBit NICs, 128MB of RAM, IDE, USB etc eat a whopping 3-5 Watts for the entire machine, was x86.
I have a WRAP machine based on a 266MHz Geode. It is not a P1-class machine, it's a fast 486, with some minor tweaks, but still a fairly nice CPU given that it's x86. I believe some of the new Geodes (500MHz sort of speed) are based on a newer design, which is 586-class.
The FreeBSD kernel I run on them seems to think so too.
OpenBSD runs very nicely on them too, which makes them ideal for firewalls. Recently, OpenBSD dropped support for i386, in the basis that it cluttered up the tree (potentially hiding bugs) and only half a dozen people were using it. A
Real Low-Power CPUs (Score:2)
Of course, in consumer electronics devices, you could just use any kind of MIPS or ARM or whatever other CPU you want, and have even lower power usage and/or better performance. It's not like you're gonna be running Windows on these devices, anyway, which is pretty much the only reason you would need an x86 in my book.
Happy (Score:2)
x86 compatible, yes (Score:2)
I mean, 500 MHTZ wi
Re: (Score:2)
With 400Mhz FSB and 128 KiB L1 and L2 they have compromised, but they are more Pentium-M then 286.
The companion North/South bridge chips are even better, with integrated video, network, memory, sata, and hardware MPEG2/4 acceleration for a couple of watts more. You could have a responsive desktop system including memory, video, network etc. etc running ~10W full load (So long as you don't spin up the hard disk). Couple that with very aggressive power sav
Re: (Score:3, Insightful)
So like. . an intel 2 duo that takes a room and miles of cable?
Re: (Score:3, Funny)
Re: (Score:2)
Re: (Score:2)
Every Beowulf cluster comment is automatically redundant because that joke is old and tired. It's not clever; it's overused, and it's irritating that we can't see a single story posted about a processor without some jackass saying to themselves, "lol it wud b sooooo funny if i made a beowulf joke lololol!!!!!!1!!1!!!one"
Re: (Score:2)
Yes, the Geode has generally better performance per MHz than even the VIA C7, according to this comparison of Celeron, VIA C7, and AMD Geode [tolly.com]. Geode beat the VIA C7 on the SPECint and SPECfp at performance per MHz, but VIA C7 beat the Geode slightly at performance per MHz in the CPU score in that article. The article also notes the Geode does not support SSE instructions. On the other hand, Geode also seems to use more power per MHz than VIA's chips, according to the Wikipedia article on Geode [wikipedia.org].