Power-Light Power Chips

DD writes to tell us ZDNet is running a story about a new Santa Clara, CA based startup that is boasting a new line of low-power, Power chips, the same architecture found in current day Macs and IBM servers. From the article: "The company's first so-called PWRficient chip will feature two processing cores, run at 2GHz and consume on average about 5 watts, thanks to an emphasis on integration and circuit design. At a maximum, it will consume 25 watts, far less than the single-core Power chips that can hit 90 watts found on the market today."

Comments

Lower power, but cooler?

[Tikicult]

I wonder if it will run cooler. TFA doesn't say.

Re:Lower power, but cooler?

[P3NIS_CLEAVER]

unless the die is incredibly smaller than a regular POWER chip it will be less.

Re:Lower power, but cooler?

[serialdogma]

ICs pretty much just turn electric energy into thermal energy, power computation wise.

Re:Lower power, but cooler?

[poopdeville]

I don't get this. POWER is an IBM trademark. Macs don't use POWER processors. Does the submitter mean PowerPC?

Re:Lower power, but cooler?

[P3NIS_CLEAVER]

I think POWER is an IBM specification. anyone can build a chip to the spec.

Embedded market

[Thanatopsis]

According to the article they are going to focus on the embedded market. I guess they mean the embedded market that need 2 GHZ embedded chips.

Re:Embedded market

[supabeast!]

From the article:

"The PWRficient actually won't come out for two years, so it's hard to predict exactly how it will stack up against the competition."

In two years a 2 GHZ dual core will probably be a good option for a high-end embedded CPU.

Re:Embedded market

[merreborn]

According to the article they are going to focus on the embedded market. I guess they mean the embedded market that need 2 GHZ embedded chips.

Yeah, who could ever use 2 GHZ processing power? It's not like Tivo's video encoding/decoding takes up any processing time. And lord knows I could never use that much processing power on any sort of mobile computing device, like a Palm Pilot, or Treo.

I mean, who wants to be able to process large amounts of data, fast?

Re:Embedded market

[afidel]

Uses
Print rasterizers: I have printers with imaging engines capable of 30+ppm but I rarely achieve it in the real world because the printers are hobled by a measly ~500Mhz rasterizer.
Networking equipment:If you want to do any kind of complex routing or switching in a truely flexible manner without ASICS you are going to need as fast of a processor as possible.
Complex analyisis of data in an appliance:Antispam appliances are often limited in the algorithms they use because the cost in processing time for some of the better ones are too expensive to apply to the volume of messages they are supposed to handle.
etc.

While I am aware that there are large swaths of the embedded market where nothing more complex than a microcontroller is needed I am also cognizant of the fact that there are many areas where a more powerfull embedded processor which is still energy efficient is still very usefull.

Re:Embedded market

[wowbagger]

Actually, YES, the embedded market that needs 2GHz chips - folks like me doing signal processing for communications, among other things. Do you have any idea how many operations per second it takes to do an echo canceler for a phone, or to do GSM or CDMA decoding in software (if you want a system that can adapt to new protocols - a software defined radio or SDR - you need to use a more general purpose part than the dedicated ICs for this), or to do the latest 802.11 protocols, or to do video decompression, or ....

Yes, Virginia, there is a market for 2GHz processors in the embedded space.

Re:Embedded market

[Red_Foreman]

... in the industrial sector, where decisions may need to be made at real time, yes, a 2GHz embedded processor could be desirable, and there's a bigger market for that than what you likely know.

Sorting machines, monitoring stations, there's a significant market for these processors. Especially at a low power / low heat point.

Re:Embedded market

[Anonymous Coward]

YES! Finally a 2ghz PocketPC!
Now I will be able to write three times as fast with my little stylus!
Will I not?
Am I missing the point???

Re:Embedded market

[TheRaven64]

It's not like Tivo's video encoding/decoding takes up any processing time.

No, it's not if they've got any sense. At least, not general purpose CPU time. Dedicated video compressor / decompressor chips get much more performance per watt, and usually more performance per $ as well when compared to general purpose hardware. The iPod video can play H.264 clips that a moderately fast G4 struggles with - and not because the iPod has a faster CPU.

Re:Embedded market

[TheRaven64]

The chip comes with two 10Gbit ethernet controllers and four 1Gbit ethernet controller on die. I would imagine that just routing packets between these could use a fair amount of a dual-core 2GHz PowerPCs available processing power.

Re:Embedded market

[Thanatopsis]

Jeesh guys - I guess I should have named said markets such as signal processing. I feel like a pinata.

Re:Embedded market

[AKAImBatman]

Actually, YES, the embedded market that needs 2GHz chips - folks like me doing signal processing for communications, among other things.

I'm sorry, WHY are you doing DSP on a general purpose CPU again? Methinks that's the precise reason why DSP vector processors were invented...

Just in case you're being cheeky, comparing a 2GHz DSP to a 2GHz general purpose CPU is a bit disingenous, don't you think? I mean, DSPs can have their clocks ramped up to ungodly levels because they tend to need less sophitication in the silicon than their general purpose counterparts. But if you chose one a DSP processors for a DesktopOS (or even something like a network appliance), you'd get absolutely horrible performance.

Re:Embedded market

[Doctor Memory]

Of course, it's a little harder to upgrade a hardware codec, so you're locked into supporting whatever yesterday's killer format was, instead of what people want today. What you really want is some general-purpose hardware you can reconfigure without too much pain. Say a speedy processor to pump data to a DSP chip or two....

Re:Embedded market

[Anonymous Coward]

I'm sorry, WHY are you doing DSP on a general purpose CPU again? Methinks that's the precise reason why DSP vector processors were invented... (if you want a system that can adapt to new protocols - a software defined radio or SDR - you need to use a more general purpose part than the dedicated ICs for this)

Re:Embedded market

[Jeff DeMaagd]

Flight control electronics use PPC chips. The amount of data to track flight variables increases with every generation of aircraft.

Re:Embedded market

[AKAImBatman]

(if you want a system that can adapt to new protocols - a software defined radio or SDR - you need to use a more general purpose part than the dedicated ICs for this)

That explains why you'd use a Digital Signal Processor instead of a hardware codec. It doesn't explain why you'd use a GP instead of a DSP.

Re:Embedded market

[bsd4me]

Lots of people do DSP on general purpose CPUs. One reason is that some large DSP systems need the features in a full RTOS that only runs on a real CPU. Another is portability: the transition from desktop prototype to real system is often easier when the taret CPU is non a true DSP. Another reason has to do with memory and I/O. General purpose CPUs can have gobs of memory and have an easier interface to things like PCI and other standard busses. DSPs often have limited memory and require custom bridges to get to standard busses.

DSP on GP CPUs isn't always best, but I think you would be surprised at the number of people doing it.

Re:Embedded market

[Doctor Memory]

IIRC, the Power instruction set has some DSP-like instructions (e.g., MAC, multiply and accumulate). And, of course, they could have licensed the AltiVec APU.

Re:Embedded market

[AKAImBatman]

Sounds a) expensive and b) slow. If you just need pluggable codecs, it seems to me that DSPs would be cheaper and faster. One might be tempted to combine the CPU and DSP in something like a cell phone, but it strikes me that simple screen updates, button handling, and other standard chores would impinge on the proceesing ability of the CPU, requiring that the CPU be much larger and more expensive. That can be particularly problematic for cell phones because they have to run on such razor thin profit margins.

Then again, what do I know? I don't design them for a living. It just seems wasteful is all. :-)

Re:Embedded market

[TheRaven64]

Many of these ASICs are in fact FPGAs. In volume, a decent sized FPGA can be bought for under $2. The contents of the chip is stored in a 2Mb flash chip, and can be re-flashed if new codecs are required. Some FPGA manufacturers even give you the design for a simple PowerPC core, so you can run a general purpose OS on part of the chip and feed data to the hardware CODECs.

Re:Embedded market

[AKAImBatman]

SIMD instructions have a lot in common with DSPs. In fact, they really are an attempt to add DSP features to a General Purpose processor with the intent of running multimedia through the main CPU. This has been particularly useful for things like software video players, but CPUs of today still have a hard time competing with dedicated hardware like an MPEG2 hardware codec.

Re:Embedded market

[Anonymous Coward]

It is currently late 2005, not late 2004. This chip is sampling in under a year.

Re:Embedded market

[kurtdg]

You mean the embedded market that needs a chip that can fill a 10Gbps pipe with a choice of IPSec or SSL traffic all on its own?

http://www.realworldtech.com/page.cfm?ArticleID=RW T102405055354

Re:Embedded market

[NekoXP]

Why do DSP on a general purpose CPU?

Usually, to reduce part count. AltiVec on a PowerPC performs better than a dedicated DSP in a lot of operations, for the cost involved, and you get to do all kinds of other stuff with the unit besides specialised FMAC or encryption or so on (which AltiVec has, and does just fine).

If these chips have the AltiVec/VMX unit, are 64bit like they say, and low-power like they say, then they will find uses just like Freescale are pitching 1.5-2GHz single and dual core chips (7448, 8641, 8641D) for the same markets now.

With regards to an SoC with PCI-Express channels though I disagree with the use of power ratings; PCI-Express is power hungry by design.

5-15W for the dual core chip is one thing but I/O voltage is going to be what decides whether they use the chip or not. The Freescale solution is that they can carry on offering the 7448 with the same processor core, which has no real I/O capability, so people can stay with low power peripherals and northbridges. You sacrifice part count for power consumption there.. but trade-offs are trade-offs.

Re:Embedded market

[Chirs]

Embedded does not necessarily mean "low-power" (although it's a benefit).

I'm currently working on an embedded board with 8 GB of RAM and dual Xeons, with 4 gigE ports, as well as another one that has dual 970FX chips and 4GB of memory with 4 gigE ports.

Re:Embedded market

[wowbagger]

Because the difference between a DSP and a normal CPU is very small now-a-days.

It used to be that only DSPs had multiply and accumulate instructions - now many CPUs do (the Power being among them).

It used to be that only DSPs had the register count to do an FFT without having to spill to memory during the butterflys - the Power also has enough registers to avoid having to spill to memory in the innermost butterflys.

It used to be that only DSPs had the fast barrel shifters for single-cycle shifts of more than one bit position - now most CPUs have them.

I can go on and on - but simply put, the only real difference between a DSP and a modern CPU is that very few DSPs are clocked at 2GHz, while many CPUs are.

The really fast DSPs are the ones like the TI C6X family - which get their "speed" from being very long instruction word processors, much like the Itanium. They don't have a very high clock speed - the fastest C6x is running about 1GHz. They are benchmark queens - the will do a 4096 point FFT blindingly fast. Oh, you wanted to do something ELSE with the data after you did the FFT? Sorry, but now you are going to lose most of that speed as the code falls out of cache, and as you run out of vectorizable code and stall most of the cores. Besides, you can get just as much speed-up using the vector instructions of a modern CPU (Altivec/SSE etc.) as you do from the C6X processor.

They also suck when you are doing protocol as opposed to signal processing - DSPs *hate* jump instructions, and don't EVEN think of asking them to do a context switch - they are like a drag racer, they go fast until you ask them to TURN.

In short, the days of the DSP as the king of signal processing are past - you can do more with a general purpose processor and an FPGA than you can with DSPs for the same amount of board real-estate, bill of materials cost, and power consumption.

Sorry, but since this is actually what I do for a living, I know from first-hand experience that DSPs really aren't all they are cracked up to be with respect to regular processors now-a-days.

Re:Embedded market

[AKAImBatman]

Fair enough. But it still seems like you're replacing a rather inexpensive part ($1-$5 DSP or $3-$10 FPGA) with a God-awefully expensive CPU ($20-$50). Perhaps you're making up by combining enough hardware? Also, doesn't a regular processor like a PowerPC or x86 significantly complicate the Bus design? A lot of embedded hardware tend to use a custom bus, or at least a far simpler design than a Desktop PC. Unless you have a customized CPU with simpler pinouts, you're going to need a bus similar to the one the CPU is expecting.

It used to be that only DSPs had the fast barrel shifters for single-cycle shifts of more than one bit position - now most CPUs have them.

I was under the impression that barrel shifters were relatively uncommon on CPUs. For example, the Pentium IV actually removed the hardware barrel shifter that had been added to the original Pentium, and reimplemented it in microcode. Google is being unhelpful at the moment, but I seem to remember that the PowerPC either lacked a barrel shifter, or that the performance was considered sub-par. Is this just my imagination?

Re:Embedded market

[wowbagger]

"But it still seems like you're replacing a rather inexpensive part ($1-$5 DSP or $3-$10 FPGA) with a God-awefully expensive CPU ($20-$50)."

Oh, that was FUNNNEEEEEYYYY! Tell me another one, please!

The 400 MHz version of the C6x runs about US$21 in 1K quantity.

There is no price advantage, no real performance advantage, no real board real-estate advantage, no real power advantage to DSPs. They are harder to do REAL work with than general purpose CPUs. The flat SUCK at context switching or interrupt service.

Like I said - I do this for a living - have been for over a decade. I am not just some /bot wanking off in his parents' basement spouting off on subjects he read about on Wikipedia. I've designed several pieces of gear, some of which have flown on the Shuttle, and which have sold many tens of millions of dollars worth of gear. I work with prototype silicon from TI, IBM, Xilinx, Altera, and other vendors.

Re:Embedded market

[farnz]

Even if your figures are reasonable (and as wowbagger's said, they're not), don't forget that you're not replacing a cheap DSP or FPGA with an expensive CPU. You're replacing a cheap DSP, cheapish CPU and associated glue components with an expensive CPU. When you're removing a $5 DSP, a $10 CPU, $5 of glue logic to join them, and $5 of PCB and manufacturing costs to go down to a $25 chip, you're saving nothing per-unit, but simplifying your development costs. Chances are that over the lifetime of your unit, the $25 chip, the $5 DSP and the $10 CPU are the only items that'll fall in price; if they all halve, you've made a saving.

Re:Embedded market

[LionKimbro]

One day, we'll be automatically generating 3D textured models by just waving our mobiles around the objects. Already in Japan, they're OCR'ing text off printed pages with their mobiles.

There are plenty of things to do with computation in mobiles...

Re:Embedded market

[Bastian]

It actually reminds me of two SiliconGraphics Indigo2s (the teal ones) that we had at work. One had the the 2D graphics accelerator, and the other had the low-end 3D accelerator.

While the one with the 3D acceleration was super-fast for wireframe work, rendering any 3D graphics with a fill was noticeably faster on the computer with the 2D card. The problem was that there was no 2D acceleration on the 3D card. Any speedup you gained on the 3D coordinate transforms was more than lost when the time came to draw all those pixels.

Re:Embedded market

[Bastian]

True, but only if you're in the kind of application domain where there are cheap off-the-shelf hardware-based options.

If you're a relatively small company making embedded products that aren't designed for the consumer market, chances are designing and fabricating all of that custom hardware is going to be much, much, much, much more expensive than simply slapping in a fast CPU and doing most the work in software. MAYBE an FPGA or two might be useful for some of the tasks that are easier to implement in hardware.

Also, the TiVO does use hardware video encoding/decoding.

Re:Embedded market

[Superfarstucker]

This reminds me about a story of an engineer, a toaster, a king and a computer 'scientist'.

Once upon a time, in a kingdom not far from here, a king summoned two of his advisors for a test. He showed them both a shiny metal box with two slots in the top, a control knob, and a lever. "What do you think this is?"

One advisor, an engineer , answered first. "It is a toaster," he said. The king asked, "How would you design an embedded computer for it?" The engineer replied, "Using a four-bit microcontroller, I would write a simple program that reads the darkness knob and quantizes its position to one of 16 shades of darkness, from snow white to coal black. The program would use that darkness level as the index to a 16-element table of initial timer values. Then it would turn on the heating elements and start the timer with the initial value selected from the table. At the end of the time delay, it would turn off the heat and pop up the toast. Come back next week, and I'll show you a working prototype."

The second advisor, a computer scientist , immediately recognized the danger of such short-sighted thinking. He said, "Toasters don't just turn bread into toast, they are also used to warm frozen waffles. What you see before you is really a breakfast food cooker. As the subjects of your kingdom become more sophisticated, they will demand more capabilities. They will need a breakfast food cooker that can also cook sausage, fry bacon, and make scrambled eggs. A toaster that only makes toast will soon be obsolete. If we don't look to the future, we will have to completely redesign the toaster in just a few years."

"With this in mind, we can formulate a more intelligent solution to the problem. First, create a class of breakfast foods. Specialize this class into subclasses: grains, pork, and poultry. The specialization process should be repeated with grains divided into toast, muffins, pancakes, and waffles; pork divided into sausage, links, and bacon; and poultry divided into scrambled eggs, hard-boiled eggs, poached eggs, fried eggs, and various omelet classes." "The ham and cheese omelet class is worth special attention because it must inherit characteristics from the pork, dairy, and poultry classes. Thus, we see that the problem cannot be properly solved without multiple inheritance. At run time, the program must create the proper object and send a message to the object that says, 'Cook yourself.' The semantics of this message depend, of course, on the kind of object, so they have a different meaning to a piece of toast than to scrambled eggs."

"Reviewing the process so far, we see that the analysis phase has revealed that the primary requirement is to cook any kind of breakfast food. In the design phase, we have discovered some derived requirements. Specifically, we need an object-oriented language with multiple inheritance. Of course, users don't want the eggs to get cold while the bacon is frying, so concurrent processing is required, too."

"We must not forget the user interface. The lever that lowers the food lacks versatility, and the darkness knob is confusing. Users won't buy the product unless it has a user-friendly, graphical interface. When the breakfast cooker is plugged in, users should see a cowboy boot on the screen. Users click on it, and the message 'Booting UNIX v. 8.3' appears on the screen. (UNIX 8.3 should be out by the time the product gets to the market.) Users can pull down a menu and click on the foods they want to cook."

"Having made the wise decision of specifying the software first in the design phase, all that remains is to pick an adequate hardware platform for the implementation phase. An Intel 80386 with 8MB of memory, a 30MB hard disk, and a VGA monitor should be sufficient. If you select a multitasking, object oriented language that supports multiple inheritance and has a built-in GUI, writing the program will be a snap. (Imagine the difficulty we would have had if we had foolishly allowed a hardware-first design strategy to lock us into a four-bit microcontroller!)."

The king wisely had the computer scientist beheaded, and they all lived happily ever after.

BUT... BUT..!!!

Re:Embedded market

[Anonymous Coward]

dang...you sure do get polite when somebody with a clue calls you on your craploads of hot air.

see, you should start off that way: polite and seeking knowledge (as opposed to aggressively trying to lay the smack down) ...you're well-informed for a non-expert...that's nothing special and you need remember that. it builds character.

Re:Embedded market

[PeteABastard]

The parent is completely correct; My former employer was using COLDFIRE processors with FPGA's to run fruit sorting machines. It was far cheaper for us to use a general purpose CPU to do exactly what we wanted. One of our biggest limitations at the time was lack of fast floating point calculations we needed for video processing. The low power requirement and a code base written in COLDFORTH kept us on the COLDFIRE platform.

We were migrating to uCLinux to allow quicker development, in a couple of years this chip family could be a good growth path.

Peter

Re:Embedded market

[Jeremi]

button handling, and other standard chores would impinge on the proceesing ability of the CPU, requiring that the CPU be much larger and more expensive

Nah, not if it's done right. You either use a real-time OS and run the audio processing in a real-time priority thread, or just do the audio processing in an interrupt. The button handling and other stuff happen in lower priority threads, so no matter what they will never slow down audio processing. (audio processing might slow down button handling, of course, so you just have to be careful to make sure your real-time code is well-behaved)

It just seems wasteful is all.

Not at all -- what would be wasteful is putting in two chips that each spend all their lives idling or at 50% utilization (read: 50% wasted). The ideally efficient system would include exactly as much processing power as needed, and not a bit more.

Re:Embedded market

[eonlabs]

Come on, in two years, you're going to need 2ghz in your fridge so the new version of windows can keep your food cool.

Re:Embedded market

[Anonymous Coward]

Right on the money! TCP/IP attack mitigation appliances with packet inspection need
a pair of 1.6 GHz processors. A pair of 1.6 GHz processors are needed to keep pace with 4 gigabit/sec ethernet NICs. Fun stuff!

What the future may hold

[xgadflyx]

Well with a savings on average of 2.64 million dollars per year, it's really interesting to see where the performance hits are and if they're acceptable or even true. I could say/promise anything I wanted given a two year window to produce. If all does prove to be well, this rollover savings might fuel the movement for a solar-powered farm (ha). Shh..don't tell the tree huggers just yet.

Re:What the future may hold

[Technician]

If all does prove to be well, this rollover savings might fuel the movement for a solar-powered farm (ha). Shh..don't tell the tree huggers just yet.

Shh... Don't tell the geeks, but all the farms here depend on solar power. There is not much that grows in the dark. ;-)

Re:What the future may hold

[Anonymous Coward]

Mushrooms maybe?

Re:What the future may hold

[Technician]

Mushrooms maybe?

Solar heat. They don't grow in permafrost. ;-)

It's Like Free Money!

[Anonymous Coward]

I'll wait to see benchmarks before I decide it's better. Nothing new about low performance + low power, so until I believe that it's high performance + low power, I'll wait.

Amazing

[Namronorman]

This is simply amazing, and if they're even remotely as powerful compared to their future competitors and their initial cost is not so bad, you could easilly factor in the energy savings for spending more on hardware versus spending more on electricty.

Re:Amazing

[Thanatopsis]

It is amazing expecially when you consider that many current Intel chips suck down 150 watts at 2.8 GHZ. This isn't like Transmeta either. The team at PA Semi are some pretty heavy hitters in the chip design world.

Re:Amazing

[Anonymous Coward]

It's not so amazing when you consider that your post is a troll... a dual core Xeon with several
megs of cache is not in the same class as this embedded chip. It's even less amazing when you consider
that Intel Pentium-M's that you can buy today (not 2 years from now but today) are at least as fast as this
chip and use about the same amount of power.

Re:Amazing

[masklinn]

Pentium-M's TDP is 27W, Turion64 MT are ranked for 25W

AMD's Geode LX on the other hand has a 2.4W TDP and 1.6W typical consumption @500MHz

Re:Amazing

[hattig]

Except we should compare like with like.

This new processor gets >2000 SPECfp per core, >1000 SPECint per core, has 2MB shared L2 cache, dual 10GigE and Quad GigE interfaces, TCP/IP offload, security offload, etc. All in 5-13W typical.

Dual-core Xeon has ... two cores, 2MB L2 per core, no TCP/IP offload, no security offload. All in 135W typical.

Sorry, you can't get dual-core Pentium M processors yet, certainly not ones with 64-bit capability or SIMD instructions. By the time this new processor comes out, Merom will be coming out of course, and include these. However Merom is 35W max, this chip is 25W max and includes an entire northbridge and more as well.

Re:Amazing

[Aqua OS X]

I don't know about "simply amazing"... a cure for cancer would be simply amazing. This is more along the lines of "pretty cool." ;)

Re:Amazing

[NekoXP]

It's not really amazing, PowerPC is meant to be low power, it's a design feature of the core across all the processor lines. Why does the 970 take 90W? Because it has a front side bus that is half the core clock and two FPUs with a seperate 4-unit vector processing system :)

The front side bus and it's i/o voltage and I suspect whatever serialiser/deserialiser is stuck to it is what gives the "unwanted" power consumption in my opinion. Moving the memory controller inside the chip removes a high speed external bus (and all it's problems) from the equation, allowing it to be clocked at core speed and leaving only the RAM interface as being power hungry.

the irony!

[Foktip]

oh the irony... this must be true since it follows murpheys law perfectly.

Apple

[ZachPruckowski]

So are they going to be regretting moving away from that? I mean, that would have an appeal in a low to middle end laptop that can run for 12 hours or something. I'd certainly pay for it. I'm impressed with my iBook battery as it is, but it is just shy of being able to cover all my needs in a day. Or at least, usually have to think about charging it. An 8 hour laptop would be great for people on the move, like students, or amateur filmmakers.

Re:Apple

[masklinn]

Ever heard of AMD's Geode chips?

Re:Apple

[Jeff DeMaagd]

The problem of excessive CPU draw has been taken care of by staying away from the "desknotes". The current limitation of battery life is not CPU draw, but LCD draw.

Cutting CPU current draw by half will not net a doubling of the current battery charge life.

Re:Apple

[tktk]

I'd only start regretting once PWRficient chips started shipping and in quantity. Course, I know that Mactels systems aren't shipping yet either, but it's too early to make a judgment either way.

Re:Apple

[TinyManCan]

I don't think that thew CPU is the main current draw in laptops. It used to be the LCD display (+ backlight) ate the huge majority of juice, but the new GPUs are proficient power eaters as well.

With the direction OS X is heading, the GPUs are going to be seeing more use than processors in the future.

I.E. I still think Apple is headed the right direction with OS X on Intel. Intel has the best chance for developing a chipset (including graphics) that provides the horsepower Apple needs along with the the ability to scale power requirements down.

Re:Apple

[wvitXpert]

This isn't going to be a laptop class chip, when it comes out it will be better suited to herald the comeback of the Newton.

Re:Apple

[Egregius]

They won't regret moving away. This startup is aimed at embedded chips, not desktops. Furthermore, Apple hasn't completely *abandonned* Power just yet. And it's not like Apple can base it's productline on a small start up with no real guarantees it can meet Apple's demands.

Re:Apple

[yamla]

Hardly. These chips aren't due out for two or three YEARS. Let's assume for the sake of argument that they ship IN BULK in two and a half years, an obviously optimistic estimate. Should Apple be satisfied with dual-core 2 Ghz laptops in the spring of 2008? I certainly hope not. While the power usage is sweet, we are looking at less than a 20% increase in speed (assuming you can safely compare clock speeds which, as we know from Intel and AMD, is not a good assumption) for a single core over that time. Even with dual core, that's pretty pathetic.

Now, if these chips were shipping in bulk TODAY and were able to be ramped up to 3 or 4 Ghz over the next six to twelve months, then maybe Apple might start regretting moving away from the G4 and G5 CPUs. That is, it'd be a toss-up at that point. As it is, this is far too little too late for Apple's laptops.

Of course, this rests on the assumption that Apple cares about processing power.

Re:Apple

[Jozer99]

2008? By then, we will have 1.8 GHz powerbooks! The world we live in today!

Re:Apple

[hattig]

PA Semiconductor Release

THE PWRficient PROCESSOR ROLLOUT

The first PWRficient chip, the PA6T-1682M, which dissipates between just 5-13 watts, depending upon the application, is a dual-core implementation running at 2GHz with two DDR2 memory controllers, 2MB of L2 cache, and a flexible I/O subsystem that supports eight PCI Express controllers, two 10 Gigabit Ethernet XAUI controllers, and four Gigabit Ethernet SGMII controllers sharing 24 serdes lanes. It will sample in the third calendar quarter of 2006, with single-core and quad-core versions due in early and late 2007, respectively, and an eight-core version planned for 2008.

So please can you explain your "These chips aren't due out for two or three YEARS" statement.

yes, it is another shitty slashdot story that links to the worst possible source for a story whilst ignoring all of the more useful possible sources, including the people themselves.

Re:Apple

[yamla]

Well, late 2007 is two years away, and I was talking about the dual-core models. The question of the day, though, is when the company would be able to ship these IN BULK. Do you really think they'll be shipping the dual-core models by the end of 2007 in sufficient quantities for all of Apple's laptops?

Re:Apple

[Bloater]

That is LCD backlight draw. Roll on long life LEPs - or give me 5 cheap and easily swappable screens. They got the right wavelengths to replace LCD, they just haven't got the lifetime. But the extra waste of disposable screens can be easily offset by the greener production process - and the money made could help produce longer lived screens sooner. Gimme gimme gimme gimme gimme gimme.

Re:Apple

[hattig]

Reread it.

Sampling 3Q 2006. That's under a year away.

2006 != 2007

Sure, it is only sampling, but will they be sampling it for a whole year?

Re:Apple

[yamla]

Apparently, yes, because they haven't scheduled the dual-core processors until late 2007, in other words more than a year later. And will they be able to produce the chips in sufficient quantity should Apple go with them? Not in 2007, you can be sure.

Re:Apple

[masklinn]

failed, 2007 is the deadline for the single core, dual core will be the first released at Q3 2006.

Re:Apple

[yamla]

You are right, I was misreading quad core as dual core.

I still don't think this would make much difference even if Apple had known about this last spring. But it isn't as bad as I had initially thought.

Re:Apple

[hattig]

Indeed, it is too late for Apple right now ... just as they switch to Intel something comes out that *may* be vastly more performant per Watt and far more integrated (meaning cost savings).

Anyway, for more information on the core of this new processor: http://www.realworldtech.com/page.cfm?ArticleID=RW T102405055354&p=3

The last line is very interesting: "per core typical power at 4W and worst case at 7W"

I feel a great disturbance in the Force

[katana]

As if millions of Apple customers suddenly cried out, and were silenced.

Re:I feel a great disturbance in the Force

[John Nowak]

"Nooooooooooooooo!" - Steve Jobs

Re:I feel a great disturbance in the Force

[Salandarin]

"As if millions of Apple customers suddenly cried out, and were silenced."

I think you mean "As if dozens of Apple customers..."

*hides*

Re:I feel a great disturbance in the Force

[Anonymous Coward]

What? Did the fans fail?

Re:I feel a great disturbance in the Force

[masklinn]

Flash news, Steve Ballmer has just vowed to Fucking Kill(TM) PWRficient Chip. He will Fucking Bury(TM) them. He has done it before, he'll do it again.

Re:I feel a great disturbance in the Force

[slimak]

As an apple customer, I can tell you that I don't give a crap what kind of chips they use as long as
1) I can get a faster powerbook
2) The cost does not go (any) higher
3) Any migration is relatively painless
4) The products still look cool

actually, now that I think of it, I really only care about #4

Re:I feel a great disturbance in the Force

[plasmacutter]

As another apple customer, i care about the fact that apple does not currently use kludgy DRM ridden intel chips, you do know the "D" in pentium D stands for DRM, and that apple is apparently depending on treacherous computing to do their dirty work.

What's to stop the temptation to use treacherous computing further to implement restrictions as per windows vista?

Apple's commercials were all about "think different"... now apparently they've forgotten what that slogan means.

finally

[kevin.fowler]

What a relief. Implement this en masse and a dormitory full of idling computers running aim won't use as much energy as a small country anymore.

Re:finally

[joib]

Considering the sort of movies they dl and watch at those places I guess heat won't be a problem

I'm just wondering:

[mctk]

How much power do processors use relative to the rest of the computer? It seems that hard drives and fans would use the majority of power (not to mention monitors and speakers if present).

Re:I'm just wondering:

[Wesley Felter]

A hard drive only uses about 10W, and a typical PC only has one. IIRC, fans use about 1/3 as much power as the components they're cooling. So processors are still using a large fraction of a computer's power.

Re:I'm just wondering:

[Anonymous Coward]

Obviously, it depends on the hardware. But for current desktop computers, processors use up about half the energy consumed.

Re:I'm just wondering:

[masklinn]

It seems that hard drives [...] would use the majority of power

The average 3"5 (desktop) hard drive (aka 7200RPM SATA/ATA133) runs around 7W idle and about 10W in seeking, high-perfs being a bit higher (12W seeking for 72Gb 10000RPM Raptor drive)

Notebook 2"5 5400RPM drives run around 1W idle (0.8W for a Samsung M40 MP0402H) and around 3W seeking.

It seems that [...] fans would use the majority of power

The fans I can check right now all fall between 0.15 and 0.30A, 12V.

This means that running them at max tension (12V) you're looking at 1.8W to 3.6W. Undervolt them at 7V and you fall between 1 and 2W.

And these are specs for 80mm to 120mm fans

So no, hard drives and fan often ain't the worst offenders as far as power consumption goes.

Re:I'm just wondering:

[toomim]

Nope, the display backlight uses the majority of the power. (See data here.)

• Power usage difference from display backlight:
  • 18% (Set at 1 bar)
  • 36% (Set at 8 bars)
  • 60% (Full Intensity)
• From running iTunes:
  • 12% (Reduced Performance)
  • 24% (Highest Performance)
• Airport: 14% (Relative to Airport off)
• Internal Speakers: 3.5% (Relative to headphones)
• Keyboard backlight: 3% (Keyboard lit all the time)
• Bluetooth: 2% (Relative to Bluetooth off)
• USB Load: 0% (A powered hub draws no current from the computer)

Posted by the CEO?

[Anonymous Coward]

I wonder if the orignal poster "DD" is "Dan Dobberpuhl", the CEO of the company, looking for free advertising on slashdot?

Re:Posted by the CEO?

[WindBourne]

And you would be surprised? There are many marketers that hit here. Some deliver useful info, and others deliver FUD. So even if it is the CEO, I say great. Hopefully, it is one with a clue and will answer questions that are posed here.

Re:Posted by the CEO?

[hattig]

Very unlikely, if it was him he wouldn't have linked to the WORST story about the new processor you are likely to find.

Re:Posted by the CEO?

[diegocgteleline.es]

If it was him, he'd probably know that we love Alphas, and he was one of the main designers behind DEC's Alpha, and lots of their engineers are ex-DEC engineers

This vs ARM Cortex A8?

[Sam Haine '95]

I wonder how this will compare to the ARM Cortex A8 in 2007?

Re:This vs ARM Cortex A8?

[WouldIPutMYRealNameO]

It doesn't. The ARM chip is in a different market, they're talking 0.3W and the PPC chip is talking 5-25W, an order of magnitude difference.
Some of the guys involved in this startup worked on the Alpha & StrongARM chips - both chips were ahead of their time & very good CPUs.
I think they'll make nice PPC chips in a couple of years time.

Re:This vs ARM Cortex A8?

[Egregius]

Well it will be interesting: one chip at 1 GHz on 300 milliwatts, vs 1 dualcore 2 GHz at 13 watts. I'm assuming the arm processor can be used for SMP as well. Any chance any of these will be used for laptops (running Linux/BSD)?

Re:This vs ARM Cortex A8?

[Sam Haine '95]

Any chance any of these will be used for laptops (running Linux/BSD)?

ARM are big in PDAs, so maybe a tablet/laptop.

G5 Powerbooks!

[ejoe_mac]

Come on, please let this be true!

Vaporously Delicious

[brogdon]

"The company's first so-called PWRficient chip will feature two processing cores, run at 2GHz and consume on average about 5 watts, thanks to an emphasis on integration and circuit design. At a maximum, it will consume 25 watts, far less than the single-core Power chips that can hit 90 watts found on the market today."

Also, thanks to our patented Vapor-based architecture, we've been able to build our level-2 RAM cache out of a giant cloud of gaseous water! And we've licensed our chips to be in the Phantom Game Console! And they'll even run Duke Nukem Forever! As we speak the SCO group is printing out some infringing Linux code with them to use as evidence in an actual trial!

Re:Vaporously Delicious

[Thanatopsis]

Well except the guys at PA Semi have actually designed and shipped chips.
Here's some Bios
Dan Dobberpuhl, President and CEO

Dan Dobberpuhl, President and CEODan Dobberpuhl, who cofounded P.A. Semi in July 2003, has been credited with developing fundamental breakthroughs in the evolution of high-speed and low-power microprocessors. Prior to founding P.A. Semi, Dobberpuhl was vice president and general manager of the Broadband Processor division of Broadcom Corporation. He came to Broadcom via an acquisition of his previous company, SiByte Inc., founded in 1998, which Broadcom acquired in 2000. Before that, Dobberpuhl worked for Digital Equipment Corporation for more than 20 years, where was credited with some of the most fundamental breakthroughs in microprocessor technology. In 1998, EE Times named Dobberpuhl as one of the "40 forces to shape the future of the Semiconductor Industry." In 2003, he was awarded the prestigious IEEE Solid State Circuits Award for "Pioneering design of high-speed and low-power microprocessors."

Dobberpuhl holds 15 patents and has many publications related to integrated circuits and CPUs, including coauthorship of the seminal textbook Design and Analysis of VLSI Circuits, published by Addison-Wesley in 1985. He holds a bachelor's degree in electrical engineering from the University of Illinois.

Nah - he knows nothing about processor design - but random dude at slashdot know more.

Rest of the team's bios

BSD

Re:Vaporously Delicious

[Muevelo]

Anyone who buys into this, please contact me asap. I have a wonderful bridge in brooklyn for sale (CHEAP!) that might interest you... *ROLLEYES* LOL

Re:Vaporously Delicious

[ron_ivi]

A great interview with Dan Dobberpuhl and David Ditzel (Transmeta's CTO) back in 2003 where he basically pre-announced this startup.

Dobberpuhl has a nice record of creating awesome chips and selling them for a lot (IIRC he sold his last chip company to Broadcom for $2billion) - and has pretty much re-assembled many of the better parts of the StrongARM team for this one - so I bet it does well.

I predict IBM'll buy them for their answer to StrongARM.

Re:Vaporously Delicious

[brogdon]

Just a joke, dude. What is this guy, your cousin or something? :)

Vaporous Bullshit

[FatSean]

They haven't even begun working on this. They just now came up with the PLAN!
I guess there isn't enough iPod or Blog news today to keep the Slashdot main page moving...

Re:Vaporous Bullshit

[Anonymous Coward]

Uhm, no.
When /any/ fabless or fab-holding chip company announces a new processor, they've already run it through /at least/ one spin, perhaps two, in a test plant.

Not only have they designed the processor, and exhaustively tested it on simulation tools, but they've also ran test batches - probably at two different small fabricators - and exhaustively tested the production of those runs.

The two year lead time is because they must shop around to find a fabricator that /can/ make their processor, to their specs (65 nm process node), and will do it in their time frame. Oh, and there's also making sure the first batch from the production fab isn't shite, either - They're claiming low power dissipation, and one requires expensive, difficult-to-produce near-flawless silicon to do that. Which takes time.

I work as a semiconductor researcher - which makes it /de rigeur/ that I post this as A.C.; When these guys popped out of stealth mode today, I stopped what I was doing to read and re-read their site - as did everyone I work with.

They are /big news/. Their product is the prodigal son, the promise of Transmeta with an environment (economic, technological, and market) that will welcome them. These, or something /very like/ these, will be at the heart of the much touted 'ubiquitous computing', and will be the processors your children's children are coding for. Dedicated ASICs, and even FPGAs just aren't cutting it for what the industry wants to do three years from now - in terms of flexibility, speed, and power consumption.

Everyone is clamoring over the time it takes to design and spin chip logic, as well as the failure rate at new process nodes. Every new piece of dedicated logic requires extensive, and expensive, preliminary verification - and may fail when coming off the line anyway, due to an inability to accurately model the physics of the chip.

The answer is, of course, the time-honored cycle of moving back to general-purpose chips, and developing your logic as software.

And if a general-purpose chip can do what you want, at the speed you want, at the power consumption you want - why run through chip production when you have verified, proven silicon?

But will it have AltiVec?

[lax-goalie]

The big question is will it have a vector processor? If so, it could end up in an Apple design, if for no other reason, to keep the pressure on Intel. If not, this is simply another PowerPC embedded CPU...

What a business plan!!!

[IGoChopYourDollars]

1) design a low-power-consumption high-performance PowerPC chip that would be ideal for Apple to use
2) keep the development so secret that spouses are kept in the dark
3) launch the product after Apple has already abandoned PowerPC
4) ???
5) PROFIT!

Re:What a business plan!!!

[Anonymous Coward]

You look pretty stupid when you don't read the fvcking article, or understand what 'embedded' means.

Apple compatability (or Windows for that matter) DO NOT MEAN SHIT in the embedded space. Want to know why? Google some and learn, instead of being a dolt with a submit button.

Re:What a business plan!!!

[wannabgeek]

Just because you're a fanboy of Apple, doesn't make Apple the be-all and end-all of PowerPC. Apple's share in the whole PowerPC market was miniscule (less than 5% AFAIK). These guys can still make profits big time without caring a fuck about Apple.

But yeah, this is /. and if you say the world comes to an end if it doesn't care for Apple, you are bound to be modded up.

I, for one

[PresidentEnder]

Welcome our new high-efficiency processor overlords.

Re:Fisht Posht!

[Anonymous Coward]

lol bs evarayone nows that teh amd rul0rx!!!1!

Re:I call BS

[Duhavid]

Since you called, you must know the number... What is it?

P-p-p-power

[InvisibleSoul]

Maybe this new Power chip can power the P-P-P-Powerbook...

http://www.p-p-p-powerbook.com/

Embedded, as in, not Asus

[Anonymous Coward]

By "embedded" they mean "niche".

That's not really fair, now that I think about it.

They just don't expect to sell to someone building a PWRficient-based PC. They want to sell to people building small, powerful, Linux-based devices like wireless routers and cow-milking machines, that operate in environments where a 200W CPU is a problem.

To try and forestall the Apple hand-wringing...

[diamondsw]

...since this is an embedded processor, it likely won't have things like vector processing units (Altivec) or possibly even out of order execution (as I recall the XBox 360 dropped). Take that along with the fact that it's a couple years away, and it really doesn't affect Apple at all.

Re:To try and forestall the Apple hand-wringing...

[klapton]

That and only 2GHz two years from now is just too little, too late. There won't be a desktop or laptop market for a processor that lags in performance as soon as it hits the street. And being a new design, it will likely be out of the cost-conscious market range as well.

Yes

[Wesley Felter]

According to the Web site it has AltiVec. By 2007 I think Apple will have switched completely to Intel, never to look back.

That would be my bet.

[Anonymous Coward]

I wonder if the orignal poster "DD" is "Dan Dobberpuhl", the CEO of the company, looking for free advertising on slashdot?

Startups just can't get enough Slashvertising you know. This article should be moderated. Score: -5 Slashvertisement

1. Make bogus hardware claims.
2. Get free advertising on Slashdot.
3. Profit
4. Fold.

Re:Amazing?

[TopSpin]

This is simply amazing

What is simply amazing? At this point we some handwaving about "emphasis on integration and circuit design." Projections from venture capital PowerPoint presentations tend not to pan out in production units. This looks like Transmeta II, except they are aiming lower.

Honestly, I don't care...

[OneByteOff]

I've been using a laptop in some fashion for the past 5 years. Honestly to me power and peformance are significantly more important than battery life. While I may stand alone or even stand with a small crowd of like minded people, I believe that current battery life is sufficient. Honestly I can't imagine sitting anywhere for more then 2 hours (roughly my Inspiron 9300's battery life) and if I did find myself in that situation I'd just power down, pop in my spare battery and go about my business for another two hours.

What I believe is important and newsworthy is the introduction of Dual-Core Laptops and Dual-Dual-Core Laptops which may not get 2-3 hours battery life, but can be used by power users to get close to desktop performance out of their laptop. When I'm onsite at a clients and need to run John the Ripper or encrypt/decrypt some folders in Windows, I really really wish my notebook had more raw processing power.

Re:Honestly, I don't care...

[Anonymous Coward]

Do you think he's speaking honestly?

Re:Honestly, I don't care...

[SlightlyOldGuy]

Two hours? Sounds like your computer is more of a portable than a real laptop. So yeah, go for the fast, hot CPU. My neighbor has a "laptop" like that. Whenever he carries it across the room he takes the power brick along and plugs in. Not much good for cross country flights, though. You'll even be looking for somewhere to plug it in at your local coffee shop, if you ever do the working-at-the-cafe thing. Usage profiles vary.

Re:Honestly, I don't care...

[Anonymous Coward]

"Clients" run "John the Ripper"? WTF!?!!

Eeehhhh..... What line of work are you in again?

25w is way too high!

[mattnuzum]

I just designed a complete computer that uses less than 3 watts! (more details)

Admittedly, it probably does far less than a power based computer. It runs at 1 MIPS, has only 64 bytes of RAM and spends most of its time sleeping, but on the plus side, it costs less than $10 to build and while sleeping uses about .05 watts of power.

Imagine a beowolf cluster of these babies!

Re:25w is way too high!

[tzanger]

Sorry to tell you this, but your LEDs won't turn on, ever. And you're using more power than you need anyway. The only transistor necessary in that entire design is the one driving the speaker, which may not work anyway due to inductive and/or high current.

Re:25w is way too high!

[Jeff DeMaagd]

The chip that you use should only consume a few nanowatts of power, so there's room for refinement.

I've used those 8 pin PICs, they are pretty nice. I ran straight from battery power, a linear regulator is too wasteful. It was for a hazardous material area where running power was undesirable and the battery cell had to last a year. I powered sensors directly from the output of one pin, so I can turn the sensors on only when taking a reading. Obviously, it's only useful for low current devices, but other than maybe the speaker, I see no high current devices preventing the technique from being used on several of the inputs.

POWER != PowerPC

[frankie]

These are (theoretically, since they don't exist yet) based on the POWER architecture used by IBM big iron servers, which is related but incompatible to the PowerPC chips in Macs. Different pinouts and almost certainly no Altivec.

Perhaps if this company had existed a couple years ago, Lord Steve might have given them an audition before jumping to Intel. But even if they somehow got their current chips to mass production in industry-record time, they would still be years away from being able to ship a PowerPC version.

Re:POWER != PowerPC

[bloosqr]

They are reasonably binary compatible. That is the PPC instruction set is a subset of the PWR instruction set. Moreover with the default xl* compiler settings the same executable could be used on any of the of PWR and PPC chips. They are compatible in the sense that the intel chips themselves are compatible :

i.e. the entire line of Socket A,1,2,3,4,5,6,7,8,370,423,462,478,603,604,754,775, 939,940,M pin incompatible chips that may or may not have "extra" commands (like MMX, 3DNow, altivec or even hardware floating point) but otherwise reasonably compatible.

Re:POWER != PowerPC

[greed]

IBM dropped the classic meaning of the POWER term with the Power3 CPUs; they are actually from the 64-bit PowerPC family.

Comments about pinout and AltiVec are relevant; TTBOMK, the CPUs IBM is using do not have the same SIMD unit that the Mac G[45] chips use.

And I would sure hope there's a different pinout on the embedded chip....

What really makes the two lines mostly distinct is the string ops in POWER which are not present in the PowerPC family (except the 601), and the presence of single-precision floating point in PowerPC which POWER did not have. (Resulting in a SPECfp "optimal option" flag of "-Dfloat=double" for performance reasons....)

There was enough similarity in the instruction set and CPU architecture that other distinctions are more of a nit-pick; like changing the assembler mnemonics for some operations. The PowerPC is bi-endian; the POWER is big-endian; some of the "lost" instructions on PowerPC are only lost in little-endian mode....

OK, some of those are big nits....

Re:POWER != PowerPC

[raftpeople]

From the IBM website:

About Power Architecture
Power Architecture(TM) technology is an instruction-set architecture that spans applications from consumer electronics to supercomputers. Power Architecture encompasses PowerPC®, POWER4(TM) and POWER5(TM) processors. What makes this microprocessing technology so different -- and such a breakthrough? Scalability, flexibility and customization. Power is built on an open architecture, making it an open ecosystem that supports freedom of action in the marketplace.

Power Cycle

[Doc Ruby]

Didn't Apple dump the PowerPC line because its power demand projections in upcoming generations meant more heat (less efficiency) than the Intel x86 competition? Why not just use these chips? Will we actually see a cross-platform Mac strategy, with Apple playing Intel(/AMD) against PPC makers like this, with the same MacOS running on either? Will Apple actually pull off delivering a simple interface for installing software from source, with consistent builds/runs across both Mac platforms? Who knows what to believe anymore?

Re:Power Cycle

[Tavor]

A startup is not going to have near the capacity that IBM or Intel has. Even given the two years that this processor is slated to come out in, I don't see that changing dramatically enough to suit Apple. Just my $0.02 USD.

Re:Power Cycle

[Anonymous Coward]

Will Apple actually pull off delivering a simple interface for installing software from source, with consistent builds/runs across both Mac platforms?

Why?

Re:Power Cycle

[necro81]

I seem to recall another problem that Apple had with IBM, which spurred the switch to Intel, was the speed roadmap on the G5 processors. Not only were the projections on power consumption not coming down, but the horizon for faster and faster clockspeeds did not look good. It has taken a long time for the POWER line, expressed in Apple hardware, to move from 1 GHz to 2 GHz, and it seems to be have been stalled at just below 3 GHz for a while now.

I know that clockspeed is hardly the best measure of a CPU's capabilities, but within the same line of chips it is a decent measure to show how much further you can push the architecture, and how well you are smoothing the kinks in your fab lines. Apple have been able to press on with more powerful computers by going to dual processor, and now dual core dual processor machines. Those machines are real powerhouses though, both in terms of number crunching and electrical demand.

Re:Power Cycle

[Doc Ruby]

Good point. But I don't think Apple would have to be constrained by the startup's capacity. If the tech is good, then getting it fab'ed elsewhere, even by IBM (or Motorola) would be cheaper than switching to Intel. Especially from IBM's POV ;).

Re:Power Cycle

[Doc Ruby]

Because fat binaries are fat. And even if they weren't too fat, generating them requires at least as much finesse as generating builds for one or another platform. Also, "binaries" of any kind aren't completely acceptable on MacOS, at least not as the only option. Building from source is the way Apple (and most other *n?x) is going - fat binaries generated from source is redundant and a waste.

POWER, or PowerPC?

[Onan]

I have a pretty hard time taking seriously the claims of a company that appears to consider the POWER and PowerPC chips interchangeable. Yes, they are related, but they're pretty substanially different beasts--especially when it comes to power consumption. I seem to recall that current POWER units consume over a kilowatt each. Yes, really.

Re:POWER, or PowerPC?

[Thanatopsis]

I think you are confusing the article with the company itself. They are pretty clear with what they are about. Try their about page or check out the company bios. The CEO has designed a chip or two in his time - namely the DEC Alpha processor.

Re:POWER, or PowerPC?

[TheRaven64]

I have a pretty hard time taking seriously the claims of a company that appears to consider the POWER and PowerPC chips interchangeable

You mean like IBM? Since the POWER3 (I think, maybe the POWER4) they have used POWER and PowerPC more as marketing terms (POWER means expensive, PowerPC means cheap) than as an indication of instruction set. The current POWER4/5 chips use the PowerPC instruction set. AIX on these systems will trap the few POWER instructions that are not part of PowerPC and emulate them.

Re:POWER, or PowerPC?

[be-fan]

Holy crap. That bio page is impressive. Sounds like the upper-brass is chock-full of x86-64 and Alpha guys!

Re:POWER, or PowerPC?

[Anonymous Coward]

In addition to their experience, they also all look like absolute studs!

http://www.pasemi.com/about/team.html

Re:POWER, or PowerPC?

[raftpeople]

PowerPC instruction set is a subset of Power
AIX runs on Power, not PowerPC
How can AIX trap an instruction that is already supported by it's processor?

Re:POWER, or PowerPC?

[TheRaven64]

PowerPC instruction set is a subset of Power

Neither PowerPC nor POWER is a subset of the other (check the IBM docs). They do, however, have a common subset. Most POWER code only uses this subset, and GCC can be configured to emit code which only uses this subset. This code will run without modification on both POWER and PowerPC chips. The later POWER-branded chips (POWER4 and later) actually use the PowerPC instruction set, not the POWER instruction set.

AIX runs on Power, not PowerPC

AIX runs on both POWER and PowerPC branded CPUs. IBM even sells PowerPC branded systems that run AIX. The execution core of the PowerPC 970 (G5) and the POWER4 is the same - the only difference is the cache logic and size, and the presence of the vector unit on the 970 (and the fact that the POWER4 series chips were dual-core before the 970 was released).

How can AIX trap an instruction that is already supported by it's processor?

Some legacy code designed for POWER3 and earlier uses the POWER instruction set, not the PowerPC instruction set used by later POWER CPUs. Most of the time, this is not important since the majority (or even the entirety) of the code will contain instructions in the subset supported by both. On the occasions when the CPU encounters a POWER instruction not present in PowerPC, it will call AIX's illegal instruction trap handler and replace the missing instruction with a small sequence that will have the same effect.

5 watts?

[Egregius]

This article says it's 13 watts, with 25 watts at peak. A little early start on the number juggling eh?

Re:5 watts?

[Anonymous Coward]

If you read the white papers on the website, the five watt number is without using the gigabit ethernet while thirteen watts would be using them in many applications.

Re:I call BS

[Anonymous Coward]

677874.

PR from start-ups...

[Gruneun]

will feature... will consume 25 watts

versus

far less than the single-core Power chips... on the market today

You be the judge.

Amazing?

[jmichaelg]

What's so amazing about vapor? The chips don't exist. From the article:

The PWRficient actually won't come out for two years, so it's hard to predict exactly how it will stack up against the competition.

Wait until they deliver the goods before you believe the hype.

Re:Amazing?

[hattig]

Hmm, what is all this "2 years" crap everyone is writing? The QUAD CORE won't be out for two years, the SINGLE core won't be out for 18 months, but the DUAL CORE will be sampling in Q3 2006, that's what all the articles I've read today have said very clearly. Even if it was sampling for 6 months, it would only be 15 months away. Not 2 years.

In addition, this company is headed by extremely skilled guys. This isn't Transmeta in any way. However it won't ever be seen in consumer computers, but maybe other consumer devices will include it inside.

But yes, we should wait for them to deliver before passing verdict.

this is an easy game!

[Thud457]

801-765-4999

Power Architecture == PowerPC

[Wesley Felter]

The old POWER instruction set is dead; no one uses it any more. These days Power Architecture is PowerPC. And this new processor does have AltiVec. Pinouts are irrelevant since they were never standardized in the first place.

Hmmmm, paper paper release?

[loose_cannon_gamer]

Don't get me wrong, I think if they can get it to work, it is fantastic, but... I could release right now my plans to develop a 64-core 2 THz processor available in 2 years too. Fine idea, but this sounds more to me like a small company looking for publicity to raise investment money rather than something I'll plan to see on the established timeline.

And since Intel has redefined its roadmaps in terms of performance / watt, I have to say in terms of brain trusts and sheer man power, if Intel can't do it, a little startup probably can't either, or if they do, by the time they get there, Intel will have been there and done that.

But I'll be rooting for the little guy. I favor good competition in the workplace, as long as they're not competing against me. :)

Did anyone else read that as poker chips?

[jonfelder]

I kept read that as poker chips, and couldn't figure out for the life of me what technology was in IBM servers that would be utilized in a poker chip.

Re:Did anyone else read that as poker chips?

[sconeu]

Stackable servers? :-P

Re:Did anyone else read that as poker chips?

[The Standard Deviant]

Help

Apple doesn't matter

[DreadSpoon]

They were a very, very small consumer in the PPC world. For every Mac that shipped, many many many embedded devices using PPC shipped in turn.

Why do you think Apple has such a hard time finding PPC manufacturers willing to keep up with their demands? There just isn't all that big of a profit to be had by supplying Apple compared to the embedded space.

(F)MAC

[joib]

Actually, many "normal" architectures have FP MAC units instead of MUL and ADD (at least IA64, PA-RISC, x86 being the major one without it). Many algorithms used in FP calculations can be efficiently implemented in terms of FMAC operations (matrix algebra, FFT etc.). Also, it is more accurate since it avoids the double rounding associated with a separate ADD and MUL, and thus also many "intrinsic" operations such as division and sqrt can be more efficiently implemented.

Feynman strikes again

[Anonymous Coward]

I just read an interesting Richard Feynman essay on this very topic. He claimed with better design, we could use much lower power consumption.

The little bastard was right. And he has prior art, so his children will be suing.

these PWRefficient chips will also be low cost...

[Anonymous Coward]

because the company has hired a mRk3ting department made up of minimum wage earning t33NaGerz! Hooray

Hmm....

[mattnuzum]

It works on the breadboard just fine.

Re:Hmm....

[mattnuzum]

I guess that's because on the breadboard the LEDs are the correct direction. In the schematic they're backwards. Ooops. :-]

Re:Hmm....

[tzanger]

Not with the LED cathodes connected to +5, it isn't.

Impressive?

[jiushao]

So, we get dual-core 2.0 GHz and 25 watts in two years? Without any more information this is far from impressive. Intel will have Yonah out in volume early 2006, which is dual-core, expected to clock to well over 2 GHz and with fairly low maximum power requirements (the current rumor is that the 2 GHz version will be in the ballpark of 30 watts TDP). In another two years this POWER chip has better offer some pretty kicking IPC or it'll be fairly uninteresting.

Re:Impressive?

[vga_init]

the current rumor is that the 2 GHz version will be in the ballpark of 30 watts TDP

So basically you're saying that this currently existing chip pales in comparison to your chip that doesn't exist yet and consumes more power.

Re:Impressive?

[Anonymous Coward]

Yonah doesn't include the memory controller, dma controller, crossbar, uarts, 4 gige, 2 10gige, and offload engines. I doubt the power figures as well, but one never knows until they get silicon.

We don't even know if this pwrefficient thing is out-of-order or not. If it isn't it'll be even less impressive than the cell PPE.

Re:Impressive?

[GotenXiao]

I doubt Intel will be able to get below 60W for a single core CPU, let alone a dual-core 2GHz model. Just look at Paxville. ...Then laugh and laugh and laugh.

Re:Impressive?

[jiushao]

"This existing chip" if you happen to live in late 2007. If this Power chip happens to actually match the Pentium M in IPC (which is far from easy) they will still have to count on Intel being unable to squeeze out another 5W out of the Pentium M for a full 20 months.

Very Funny!

[Duhavid]

http://www.google.com/search?hl=en&q=+801-765-4999 &btnG=Google+Search

Re:Apple - I agree

[olddotter]

Apples descision could have been based on a Jobs saying "They will be out when?".

In 2 years Apple might put them in dual core iPods or something. You know the pocket sized holo-suite projector model!

PS. As I sit here ripping DVD's to iPod video format, I have to say Apple NEEDS to care about speed. A mac mini is not a video editing box!

Sony PSP2?

[Tumbleweed]

Pair this with a 720x480 (24-bit) OLED display, with a 1 or 1.8" HD (whatever the size is), and you'd have a pretty killer base for a PSP2.

Here, let me help you:

[leoxx]

1) design a low-power-consumption high-performance PowerPC chip that would be ideal for Apple to use
2) keep the development so secret that spouses are kept in the dark
3) launch the product after Apple has already abandoned PowerPC
4) Ignore Apple because they are irrelevant. Instead, sell stuff to the many companies who consume more PPC chips than Apple ever could now or in the forseeable future.
5) PROFIT!

Intel processing cores?

[phorm]

Wouldn't those intel chips be single-core as well? These chips might give them a run for their money.

PWRgads!

[fanblade]

The company's first so-called PWRficient chip...

"PWR" is thPWR nPWRw "e".

If they can give me 1,000 of them

[torpor]

for a nice price, at batches at a time, it might be interesting.

but its not interesting until the foundry actually has SKU's and can ship, and still survive on such order/delivery demand. too many CPU upstarts forget this rule and fail, countless times, their developers.

Oh it does not either.

[way2trivial]

http://en.wikipedia.org/wiki/Pentium_D
in part

After a week of confusion following the processor's launch, Intel officially denied a report in Computerworld Today Australia that the Pentium D includes "secret" digital rights management features in hardware that could be utilized by Microsoft Windows and other operating systems, but was not publicly disclosed. While it admitted that there were some DRM technologies in the 945 and 955 series of chipsets, it stated that the extent of the technologies was exaggerated, and that the technologies in question had been present in Intel's chipsets since the 875P.

Re:Oh it does not either.

[plasmacutter]

to tell you the truth, i simply don't believe them.
their own rep was the one that went on at length about "drm looking into the future". As such i just don't believe the news orgs pulled it out of their bums..

In many nations they're not required to label their products with what DRM is present within, and they get people to adopt less capable products by furtively sneaking in drm which can be activated/used in the future. I think they are trying to do that now, but to each his own i suppose.

Re:Dual Core Apple PowerBook

[Your Average Joe]

When/If they ever make one I am all over it. Dual core G5 might be 4 float engines! I do not want a good looking x86 32 bit Mac, dual core x64 maybe...

Re:Hmmmm, Intel the ONLY cpu innovator...

[Your Average Joe]

All I can say is do you know who DEC was? Do you know what the Alpha was? Do you know how many years DEC with the Alpha had the highest clock rates and SpecFP numbers?

DEC kicked ass. They sold on the order of 10,000% fewer CPUs when compared to the volume crap that Intel was pushing. Commodity hardware is just that, cheap shit. The kid with the Honda civic will never run the 24 hours of Le Mans while a stock Viper or Z06 can.

A much better description of the architecture

[NoNickNameForMe]

is available on Real World Technolgies http://www.realworldtech.com/page.cfm?ArticleID=RW T102405055354

SlashApple backlash?

[wannabgeek]

I'm trying to understand why there is to be so much reluctance on part of /. crowd to accept this. Every fifth post seems to be calling this vaporware, or saying the business is flawed or some such negative comment, instead of any discussion about the technology/feasibility. Contrast this with the announcement from Transmeta and the enthusiastic welcome that /. crowd gave it. Makes me wonder if this is all sour grapes, as Apple already committed to move to x86, and this seems to make Power a decent option for laptops again.

I know, I would most probably be modded a troll.

Interestingly enough, another announcement...

[Decker-Mage]

This is especially interesting in light of an announcement that I came across on PhysOrg.com: NEC DEVELOPS SPEECH-TO-SPEECH TRANSLATION SOFTWARE FOR MOBILE PHONES, October 24 NEC Corporation today announced that it has succeeded in the development of Japanese-English/English-Japanese, automatic speech translation software for single-chip multi-core processors for small devices such as mobile phones, capable of operation at high speeds with low power consumption. Full story at http://www.physorg.com/news7513.html.

Curious coincidence, neh? An application that needs a low power, multicore design and here somebody steps up to the plate and says "sure, we can do that."

Naw. Here's how this game is played.

[Anonymous Coward]

After reading all the comments here, all I can say is that we've got an amazing amount of noobs posting about this. The embedded space is irrelevant. Apple is irrelevant. Everything else is irrelevant except for one thing. Namely, the goal here is to slap something together and sucker someone else into buying it.

This whole Slashdot piece is just a part of the pre-marketing buzz in order to do this.

I've worked with a lot of startups lately, and all of them could care less about building a decent product. Everything is a slap-dash piece of crap. They can get away with this because they won't have to maintain it; the next sucker will. All they want to do is the least amount of work possible in order to sell it as faster as possible.

Or in otherwords, this is a con-job. And from the history involved here, this is no exception.

Take a look at the background of these guys. While the Alpha/Dev/Vax stuff is respectable, that was in ages gone by. Most recently, look at all the SIBYTE work. This is what is applicable.

The Sibyte stuff was hastily slapped together crap, which got bought up by Broadcom, who is consequently in the process of dropping the Linux related support of it as quickly as possible. Which is a good thing, as the Linux-related software is sheer crap.

For example, look at their gdb support. It was done in an incredibly bad fashion, completely non-standard; which is why none of that crap was ever accepted by the GNU folks.

Then there's the C compiler, which is FULL of bugs.

Oh - and there's some sort of kickbacks going on with the only JTAG support, which is done by Corelis. Don't expect to be able to use any decent JTAG hardware; you're out of luck. And stuck with Windows.

I could go on and on; suffice it to say that the software support is done by people who really don't know what they are doing. The hardware people have a little more of a clue; but they release buggy chips.

In short, I'd avoid any Sibyte chips. And I'd avoid any chips these clowns are trying to push. Anyone betting on them ought to know that it will be the standard slapdash crap that the V.C. community is into turning out these days; and it won't be a reliable product.

Sorry for the rant, but I've already been burned by these clowns once. When they actually release a well-designed quality product, let me know.

One eye watching at all times

[PurpleWizard]

If they watched and chose to change once it would be daft to then swear any such change as being for all time. If they have changed once for advantage they will probably have cause to do so again (given time).

Therefore your reasoning appears flawed.

The fact is history repeats wearing different hats. Which (continuing deviating aka. mind wander ) makes me wonder if we are stuck in a control loop and only capable of minor course change to remain stable. Anything to big and we are doomed!!!