The CPU Redefined: AMD Torrenze and Intel CSI

janp writes "In the near future the Central Processing Unit (CPU) will not be as central anymore. AMD has announced the Torrenza platform that revives the concept of co-processors. Intel is also taking steps in this direction with the announcement of the CSI. With these technologies in the future we can put special chips (GPU's, APU's, etc. etc.) directly on the motherboard in a special socket. Hardware.Info has published a clear introduction to AMD Torrenza and Intel CSI and sneak peaks into the future of processors."

Comments

huh?

[mastershake_phd]

Werent the first co-processors FPUs. Arent they now integrated into the CPU? By having all these thing sin one chip they will have much lower latency with communicating between themselves. I think all in one multi-core chips is the future if you ask me.

Re:huh?

[Chrisq]

I think it has to do with the number of configuration options. Even if technology was able to fabricate one super chip with the best possible GPU and sound processor might be great for some people, but others would be better off with extra general purpose cores, cache, etc. The flexibility of "mix and match" probably outweigh the advantages of having the separate components on a single chip

Re:huh?

[mastershake_phd]

Some people would like to be able to customize (especially sound cards), but mass producing a "super chip" would be more cost effective. You could of course have different versions of "super chips".

Re:huh?

[MrFlibbs]

The CPUs will still be multi-core. They will also integrate as many features as makes sense. However, there are limits on how big the die can be and remain feasible for high volume manufacturing. Using an external co-processor is both more flexible and more powerful.

The interesting thing about this whole co-processor approach is that the same interface used to connect multiple CPUs to each other is being opened up for other processing devices. This makes it possible to mix and match cores as desired. For example, you could build a mesh of multi-core CPUs in a more "normal" configuration, or you could mate each CPU with a DSP-like number cruncher and make a special purpose "supercomputer". It will interesting to see what types of compute beasts will emerge from this.

Re:huh?

[Mr2cents]

Adapting another quote: "If you want to create a better computer, you'll you'll end up with an Amiga". It's more or less what they're describing here. Amiga made heavy use of coprocessors back in the days. It could do some quite heavy stuff (well, at the time), while the CPU usage stayed below 10%.

One cool thing I discovered while I was learning to program was that you could make one of the coprocessors interrupt when the electon beam of the monitor was at a certain position. Pretty nifty.

BTW, for those who are too young/old to remember, those were the days of dos, and friends of mine were bragging with their 16 color EGA cards. Amiga had 4096 colors at the time.

Re:huh?

[TheThiefMaster]

It's a cost and feasibility thing. The original FPUs were separate because they were expensive, not everyone needed them, and it was impractical to integrate them into the cpu because it would make the die too large and result in large numbers of failed chips. They became part of the chip later once the design was refined and scaled down.

The same applies to trying to integrate GPUs into the CPU, at the moment a top-end GPU is too large and expensive to integrate, and not everyone needs one. The move to having a GPU in a CPU socket should cut a lot of cost because the GPU manufacturers won't have to create an add-in-card to go with the GPU, they can just design the chip to plug straight into a standardised socket.

At the same time low-end GPUs are small and cheap enough that they are being integrated into motherboards, integrating a basic GPU into the CPU seems like a good next move, and the major cpu manufacturers seem to agree. IIRC Via's smallest boards integrate a basic cpu, northbridge and gpu into one chip? AMD are definitely planning it with their aptly named "Fusion". *Checks wikipedia* Yeah, Via's is called "CoreFusion".

Still, you are right, all-in-one cpus are the future, we're just not quite there yet.

Re:huh?

[clickclickdrone]

you could make one of the coprocessors interrupt when the electon beam of the monitor was at a certain position

The Atari 800 could do that easily at the scan line level with Display Line Interrupts and somewhat harder with cycle counting at points across the line. And that was 1978 technology..

Re:huh?

[cHALiTO]

I agree the Amiga was a great piece of hardware, but the palette was 4096 colors, It could actually use 32 of them simultaneously on screen (at least the amiga 500, the amiga 2000 could go up to 4096 colors on-screen in HAM6).
EGA displays could use 16 out of 64 if I'm not mistaken.

Ahh those were the days :)

Re:huh?

[walt-sjc]

Ahh - the Amiga. My favorite machine during that era. I got my A1000 the first day it was available. Modern OS's could still learn a lot from that 20 year old OS. Why oh why are we still using "DOS Compatible" hardware????

Amiga had 4096 colors at the time.

Better put "4096" with a "*" qualifier. You couldn't assign each pixel an exact color - the scheme got you more colors by being able to set a bit that said that the next pixel modifies the previous pixel by "x". In this way, they could get more colors using less memory than traditional X bits per color per pixel schemes (Amiga was a bitplane architecture.)

Anyway, back on topic, I wish that the CPU manufacturers could finally come up with a "generational" standard socket. A well-designed module socket should last as long as an expansion slot standard (ISA,PCI,PCIe) and not change for damn near every model of chip. I should be able to go out and get a one, 2, 4, 8 socket motherboard, and stick any CPU / GPU / DSP module into it I want. Can we please finally shitcan the 1980's motherboard designs?

Re:huh?

[Jeff DeMaagd]

I think all-in-one / system-on-a-chip have been around for a long time, but they just weren't popular because that meant a significant performance hit. They may become more common as the performance becomes "good enough" for most common tasks where a desktop or notebook computer would be unnecessary and overpowered. It hasn't been a very popular idea yet, I think in part because the cost difference wasn't much. The next mainstream computer platform just might be a phone though, I understand that a lot of people in SE Asia have been doing this for several years already.

Re:huh?

[dosquatch]

You could of course have different versions of "super chips".

Thereby decreasing their cost effectiveness. 'Tis a viscious circle.

Re:huh?

[mikael]

Werent the first co-processors FPUs. Arent they now integrated into the CPU?

The Intel 8086 had the Intel 8087
A whole collection of Intel FPU's is at Intel FPU's

TI's TMS34020 (a programmable 2D rasterisation chip), had the TMS34082 coprocessor (capable of vector/matrix operations)
(Some pictures here. Up to four coprocessors could be used.

Now, both of these form the basis of a current day CPU and GPU (vertex/geometry/pixel shader units).

Re:huh?

[Tim C]

I think all in one multi-core chips is the future if you ask me.

Great, so now instead of spending a couple of hundred to upgrade just my CPU or just my GPU, I'll need to spend four, five, six hundred to upgrade both at once, along with a "S[ound]PU", physics chip, etc?

Never happen. Corporations aren't going to want to have to spend hundreds of pounds more on machines with built-in high-end stuff they don't want or need. At home, I want loads of RAM, processing power and a strong GPU. At work, I absolutely do not require the GPU - anything that can do 1600x1200 @ 32bpp and 60Hz for 2D is perfectly adequate.

Likewise, the chip builders aren't going to want to have to release these all-in-one chips in a myriad of options, for low/medium/high spec CPU/GPU/PPU/SPU/$fooPU, it simply won't be cost-effective.

It's lose-lose imho; you're either stuck buying things you don't want, or have a mind-boggling number of options to choose from (consumers/business) and support (manufacturers/OEMs/IT depts).

Re:huh?

[*weasel]

However, there are limits on how big the die can be and remain feasible for high volume manufacturing.

The limits aren't such a big deal.
Quad-core processors are already rolling off the lines and user demand for them doesn't really exist.
They could easily throw together a 2xCPU/1xGPU/1xDSP configuration at similar complexity.
And the market would actually care about that chip.

Re:huh?

[metalcoat]

My only question is that will these be needing more heatsinks or fans? Seems like if you spread it out a small heatsink would only be required but I am not an Engineer.

Re:huh?

[badfish99]

On the original IBM PC with a CGA adapter, you had to wait until the vertical flyback interval before updating the video memory. Otherwise the hardware couldn't keep up with sending data to the monitor (or something), and the monitor displayed snow.

Re:huh?

[walt-sjc]

The next mainstream computer platform just might be a phone though

Smartphones will (IMHO) evolve to a wireless portable computing device that "oh yeah, it can make phone calls too," but the problem is still that the screen is still WAY to small, and user input still sucks. Maybe they will finally be able to make LCD-like glasses that really are high-resolution, and maybe they will come up with a neural interface so we can ditch the keyboard / mouse... But I don't see those things being practical within the next 10 years. I also see the CPU speed / memory / storage requirements continuing to increase. We may be able to get everything we will want in a hand-held in my lifetime, but I doubt it.

Re:huh?

[thygrrr]

Nope, the A500 also had 4096 colours in HAM mode. They were basically the same hardware, except the A2000 had different - and more - expansion slots and was a desktop machine while the A500 was a typical home computer/console kind of thingy.

Re:huh?

[tomstdenis]

Um, you'll see noise on any array display if you write to the memory while it is drawing. It may not always show up as noise, it could show up as unsynced portions of the display (re: looking sucktastic).

Tom

Re:huh?

[Anonymous Coward]

Docking stations may then come back.

They were rather silly with laptops since they have large screen and keyboards, but if a smartphone could become as good as a standard desktop of today, then having a dock with a screen and keyboard attached would be an ideal model, since one of my biggest complaints with multiple computers is knowing where files are.

Re:huh?

[Fordiman]

But think. There is definitely money in non-upgradable computers - especially in the office desktop market. The cheaper the all-in-one solution, the more often the customer will upgrade the whole shebang.

Example: in my workplace, we have nice-ass Dells which do almost nothing and store all their data on a massive SAN. They're 2.6GHz beasts with a gig of ram, a 160G HD, and a SWEET ATI vid card each. Now, while I personally make use of it all proper-like, most people here could get along with a 1GHZ/512MRAM/16GHD/Onboard video system.

I think Intel/AMD stands to make a lot of money if they were to build an all-in-one-chip computer, ie: CPU, RAM, Video, Sound, Network, and a generous flash drive on a single chip.

Re:huh?

[Anonymous Coward]

The x87 FPU architecture is kind of interesting. I don't know what Intel was thinking, exactly, I can see 3 possibilities: a) 16bit FPs are kind of useless so make it an extra b) FPs aren't used that often, even on current chips there isn't a ton of FP operations in most cases so make it extra, c) I want to have 2 classes of users, cheap mass users and high dollar specialty users and this costly extra will differentiate them.

When Intel finally did integrate the x87, they still had "sx" and "dx" parts, basically the same part at different price points with different artificial limits. Things change of course but now we have x87 FP, MMX, SSE, SSE2, SSE3 instruction sets, all of which have very limited usefulness in general so I'm inclined to think Intel doesn't mind useless stuff so much as they want two price points, Core and Xeon (and itanium, unless its dead.)

It's not a bad idea to have coprocessors but it's not a great one. You have to code for them, basically anything you code to is going to get much more limited use. Basically I see only 2 markets right now for them, graphics and physics, all the DSP type stuff can be done much faster and more cheaply with SSE like instructions. I kind of see graphics petering out in the next 3 years, 3D will reach some state where everybody is more or less satisfied and the quality is more or less the same (similar to 2D acceleration) and at that point, just integrating it to the CPU makes a lot of sense. Physics engines on the other hand are a much more limited and specialized area. Maybe some kind of pro-grade audio component and maybe some sort of specialized video decompression component but it's still hard to see it working. Seems to me like another tool to price parts at different price points.

Re: huh?

[Dolda2000]

Still, you are right, all-in-one cpus are the future, we're just not quite there yet.

Actually, no thank you. I've had enough problems ever since they started to integrate more and more peripherals on the motherboard. I'd be troubled if I'd have to choose between either a VMX-less, DDR3-capable chip with the GPU I wanted, a VMX- and DDR3-capable chip with a bad GPU, a VMX-capable but DDR2 chip with a good GPU, or a chip that has all three but an IO-APIC that isn't supported by Linux, or a chip that I could actually use but costs $500.

Instead of gaining those last 10% of performance, I'd prefer a modular architecture, thank you. Whatever is so terribly wrong with PCI-Express anyway?

Re:huh?

[TheThiefMaster]

The way I see it is that the hybrid cpu+gpu chips would be about the same size and thermal output as a modern dual-core chip, and the gpu-in-a-cpu-socket would be about the same thermal power as a normal cpu for that socket, so would take the same heatsink.

So one is replacing one core of a dual-core cpu with a gpu and the other is replacing one cpu of a dual-cpu machine with a bigger gpu, with little change in power or cooling requirements in either case.

Re:huh?

[sjwaste]

Intel's quad cores are, and they're actually two Core2 dies connected together I believe. "Native" quad core is in the works by AMD and Intel, but is currently not on the consumer market.

Now, if there are other CPU's out there doing native quad core for general purpose computing, I'm unaware and withdraw my ignorance if so :)

Re:huh?

[evilviper]

"If you want to create a better computer, you'll you'll end up with an Amiga". It's more or less what they're describing here.

That's what he's describing, but I don't believe for a second that's what it's going to be...

I don't believe for a second practically ANYONE is going to buy an expensive, multi-socket motherboard, just so they can have higher-speed access to their soundcard... Ditto for a "physics" unit.

This exists solely because CPUs are terrible at the same kinds of calculations ASICs/FPGAs are incredible at. That will be the only killer app here.

Video cards are a good example on their own. CPUs are so bad, and GPUs are so good, that transferring huge amounts of raw data over a slow bus (AGP/PCIe) still puts you far ahead of trying to get the CPU to process it directly. And it works so well, the video card companies are making it easier to write programs to run on the GPU.

And GPUs aren't remotely the only case of this. MPEG capture/compression cards, Crypto cards, etc. have been popular for a very long time, because ASICs are extremely fast with those operations, which are extremely slow on CPUs.

The situation is much more like x87 math co-processors of years past, than it is like the Amiga, with independent processors for everything.

It is likely that, in time, integrating a popular subset of ASIC functions into the CPU will become practical, and then our high-end video cards will be simple $10 boards, just grabbing the already-processed data sent by the chip, and outputting it to whatever display.

Then maybe AMD and Intel will finally focus on the problem of interrupts...

Re:huh?

[Ngarrang]

What is new is old, and what is old is new. For a while, the mainframe was declared extinct. Now, the mainframe is running Linux and is all the rage again. Same for boards. What is integrated becomes separate, becomes integrated again.

For low-end computers, a board that integrates the CPU, GPU, ATA, etc, makes sense for that segment. But, there is a market segment that wants to be able to upgrade their boards more easily. I would very much prefer a board with a separate GPU socket, if I knew that it was a standard being followed by the dominant video makers. Upgrades would be easier. Power and cooling planning would be simpler.

My vision of the perfect PC has very little integrated onto the board. Give me a board with 8 to 12 PCIe slots. Let ME determine what USB and Firewire will be installed. Let ME determine if my system will have a DB9 serial port or not. Let ME choose SATA or SAS. I do not like having to pay for things I do not need and will not use. For that matter, put my CPU and RAM on a PCIe card, as well. Then, I am not tied to AMD's or Intel's ever-changing socket plans.

Re:huh?

[PrescriptionWarning]

so long as each of those things can be replaced easily, or if the all-in-one as a whole is itself fairly cheap and easy to replace, then it could be a good idea. The only thing I'd have a problem with is the case where the whole bundle costs 500 bucks or more, and when a single item inside breaks a year later, you have to pay 500 bucks again.

Re:huh?

[Ngarrang]

Why oh why are we still using "DOS Compatible" hardware???? Because DOS is still commonly used, far more than most would want to believe. Many boot diskettes are still DOS. A lot of manufacturing equipment still uses DOS. In a world where it seems like all computers are multi-GHz, there are embedded computers in the machines that run our world that are not.

Re:huh?

[sconeu]

Intel also had the 8089, which was a coprocessor for I/O. It's described (along with the 8087) in my vintage July 1981 8086 manual.

Re:huh?

[CastrTroy]

The problem is that there's a lot of other stuff that has to understand the CPU for the computer to work properly. You'd need to be able to snap in different chipsets so that the CPU could actually work. You'd probably want to be able to plug in new ram too. As processors get faster and faster, they require more pins. No more 80 pin CPUS for us. If they designed a universal CPU slot today, it would either have twice as many pins as we needed, or we'd run out of pins within 2 years.

Re:huh?

[Archangel Michael]

"most people here could get along with a 1GHZ/512MRAM/16GHD/Onboard video system."

Haven't tried to run Vista yet ... have you.

Re:huh?

[tmach]

Sounds like we're going BACK to 1980's motherboard designs. Heck, before the Amiga had this setup there was the good ol' C=64. Can we name the other chips SID and VIC just for old times' sake? Now we just need to put the OS on a flashable chip with its own dedicated swap RAM and we'd really be back to the days of "instant on". That'd be sweet!

Re:huh?

[drinkypoo]

Haven't tried to run Vista yet ... have you.

No, and neither has anyone else, so his point still stands.

Re:huh?

[rbanffy]

OK... Let's rephrase that:

Folks with 16-bit PCs were bragging about their 16 out of 64 color EGA cards and single-tasking OSs when even the simplest the Amigas had 32-bit processors, 32 out of 4096 colors, PCM audio and a fully multi-tasking OS coupled with a GUI.

As for the "processor socket", there are people selling computers that go into passive backplanes. If you put the CPU and memory in a card, there is little reason why you would have to upgrade the rest of the computer when you change the CPU (you would have to scrap the card, anyway, but processors are intimately related to chipsets, so, it is to be expected.

Thete are some SoC (system on a chip) solutions out there too. Those incorporate the chipset (or most of it) into the CPU, so, it would be easier to build a trans-generational socket

Re:huh?

[WorseThanNormal]

But by making specialized chips, you can limit and optimize the instruction set to allow for many more instructions per second. The performance gains of this strategy (as well as using this a a means of heat distribution) could out strip the latency gains of putting everything on one chip.

Re:huh?

[Anonymous Coward]

"Never happen. Corporations aren't going to want to have to spend hundreds of pounds more on machines with built-in high-end stuff they don't want or need. At home, I want loads of RAM, processing power and a strong GPU. At work, I absolutely do not require the GPU - anything that can do 1600x1200 @ 32bpp and 60Hz for 2D is perfectly adequate." ...and thiat machine is most likely to be the one running AT THE MOMENT on an integrated gfx platform. Essentially, once the the whole northbridge is integrated onto the CPU die mobo complexity collapses, only one cooling system is needed - everything shrinks. This is what commoditisation is all about, and it's exactly what software vendors like MS LOVE. SoCs like this are also suiperbly well suited to mobile computers, which are both already dominant and continuing to displace desktops.

Re:huh?

[Yvan256]

I like the idea of "CPU slots". Not in the Celeron 300A-style "CPU on a card" concept, but to put everything related on a card, make the motherboard "braindead".

Put the CPU, chipset and RAM slots on the "processor card", that way the only reason to upgrade a motherboard would be if new slots were introduced (PCI-Express, etc) and you actually needed them.

Isn't this called "passive backplane" or something? If it already exists for some systems, why not desktop computers?

Re: huh?

[Panzergheist]

Instead of gaining those last 10% of performance, I'd prefer a modular architecture, thank you. Whatever is so terribly wrong with PCI-Express anyway?

Three words: Small Form Factor.

Re:huh?

[walt-sjc]

As processors get faster and faster, they require more pins. No more 80 pin CPUS for us. If they designed a universal CPU slot today, it would either have twice as many pins as we needed, or we'd run out of pins within 2 years.

That assumes you are plugging in individual processors. This is why I specified "module". Using high speed serial channels, like PCI Express does, you can reduce the pin count on the module level considerably. I would assume that the module would also contain a certain amount of local RAM, such as 1 - 2G. The pin count on CPU's today is also an artifact of legacy design - wider parallel buses, integrating more "support" into the processor, etc.

In order to really advance computing, we have to think differently.

Re:huh?

[flaming-opus]

small parts of the market might be interested, but only the parts that are now served by GPUs integrated into chipsets. The advantage of discrete graphics chips, whether plugged into a cpu socket or into an expansion slot, is upgradability, and the chance to differentiate products. Putting everything on one die lowers total system costs, and circuitboard complexity, but is a hard sell in many segments of the PC market.

My question is what advantage will they get from plugging these coprocessors into CPU sockets, as opposed to PCIe risers? Does the coprocessor realy need to be cache coherent? If it is, how do you deal with interrupt handling? Does the GPU run independant threads that are peers to CPU threads?

Re:huh?

[ChrisA90278]

You are right. The distance and therefor communications time is better if the device is closer. But butting the device inside the CPU means it is NOT as close to something else. One example is the graphic cards. There, you want the GPU to be close to the video RAM,not close to the CPU. Another device is the phone modem (remember those) you want that device close to the phone wire. Now let's look at new types of processors. A Disk I/O processor that makes a database run faster. You would want that to be outside of the CPU. It should belocated between the PCI bus (or other I/O bus and the system RAM. Putting it inside the CPU will just cause more traffic on the CPU's bus.

The reason you might want the processor to be NOT insde the CPU is to keep some data off the CPU's bus. A floating point processor is an example of something you dop want inside the CPU but a RAID chip is best outside the CPU. You need to deside case by case.

Re:huh?

[ioshhdflwuegfh]

you could make one of the coprocessors interrupt when the electon beam of the monitor was at a certain position The Atari 800 could do that easily at the scan line level with Display Line Interrupts and somewhat harder with cycle counting at points across the line. And that was 1978 technology.. And Commodore 64 could do the same thing. Very cool stuff, you could switch the video mode at a particular line of the screen and split screen into text and graphics, etc, but I have never tried nor have seen anybody doing it with the points across the line...

Re:huh?

[donglekey]

1 GPU != 1 Core
The latest Nvidia GPUs actually have 128 cores/shader units, and have many more transistors than the a cpu. Also Intel Quad Cores are actually 2 dual core procs next to each other.

Re:huh?

[vivek7006]

Intel's quad cores are, and they're actually two Core2 dies connected together I believe. "Native" quad core is in the works by AMD and Intel, but is currently not on the consumer market.

At the end of the day, nobody cares whether the cpu is "native" or not. If its cheap, gives good performance, sucks less electricity and fits in "one" socket, its good enough for most folks

Re:huh?

[sjwaste]

I agree, but the post I was responding to was stating that quad core was rolling out despite demand and die space that might reduce yield. My point was only that because its two dies put together, its different than having to put four on one die and having corresponding yield issues.

Re:huh?

[default luser]

Isn't this called "passive backplane" or something? If it already exists for some systems, why not desktop computers?

Early high-end computer systems started out like this, utilizing backplanes like VME. They've been phased-out, because ultimately that modularity was too expensive, and because the shared-bus architecture hurt performance. Hardware devices that used to require multiple cards can now fit on a single chip, and have their own PCIe drop to increase performance. Memory upgrades that used to require multiple cards just to reach 1MB are now eclipsed by 8 and 16-chip configurations on a single DIMM (a specialized expansion slot), and have their own bus to improve performance.

Let's say they went with the Single-board computer design (CPU+memory+bus controller) - now your costs go up, because you have to build multiple "processor cards" for all the different backplanes you want to plug into. ISA backplane - 1 model. PCI backplane - 1 model. PCI + ISA backplane - 1 more model, and it also requires a new specification: the new bus designs have to play nice with the limited I/O space at the back of the card, so you end up either making the bus connector larger, or you end up making certain bus combinations impossible.

With the motherboard and atached bus design, your costs go down because you can provide a mixture of the busses that are the most popular. Thus, you only have one product to design and electrically verify, and only one manufacturing line to test.

Also, when you move to point-to-point architectures like PCI-Express, with a separate backplane you really limit yourself to the slot configurations you can offer. Unlike with a shared bus, with P2P interconnects you have to make sure the backplane layout matches the connector layout exactly. This means you either standardize on ONE configuration (boring), or you put the ports on the processor card (what we are doing).

The only places that still use modular bus designs today are embedded developers, and that's because they still need the expandability and modularity that end-users do not. They also need the backward-compatibility affored by these old bus specifications (VME especially). They pay for it, in terms of performance - most of them bypass the slow backplane of VME or CompactPCI with faster interconnects like Gig/10GigE, Fibre Channel, RapidIO or Infiniband.

Re:huh?

[badfish99]

No, you could update the original monochrome IBM PC display whenever you liked. The snow thing was only a (mis)feature of the colour display adapter.

Re:huh?

[Wavicle]

Anyway, back on topic, I wish that the CPU manufacturers could finally come up with a "generational" standard socket. A well-designed module socket should last as long as an expansion slot standard (ISA,PCI,PCIe) and not change for damn near every model of chip.

Unfortunately this isn't as practical as it sounds. As technologies to increase performance continue, the socket technology would quickly become a bottleneck. If we were to go back 10 years ago (1987) our boards would have 3.3v CPUs, 64 bit memory busses and plain old SDR DIMMs. We'd be seriously hamstrung with how much performance we could squeeze out of it. Hyper Transport and CSI would be entirely impossible with the frontside bus design from back then, and would have been entirely impractical to implement back then. A socket that could have grown over that time would have 100 pins for general housekeeping, 200 legacy pins to support FSB and another 700 pins for the multiple hyper transport interconnects. It would have been prohibitively expensive to make boards back then that supported 1000 pin CPUs. We did not know then how exactly we would get to our performance today. A whole memory technology (RIMM) and an expansion card technology (AGP) came and went during that time.

Re:huh?

[tomstdenis]

I didn't say it would be snow (as in electrical noise).

But think about it, if you are say 0.5 frames out of sync with the display, then only half of the display will actually have "live" content. The result (especially with movement) will look very distorted. So even with modern displays, you need to sync to vsync and draw only during vblank (or draw during the frame but use frame switching).

LCDs are not instantaneous either. Try filming an LCD without the correct shutter. You'll see the same annoying bar floating down the screen as you do with a CRT (albeit usually slower depending on refresh rate I guess). So it isn't as if BOOM the frame appears 100% all at once.

Tom

Re:huh?

[real gumby]

Werent the first co-processors FPUs

Actually there was an evolution of processor design into single, monolithic processing units; until well into the '70s it was hardly uncommon that computers would have all sorts of processing units (remember the "CPU" is the "Central Processing Unit.") Of course in this case I'm primarily talking about mainframes; one of the distinctions of the minicomputer (and later microcomputer) was that "everything was together" in the CPU. But even then the systems didn't really start out monolithic: it was not uncommon to find minicomputers with separate FPUs, writable microcode and the like.

Another element of mainframes which is reappearing is I/O processors (AKA "channel controllers"); most mainframes from the 50s on had programmable I/O processors. The ARPAnet interfaces from the very earliest days were computers in their own right.

Finally, consider that there's more to the interconnect issues than low latency between processing units. For example, if you can load a coprogram into a coprocessor (be it FPU or parallel unit, graphics unit or the like) then it can crunch away and do its own DMA (perhaps to separate banks) without (for example) cache contention or the like. You can get better performance than having all these features on-chip.

By the way the unified philosophy of the microprocessor influenced Unix and the C language, which featured a monolithic kernel (and lots of stuff in userspace) in the former case and "weird" artifacts like reduced syntactical diversity and the like (remember the line in the intro to K&R: " 'You mean you have to call a function to do I/O?' " -- I found this hilarious since I was a Lisp programmer at the time and was well used to this approach). But look back at how people were thinking back then: these were small systems for small (e.g. research, instrument control, etc) applications. Nobody, or hardly anybody anyway, back then was imagining minicomputers replacing "real" computers. In a funny way they were right, since today's microcomputers look a lot like the mainframes of yore.

Re:huh?

[blahplusplus]

"Werent the first co-processors FPUs. Arent they now integrated into the CPU? By having all these thing sin one chip they will have much lower latency with communicating between themselves. I think all in one multi-core chips is the future if you ask me."

You misunderstand displacement, first of all there are only so many chips you can pack into a CPU die before bandwidth and memory issues become a problem, high memory bandwidth devices will spam the communications channel waiting for data. Really is a displacement of data (think water, or energy). I've seen computers go from specialized chips to integrated and back to specialized chips again and so forth. It all depends on the state of the technology of the time and the performance/power requirements.

Re:huh?

[Anonymous Coward]

Then maybe AMD and Intel will finally focus on the problem of interrupts...

Isn't that why Intel introduced Message Signaled Interrupts?
http://en.wikipedia.org/wiki/Message_Signaled_Inte rrupts
http://lwn.net/Articles/44139/

Re:huh?

[LWATCDR]

Been there and done that and it just doesn't work.
Kaypro and Zenith both offered PCs that let you swap out CPU cards. I have a great poster in my office for the Kaypo PC that says "The End of Obsolescence". It wasn't
The problem is that the upgrades tended to cost as much as a new computer.

You have a smaller potental market and the cost for the new CPU board is so close to a new motherboard it just isn't worth it.
Then you add the improvements in memory systems and as you can see it just doesn't work out.

Re:huh?

[carlmenezes]

Coming from a DSP background, I can't imagine a company like Texas Instruments NOT jumping all over this idea as their entry point into the world of the mainstream desktop. They already have tons of experience in designing processors dedicated to audio/image processing and I could so easily see a "codec co-processor" made by them being one of the processors you could add to your uber system for some really amazing decoding/transcoding of multimedia content.

After all, are we not moving towards the era of where multimedia is as vital to your social life as email is now?

Re:huh?

[toddestan]

Haven't tried to run Vista yet ... have you.

I would guess that most of them could also get by just fine on Windows 2000 too.

Re:huh?

[VanessaE]

You're talking about screen jitter - the brief appearance of vertical or horizontal "cuts" in the image when something is moved or redrawn. Even Commodore 64's have to deal with this (unless you feed updates to the screen buffer more than 8 raster lines ahead of the actual raster position). The G(GGG...)P is talking about random character "snow" - like a screenful of random characters appearing briefly over the actual content of the display, most commonly seen when the screen scrolls. My old 286 had such an adapter with the snow problem.

Re:huh?

[VanessaE]

On the C64, you can make changes in screen/sprite/background colors and a few other things midway across a line, but mode changes, re-positioning sprites, etc. will only change at the beginning of the next raster line.

Re:huh?

[PitaBred]

Man, I can't WAIT to get back to the days of unprotected memory, where any program can access any other program's address space, and do all kinds of NEAT hacks!!!! Just don't connect it to the Internet.

The Amiga was a great system for it's time, and had some very good design choices with the co-processors and the like. But it's not feasible in today's malware (and crapware, and buggy-beta-ware, and...) infested world. I think these designs, that are backed by the two central players in the desktop CPU/architecture market (and heavy players in the small to mid-size server market), will be a good compromise between the robustness and speed of recent technologies with older ideas. Not to mention it's actually better having a general-purpose bus, so things are more configurable, rather than just having a system with separate chips for different tasks.

Re:huh?

[Fordiman]

Have no intention of doing so ... ever.

Why bother when you can get all the glitz for much less resource (and cashish) using Linux/Beryl, or all the functionality at the same low resource with Win XP?

Re:huh?

[mrchaotica]

My question is what advantage will they get from plugging these coprocessors into CPU sockets, as opposed to PCIe risers?

The advantage is that it's easier to do "general purpose" stream processing on them, instead of just graphics.

Does the coprocessor realy need to be cache coherent? If it is, how do you deal with interrupt handling? Does the GPU run independant threads that are peers to CPU threads?

Right now, with DirectX 9 parts? No. With Vista and DirectX 10 and CUDA and whatnot? Absolutely!

Re:huh?

[smallfries]

HAM wasn't really about saving memory. The graphics hardware was palettized (new word!) so there was a hard limit on the on-screen colours. Hold And Modify was a nice hack that used the beam-position interrupt in the copper chip to reprogram the palette on the fly.

The reason that we don't have a standard plugin socket for processors is that the chips would become a commodity. As the chip companies design the sockets it is in their interest to maintain incompatible versions. If I could just buy a "standard motherboard" rather than deciding from the ground up whether it should be Intel or AMD then the chip decision would come down to simple cost (or at least cost/performance). Keeping the sockets clashing means that OEMS are tied into a particular company for a particular line - otherwise they have to redesign the machine.

It is a nice pipedream though. Buying a 8x "standard" board and then filling it with a mixture of vector processors, x86 chips and other exotic possibilities.

Re:huh?

[walt-sjc]

That depends on whether you really innovate, or if you just rehash the standard CPU socket design. You are assuming that it would be a rehash of what we have today. I would not suggest that. I also think stuffing a raw CPU into a socket on the motherboard is a bad idea - it should be a module that has supporting circuitry, probably even "local" ram. Think of a design that is closer to PCI Express than PCI-X where you have many very high speed serial interconnects rather than a couple ultra-wide parallel. With a reasonable amount of local memory, more can be done "on module" at even higher speeds, with a lower pin count.

CPU integration with system board design has not changed much since the very first CPU. Not changing anything has the advantage of making it easy to integrate new CPU designs, but the disadvantage of what we have today - constantly changing CPU sockets requiring new motherboards for damn near every chip rev. I'm just saying that maybe there is a better way...

Re:huh?

[Mr2cents]

Unprotected memory was the standard at the time. Processors didn't have MMU's. OS's weren't designed with virtual addressing in mind. It's not like the DOS world had it.

Re:huh?

[neomunk]

Yeah, that's pretty much what I thought when I saw this too, also that it might be the beginning of the end for high-powered single-function console type computers. You know, like professional midi controllers, high end audio processing gear, video editing boards, stuff like that. Looks to me like pretty soon people are going to be able to use the same platform for all of those, with different processor combinations.

Like if you had a platform with, I dunno, say 4 chips per system....
server: 4x general purpose CPUs , at quad cores that's 16 cores of hot server love.
archiving: 1x CPU 3x compression/decompression chip, or 1x CPU 2x codec chips and 1x encryption chip ... I could go on and on like this, it's fun to think about actually.

And imagine the new market for chips! They have produced (or are in the works) things like a game AI chip, a physics simulator chip, all manner of codec chips, encryption chips, graphics, sound, servers of all types (http+scripting, network fileserver, irc, whatever)on a chip, floating point... wow.
Now we'll start seeing things like a Google search chip at Wal-Mart (you can probably get a Google rack unit at Sam's now, come to think of it, but that's beside the point) and possibly even something like game companies putting whole game engines on a chip and slingin them bad boys with a certificate to let you pick out 2 of the tens or hundreds of the games that run on it or something.

One added bonus, to ME at least, but I know there's other out there... Ever since I was a little boy with my VIC-20 I've dreamt of pretty much exactly this. I mean pretty much exactly. This mix and match bag of actual processors is a hacker(original usage)'s wet dream, allowing optimized execution of any ole processing you need done.

So, in short, like the announcer from the game Smash T.V. IIIIIIIIIIII LIKE IT!

Re:huh?

[neomunk]

I really don't know the answer to this, but what about AMD's hypertransport? Is that what you're looking for? From what I can gather it's basically PCI-e but with processor scale latencies and bus speeds.

Re:huh?

[PitaBred]

That's what I'm saying. We DON'T want an Amiga. We can learn some lessons from offloading some processing tasks, but that's it.

Re:huh?

[evilviper]

I was specifically referring to the performance penalties associated with x86 interrupts (as opposed to other platforms). Something MSI doesn't appear to address, AFAIK.

Re:huh?

[Anonymous Coward]

"I don't believe for a second practically ANYONE is going to buy an expensive, multi-socket motherboard, just so they can have higher-speed access to their soundcard..."

Then you know diddly about pro audio.

CSI?

[BigBadBus]

CSI? De-centralized CPU? Where will they be located; Miami, New York or Las Vegas?

Re:CSI?

[Anonymous Coward]

They'll be located in all three locations, of course.

Re:CSI?

[91degrees]

CSI? De-centralized CPU? Where will they be located; Miami, New York or Las Vegas?

Well, clearly, they won't. They're decentralised.

New on NBC, "CSI: Wherever". We even have a song by The Who for the opening credits - "Anyway, Anyhow, Anywhere".

Re:CSI?

[MarkRose]

Sounds like a Crappy, Stupid Idea to me.

Re:CSI?

[BigBadBus]

Maybe its the CSI Exclusion Principle ...

Re:CSI?

[Anonymous Coward]

CSI is on CBS, you insensitive clod!

Re: CSI?

[Dolda2000]

Aside from the jokes, am I the only one who am more than a bit disturbed by Intel's CSI (apparently "Common System Interface")? Did they actually find anything really bad about HyperTransport that TFA fails to mention, or is it just a horrid example of the NIH syndrome?

Re:CSI?

[Anonymous Coward]

I would use the "I live in another country" defence, but I Googled it and knew it was CBS. God knows why I put NBC.

Re:CSI?

[Jonavin]

The new Intel CSI Miami. "It looks like, there is, some sort, of, connection." WWWWWAAAAAAAAAAAA!

[minrant]Stupid David Caruso[/minirant]

Re:CSI?

[VanessaE]

nonononono if you have to mimic Caruso, it should be more like this:

"It looks like, they are connected..." [puts on sunglasses] "...and it's serial."

What goes around.......

[trancemission]

Shrugs at memory of 3 days attempting to install windows 95 on a 386... finally got there after removing the co-processor. I was young........Happy days........

Previous announcements

[G3ckoG33k]

The first details emerged half a year ago:


IBM and Intel Corporation, with support from dozens of other companies, have developed a proposal to enhance PCI Express* technology to address the performance requirements of new usage models, such as visualization and extensible markup language (XML).

The proposal, codenamed "Geneseo," outlines enhancements that will enable faster connectivity between the processor -- the computer's brain -- and application accelerators, and improve the range of design options for hardware developers.


http://www.intel.com/pressroom/archive/releases/20 060927comp_a.htm

Re:Previous announcements

[Anonymous Coward]

Since when is XML a new usage model requiring advances in processor design?

Are they going to redo the X86 instructions in XML? X86 XML ASM? The CTO will luv it!

<asm>
<data>
<db id="msg">
Hello World!
<br />
</db>
<equ id="len">
<value-of select="$msg" />
</equ>
</data>
<text>
<globals>
_start
</ globals>
_start
<mov><ebx>0x01</ebx></mov>
<mov ><ecx><value-of select="$msg" /></ecx></mov>
<mov><edx><value-of select="$len" /></edx></mov>
<mov><eax>0x04</eax></mov>
<int>0 x80</int>
 
<mov><ebx>0x00</ebx></mov>
<mov><eax >0x01</eax></mov>
<int>0x80</int>
</text>
</asm >

Re:Previous announcements

[badfish99]

Since when is XML a new usage model requiring advances in processor design?

Since it became bloatware that is capable of wasting 90% of the processing power of a modern computer.
</sarcasm>

Re:Previous announcements

[Fordiman]

Oh my GOD that needs shot in the face right now! It's a markup language, not a programming language! ...

They should use Javascript instead!

*ducks*

Re:Previous announcements

[Anonymous Coward]

That won't be an issue when you plug-in your XML acceleration chip.

Re:Previous announcements

[BlackSnake112]

The proposal, codenamed "Geneseo," ... Isn't that a town in upstate NY? I think there is a school there to.

Re:Previous announcements

[edflyerssn007]

Yeah, it is about a half our south of Rochester, NY. There is a decent school there which a couple of my friends attend.

-Ed

Boring

[Anonymous Coward]

Give me 8 core CPUs and then strap on DSP for Audio, Graphics, Physics and AI. While you're at it, do something original and innovative that will impress me.

Re:Boring

[andy_t_roo]

well the way games are going you can get the first 2 easly, the 3rd if the programmers put in some effort to create a good engine and the 5th if you get lucky, but we now have 15 years of basically non-original game play that sais that the last 2 probably won't happen (except in the occasional game - one or 2 in the next few years)

Retro-innovation

[Don_dumb]

Here spins the Wheel Of Reincarnation http://www.catb.org/~esr/jargon/html/W/wheel-of-re incarnation.html watch how everything comes back and then goes away again and then comes back . . .

Re:Retro-innovation

[Delifisek]

Amiga on chip ?
after so many new boards, cpu's, operating systems...
I'mean spending hell of money they found Amiga in cpu ?

In corporate america inventions OWNS YOU

Re:Retro-innovation

[Anonymous Coward]

"Here spins the Wheel Of Reincarnation, watch how everything comes back and then goes away again and then comes back . . ."

You mean like the cellphone is just a reincarnation of the wired phone? The thing all you "reincarnationist" forget is that while the same idea is going around, the circumstances and technology around it is changing. In some cases making what was formally unfeasable, possible. At least in this stories case, we don't have to wait decades for someone to develop the means to shrink all of it down to one chip before we can enjoy it. Just like we didn't have to wait for cellphones to become pocketsize before they were availanle to the market.

Yay Amiga!

[Anonymous Coward]

So they're reinventing the Amiga architecture?

http://en.wikipedia.org/wiki/Advanced_Graphics_Arc hitecture

Re:Yay Amiga!

[T.E.D.]

So they're reinventing the Amiga architecture?

Except that the (single-source) chips won't be soldered into the motherboard.

On the Amiga, this caused perversions like the blitter (fast memory copying) chip eventually becomming slower than the CPU at copying memory. There was no way to pop in a new faster video chip or blitter chip. If you wanted a better rig, your only recourse was to head over to West Chester and join everyone else in begging Commodore to design one.

I like this modular (co)processor idea way better.

Amiga had all processors on the main board

[CdXiminez]

Are we finally getting back to actually complete computers like the Amiga?
It had custom designed processors for sound and video on the motherboard.
And then it was sold together with a fitting OS, so you got computer and software as a complete functioning machine in stead of many loose ends in a PC.

Re:Amiga had all processors on the main board

[Goaway]

You want that, get a Mac.

Seriously, I did, and it's feeling just like the old days.

Re:Amiga had all processors on the main board

[Yvan256]

You mean like modern Macs have become? They have a CPU, a GPU, some audio chip (probably not a DSP but still). And the OS knows how to work with both the CPUs and the GPU.

Re:Amiga had all processors on the main board

[clickclickdrone]

Heck, this goes back to the Atari 800 series.
All this is really doing is bringing a more standardised set of co-processers on to the mobo rather than any number of 3rd party ones - it would make it much easier keeping the OS stable if you have a more controlled number of architectures to deal with.
On the downside, if these processors were DRM hobbled, it would make life harder too..

Re:Amiga had all processors on the main board

[CdXiminez]

I did, at home, I'm just a bit frustrated with the PCs at work.

Re:Amiga had all processors on the main board

[CdXiminez]

I did :-)
I moved from Amiga 1200 to iMac in 1999. Never had a PC in the house (except, perhaps, the bridgeboard on the A2000, which, back then, made me wonder what all the fuss of PCs was about).

Re:Amiga had all processors on the main board

[walt-sjc]

The problem is that the Mac's are all built like standard PC's now. If you replace the Apple firmware with a standard BIOS you could boot DOS. The Mac Mini is basically a standard notebook in a different form factor, ditto for the imac. The Mac pro is not much different in design than a Dell desktop. Why? Cost. They get to use standard parts / software. There is NOTHING on the market like the integrated design of the Amiga. The mac has a lot more in common with a 1982 IBM PC than an Amiga.

Re:Amiga had all processors on the main board

[Anonymous Coward]

Or get a 1960-vintage IBM 360.

Re:Amiga had all processors on the main board

[mdwh2]

Or get a PC.

Really, if GPUs and sound chips are sufficient for a comparison to the Amiga's chipset, then PCs have been doing that for at least as long as Macs.

It's not clear to me why this article is about something more Amiga-like than what modern computers already have (especially since GPUs are fully programmable). The difference about this news is that the chips can be put on the motherboard via a standard socket - but it was never the case with the Amiga that you could plug in chips you wanted, you just had the entire chipset hardwired to the motherboard, no different to a chipset on a modern PC.

Re:Amiga had all processors on the main board

[Goaway]

I don't think you understood at all what this particular branch of the discussion is about.

Re:Amiga had all processors on the main board

[mdwh2]

Yes, I did - if not, perhaps you would care to explain to the rest of us rather than playing guessing games...

Re:Amiga had all processors on the main board

[Goaway]

Quoting from a few posts up in the discussion:

And then it was sold together with a fitting OS, so you got computer and software as a complete functioning machine in stead of many loose ends in a PC.

Moore's Law

[Anonymous Coward]

is/was an over-optimistic/egotistical load of crap

Interesting

[Aladrin]

I find the idea of multiple Processing Unit slots on the motherboard that can each take different type of chips to be very interesting. I'm not sure how well it will work, though. The article mentions 5 types that already exist: CPU, GPU, APU, PPU and AIPU. (Okay, the last doesn't exist yet, but company is working on it.) There's only 4 slots on that motherboard that's shown. I definitely do NOT want to see a situation where the common user is considering ripping out his AIPU for a while and using a PPU, then switching back later. I can only imagine the tech support nightmares that will cause.

So the options are to have more slots, or make something I like to call an 'interface card'. See, there'll be these slots on the motherboard that cards fit into... wait, don't we have this already?

And more slots isn't really an option because the computer would end up being massive with all the cooling fans and memory slots. (Which are apparently seperate for each PU.)

I kind of hope I get proven wrong on this one, but I don't think this is such a great idea. Just very interesting. Having 16 slots and being able to say you want 4 AIPUs, an APU, 4 GPUs, 3 PPUs, and 4 CPUs on my gaming rig and 1 GPU, 1 APU, and 14 CPUs on my work rig would be awesome.

Re:Interesting

[eddy]

Maybe if a motherboard featured a very large generic socket to which was attached one cooling solution, it'd work out better. Processing Units, which would be smaller as to fit as many as possible, would be able to go anywhere in this socket (in a grid-aligned fashion). Easiest solution, socket is X*X square grid, and all PUs must be say X/2 (or hopefully X/4) squares which can be arranged in any fashion. Plunk them in, reattach cooling over all of them, boot and enjoy that 4CPU, 2GPU, 2FPU configuration.

Separate sockets with separate cooling, which I assume is what we're about to see [more of], is going to get messy. And loud.

Maybe in the future some day we've have "Tetris Computing" where you have to puzzle to fit the PUs optimally in the socket. "Oh, I'd really like that nVAMD GPU eXTReMe 2010, but it's an L-piece, and I really need an S-piece for 'tetris' in my bottom half of the socket..." :-)

Re:Interesting

[Overzeetop]

You are correct - sockets are just a reincarnation of slots, but less flexible because you're limited to what you can put on a single chip instead of an entire card.

Perhaps the better thing to do would be better slot designs (not that we need more with all the PCI flavors floating around right now) with integrated, defined cooling channels. If you were to make the card spec with a box design rather than a flat card, you could have a non-connector end mate with a cooling trunk and use a squirrel cage (higher volume, quieter, more efficient)fan to ventilate the cards.

Re:Interesting

[walt-sjc]

So basically have a passive backplane like industrial computers have been doing for YEARS, except that you allow multiple CPU boards. I like it.

Re:Interesting

[Joe The Dragon]

look up HTX slots

Re:Interesting

[drinkypoo]

Perhaps the better thing to do would be better slot designs (not that we need more with all the PCI flavors floating around right now) with integrated, defined cooling channels.

Adding a connector means you will have more noise. Using chips means they both have a shorter path and are electrically better connected.

Most solutions need only two things; a processor and memory. Everything else you see on the card is either there for I/O (video cards have RAMDACs for example, or whatever the chip that handles digital video is called) or there to interface to the bus. If you make the bus interface work directly with a chip, then you don't need most of that shit anyway.

Re:Interesting

[Overzeetop]

I think there are disturbingly few consumer or (volume) business applications which require additional process-specific (FPU, GPU) processors which don't interact with the ourside world. Heck, that's what makes computers great - connectivity. Just about any arguement that can be made for additional PU sockets could be made better by including them on a single die, save possibly thermal concerns, in which case you can add multiple smaller processors.

In my opinion, what the slots lose in path length and electrical noise they gain in flexibility. Right now, sockets are terribly limited in their range. I've got a server with a fairly common motherboard (okay, it's an OEM, but its basis is common), and the procesor slot has a whopping 50% speed increase from the slowest to fastest processor it can take. It's three years old and the fastest processor it can handle (a prescott P4-3.4) can barely get a sniff at group of processors hogging the benchmark charts.

No, unless you're just going to put extra processor slots in so we can add SMP chips as the computer ages, I'll pass.

Re:Interesting

[drinkypoo]

No, unless you're just going to put extra processor slots in so we can add SMP chips as the computer ages, I'll pass.

If all the chips use HT to speak to one another, and all the chips use the same package, then you can put EITHER a CPU or another type of processor in it.

Looks like CSI is continuation of Intel CSA bus

[timecop]

Intel introduced something called 'CSA' bus (http://www.intel.com/design/network/events/idf/cs a.htm), which was higher bandwidth than PCI and was to be used for "streaming" devices like NICs and such. Making this 'general purpose' and user accessible was the next logical step. Go intel!

Amiga?

[myspys]

Am I the only one who thought "oh, they're reinventing the Amiga" while reading the summary?

Re:Amiga?

[Anonymous Coward]

No, I thought it first and I already filed a patent on it.

Re:Amiga?

[DingerX]

Yeah, you were.

The diehard Amiga fans were thinking, "This would really work well if the bus ran faster than any of the cores."

Re:Amiga?

[Easy2RememberNick]

I was thinking about the Amiga too.

Re:Amiga?

[oh_my_080980980]

YES!!!

It's been done before and with great success. To bad it took 23 years for the rest of the industry to catch up!!

Re:Amiga?

[drinkypoo]

Am I incorrect in believing that the original Amigas' buses ran at the processor clock rate? At least until PPC accelerators came out anyway?

Re:Amiga?

[3t3rn4l]

Nope--First thing that came to my mind also!

Thats great new's!

[dave420]

Is there an address we can send money to get /. editors a basic grammar textbook? I'm no pro, but that's just ridiculous.

Re:Thats great new's!

[Goaway]

Slashdot "editors" do not "edit" the submissions. According to CmdrTaco, this makes Slashdot "more real".

Slashdot could benefit from a co-processor...

[Mad_Rain]

that revives the concept op co-processors.

Slashdot's computers might benefit from a co-processor, the function of which is to monitor and correct spelling and grammar errors. It would serve like an editor's job, only better, because, you know, it might actually work.

(Bye-bye karma!)

Re:Slashdot could benefit from a co-processor...

[FreakyLefty]

revives the concept op co-processors

It would also make proper use of no ops.
But they're giving us sneak peaks, so who are we to complain?

Amiga v2?

[BobLenon]

How 'bout Agnus, Denise and Paula. :)

Re:Amiga v2?

[fellip_nectar]

Or even ANTIC, CTIA and POKEY?

EOISNA

[omega9]

Everything old is new again.

Re:EOISNA

[jojoba_oil]

Does that mean goth is the new emo? Or black is the new pink? O, what a horrible, horrible world.

Amiga anybody?

[plebeian]

It is nice to see PC architecture has finally caught up with Amiga.

Re:Amiga anybody?

[Afecks]

It is nice to see PC architecture has finally caught up with Amiga.

It's nice to see you've finally caught up with all the people that have made an Amiga comment.

great writing

[Anonymous Coward]

The downside of this is that they aren't optimised for a specific task, hence their basically jack's of all trades but masters of none.

In the past the processor was the beating heart of the computer (hence the term Central Processing Unit) but with all the different developments in abovementioned areas the CPU is becoming a less determining factor for overall processing power within the PC.

That's interesting, because everyone in my parents' generation says the "CPU" is the whole box, and the chip from AMD or Intel is the "microprocessor".

CSI

[jlebrech]

Is that the infinite image enhancement chip?

Sneak Peak?

[Anonymous Coward]

Is this located near Stealthy Valley?

Re:Sneak Peak?

[LarsG]

It's certainly not in Plain View, NC.

Good Idea

[wolff000]

I like the concept and yes I know it's nothing new. I hope this thing takes off I would love to be able to just snap a single chip into place than have to deal with gigantic video cards. ALthough I suppose we would end up with more heatsinks on the mobo this way but at least my PCI slots wouldn't be so crowded.

As rumored, first addopted by the porn industry

[alta]

Prepare to see the pornprocessor soon. I'm not going to give a lot of details here, but it's optimized for specific physics, AI and Graphics.

Re:As rumored, first addopted by the porn industry

[demonbug]

Prepare to see the pornprocessor soon. I'm not going to give a lot of details here, but it's optimized for specific physics, AI and Graphics

AI? For porn? You have seen porn before, right?

Cell Clusters

[Doc Ruby]

How about the Cell uP (first appearing in Playstation3), which embeds a Power core on silicon with a 1.6Tbps token ring connecting up to 8 (more later) "FPUs", extremely fast DSPs. IBM's got 4 of them on a single chip, connected by their "transparent, coherent" bus, a ring of token rings. One Cell can master a slave Cell, and IBM is already debugging 1024 DSP versions, transparently scalable by the compiler or the Power "ringmaster" at runtime.

These little bastards are inherently distributed computing: a microLAN of parallel processors, linkable in a microInternet.

Imagine a Beowulf cluster of those! No, really: a Beowulf cluster of Cells.

Re:Cell Clusters

[spirit of reason]

And how good are those Cell yields again?

Electronic News: What's the defining factor that makes some chips better than others?

Reeves: Defects. It becomes a bigger problem the bigger the chip is. With chips that are one-by-one and silicon germanium, we can get yields of 95 percent. With a chip like the Cell processor, you're lucky to get 10 or 20 percent.

Source

Sounds like a good idea for the mass market to me...

Re:Cell Clusters

[Doc Ruby]

I don't know whether you're implying sarcastically that Cell yields are too low to be economical, which is what the short segment you quoted from out of the context you cited says. The point of the Cell is that

IBM has said it needs seven of the eight cores on the Cell processor to work for Sony's Playstation. Will there be an aftermarket for chips with fewer operational cores?
Reeves: There are a lot of chips with six cores operational, and we've been thinking about whether we should really throw all of those away. We also have a separate part number for chips with all eight cores good. The stuff that's going to be for medical imaging, aerospace and defense and data uses eight cores.

[...]

That doesn't mean there aren't good uses for a chip with four SPEs [synergistic processing elements].

Electronic News: What's the defining factor that makes some chips better than others?
Reeves: Defects. It becomes a bigger problem the bigger the chip is. With chips that are one-by-one and silicon germanium, we can get yields of 95 percent. With a chip like the Cell processor, you're lucky to get 10 or 20 percent.

They get 10-20% perfect, with 8 SPEs, that they use for medical imaging, aerospace/defense and data proc. They must get a lot more than 20% with a defective SPE, which goes into the mass-market PS3 at 7 SPEs. Which alone would pay for the entire run at $95-100 apiece: a $5B fab investment would require 5M chips sold to pay for it, if Cells accounted for 10% of Sony's production there. Sony has already shipped over 2M units at launch, with shortages defining the market. At this rate, Sony will probably get at least halfway to paying off the fab just on PS3 Cells. Before they sell the 8-SPE ones for the high-end apps (at much higher prices). And the 4-6 SPEs, and maybe even all the way down to 0-SPE Power970s for dirt cheap in all kinds of devices. And that's not even using the electronic fuses for "self healing" and even just field tailoring power consumption to just the performance needed. An expensive feature in a cheap, powerful chip.

The redundancy and degradable multiuse strategies exploit yields to segment markets. It's like the revolution in marketing different clock speeds as different products, not defects, resulting from production yields. But this time with parallel, not serial, processing.

Re:Cell Clusters

[spirit of reason]

I guess I figured that an integrated solution of CPU/GPU might be more asymmetric than the Cell and then have more problems with yield, but after thinking about how a G80 looks and works, I'm probably quite wrong there. If you started adding more flavors of specialized units, though, cost might become an issue again.

Re:Cell Clusters

[smallfries]

So the cell is the master ... and it communicates through the ring. Sort of like the ring is in charge here ... hmmm and there's only one ring. No, I thought there was an obvious joke but it's gone.

Re:Cell Clusters

[Doc Ruby]

We'll have to wait for the Easter Eggs in the inevitable Lord of the Rings PS3 sim.

Re:Cell Clusters

[Slashcrap]

They get 10-20% perfect, with 8 SPEs, that they use for medical imaging, aerospace/defense and data proc.

Can you provide a single example of the Cell being used in a medical/defence application or a news story from an independent (non-Sony/IBM) source suggesting that they will?

Because the thing about the medical, aerospace and defence industries is that they tend to be a bit conservative about the chips they use. And I find it extremeley hard to believe that the Cell is anywhere near to being validated for any of those applications. Do they actually make a Milspec Cell?

You see, when I read a quote like the above I mentally translate it to something like the following in realtime :

"They get 10% perfect, with 8 SPEs, that they would love to use for medical imaging and aerospace/defense if there was the slightest chance of anybody in those industries actually buying them in the forseable future. Which there probably isn't."

You seem to lack that ability.

The redundancy and degradable multiuse strategies exploit yields to segment markets.

Please stop breathing ASAP.

Re:Cell Clusters

[Doc Ruby]

Fuck you.

Oh, and I am working with an engineer on consumer Cell apps whose brother is an EE working with a European defense contractor already holding a contract to deliver Cell apps into a national military system. And his old medical imaging company has asked him to port their app to a Cell system. And he doesn't even work for IBM, Toshiba or Sony.

Did I mention fuck you?

can someone explain?

[ArcSecond]

Is this the same as a bus-oriented system? I remember spec'ing out systems for a defence contractor back in the 90s, and there were systems designed around "daughter-card" processors, something like a modular mainframe on the cheap. It always seemed to me that a bus-centric system had a lot going for it performance-wise, rather than forcing everything in the computer to synch to the CPU.

Re:can someone explain?

[ACMENEWSLLC]

>>Is this the same as a bus-oriented system?

That's what I was thinking.

I've got some of those now. There is a common backplane. Everything is dropped onto it. I have a card on which I plug the memory into, so I can have 20 memory chips if I like. I have a card/cards where I put the processors. So I can pull that card out and upgrade to a newer processor, or change families.

The problem with these is cost and performance. These are specialized machines and are expensive. Since I am adding another layer to the hardware, there is a performance hit. This is a industrial solution, so neither of those are a concern for me. But it wouldn't fly in the general market.

Where have you been,

[Dareth]

That is Paula, Randy, and Simon...

I mean, do you ever watch American Idol, hello!

Just plug in a spellchecker Co-Processor

[TrueKonrads]

Just plug in a spellchecker Co-Processor! I think no ordinary CPU could handle such massive mistakes

AMD competes with...

[Comboman]

AMD will compete by releasing "Law & Order: Central Processing Unit".

And the wheel turns...

[psycho8me]

http://catb.org/jargon/html/W/wheel-of-reincarnati on.html

AMIGA!

[elrick_the_brave]

This sounds vaguely like the Amiga platform of years past (with a fervent following today still)... how innovative to copy someone else!

Re:AMIGA!

[Slashcrap]

This sounds vaguely like the Amiga platform of years past (with a fervent following today still)... how innovative to copy someone else!

How ironic that you should say that when you're already the tenth fucking idiot to post a one line comment about how great the Amiga was.

Just so you know, I have the motherboard from an original A1200 here with me. I was going to give it away to anyone who might be able to make use of it. However after reading all the comments here, I've decided to snap it in half and throw it in the trash. I've had enough of whiney Amiga bitches. You're getting worse than fucking Mac users.

HTX

[Joe The Dragon]

HTX slots are better the pci-e ones and right now amd can make there desktop cpus driver 2 of them and it the 4x4 system you can have 2 chip set links and 2 htx slots.

I have a phone dock for that already

[DrSkwid]

It's called the LAN.
I turn on the WiFi and my phone is part of the internet.
It serves it's flash/ram/rom/sd via the 9p protocol.
My terminal can boot from it, if I wanted it to, yours could to if you were in my authentication server.
I can store encrypted data on it useless without TCP access the dock.

Re:I have a phone dock for that already

[walt-sjc]

Does your phone dock work in the car? On a plane? Does it add CPU power and memory? Does it give you a bigger screen? The answer to all of these is No. Phone docks basically allow you to charge your phone and sync with your desktop. The discussion is about replacing the desktop and using your phone as your main computer. Palmtop's have had add-on keyboards for years - they tend to suck as a primary keyboard however, and are still not very portable (folding keyboards are larger than the device by far.) Phone-style portable computing devices are no where near as powerful as desktop's, and heat / power issues will ensure that they remain that way for the foreseeable future.

OMG not again

[Anonymous Coward]

Just what everyone needs: another spin-off of CSI.

Unification

[IPFreely]

While there are a wide variety of co-processor options (or at least ideas) right now and few sockets into which to put them. I suspect the solution will more likely come in the form of unified co-processors rather than multiple sockets.

Mother board shipsets are becoming the union of a lot of functionality (Disk, Ethernet, Sound, UDB, PCI/e and graphics). Even though you can still get best of breed addin cards for many of these functions, the majority of desktop systems do just fine with what the chipset offers.

These coprocessors will also become unified. AMD and nVidia already are talking about doing physics. In the end, you are likely to get a single processor that does graphics, physics, AI, advanced math, and probably Java, sound and a few other things we have not thought of yet. As a single entity, it takes a single slot. So long as all the functionality is all accessable through some standardized interface (DirectX in the MS world, something else for everyone else) then the difference between the competing manufacturers will be about the same as the difference between graphics cards now.

You can buy Torrenza today

[LordMyren]

To the best of my knowledge, Torrenza is already implemented. The HTX port on many Opteron motherboards is a HyperTransport connection. You can already buy FPGA dev kits from U. of Mannheim that plug into this HyperTransport slot and interface with the rest of your system. Torrenza may continue to advance the HyperTransport / Coprocessor war, but as far as I'm concerned, Torrenza is already here.

Re:You can buy Torrenza today

[mzs]

Yes! Can you do battery backed SRAM behind this gate array?

Re:You can buy Torrenza today

[TheRaven64]

HTX is currently limited to 16-bits/800MHz, which is pretty anaemic by today's standards. This is the sort of bandwidth processors had back in the late '90s. For comparison, the main CPU is sitting on the end of a point-to-point HT interconnect that runs at 2.6Ghz, and is twice the width, giving a total of almost eight times the bandwidth. Since HT messages are comprised of 32-bit words, this width will also give a much lower latency (one cycle per word rather than two, and most messages are several words).

If you want a co-processor to run at a decent speed, you'd be better off designing it to plug directly into one of the CPU sockets (some FPGAs support this already), and getting a decent amount of bandwidth. HTX is really designed for high-speed peripherals such as infiniband interfaces, not for cp-processors.

Re:Vista

[chrwei]

Last night I had my first long session with Vista on my sister-in-law's new laptop, a Compaq with a Celeron M and 512M ram. Without disabling "user" functionality (the things that non-tech users like about windows) I was only able to get it down to using about 350MB ram at boot up. I didn't check how much ram the Theme service uses but I doubt that it is significant. The default config used over 400MB on bootup and I find this appalling for 2 reasons: 1) Why do so many tray icons need to use so much ram? With the amount of ram used, startup time was SLOWER than if these icons were just launchers! 2) How does it make sense to sell a product that just by turning it on the system resources are maxed out and so everything else you do runs super slow? These leaves a very "oh my this pc is slower than my old one" feeling, how does HP think the $50 they shave off the price for that extra 512M is worth this? To [ab]use the car analogy, this is like selling a car where the engine can only handle pushing the car with one passenger and the radio, fan/heat/air, and lights off. Even though it has seating for 5, adding just one more person or turning on the radio causes it to take so long to get up to highway speed that you'd just rather not drive it at all, but for a $50 upgrade it'll do anything you want!

The things in Vista I like are the Windows Explorer interface and the new start menu. Yeah that's about it, Well, at least I like them initially, after only a few hours with it. Plus most users don't understand file system usage concepts (and I don't think the new interface will make the concept any easier to follow) and they don't really care how the start menu looks as long as they can start their apps (or drag the icons they use most to the desktop), so this is not a significant improvement over XP. Of course, I said the same kinds of things about XP compared to Win2k when it first came out too, but then SP2 was released with the better-than-nothing-firewall and better wifi tools and that made it enough better that XP is a must on laptops. MS needs to find the same type of "must" features for Vista, and I wish them luck.

So guess what? the office pc's I manage will stay at P3's with 512Meg and Win2k for a while longer. What can the old, slow pc's do? AutoCAD Lite, Adobe Illustrator, ACT, all office apps, any all the custom stuff I've got for CRM and order management. They are not maxed out on RAM, ever, and a faster CPU does not significantly increase performance.

Plenty of papers have been written on the 2 basic ways to tackle upgrades. You can upgrade everything ever 2 years, probably with it all leased and support contracts, and have an affordable monkey to handle all that, or you can have a high paid sysadmin that can fix anything quickly. The costs overall are about the same, but the monkey will spend time on the phone with the jokers in tech support and wait for parts to ship, where the sysadmin will have it working in no time with the spare parts from old systems he keeps around. Since I can pick up the P3's for well under $150, including a COA, the CFO is happy too, he can just "expense" them and not have to make a lease payment or deal with depreciation.

Re:sockets

[chrwei]

As for the "processor socket", there are people selling computers that go into passive backplanes. If you put the CPU and memory in a card, there is little reason why you would have to upgrade the rest of the computer when you change the CPU (you would have to scrap the card, anyway, but processors are intimately related to chipsets, so, it is to be expected.

you mean like the G3 macs, but add memory? This is actually a pretty good idea, put the cpu/ram/chipsets on a card that would plug into the pci/ata/legacy chipsets using a standard interface. Many chipsets do this anyway with these things having their own chipsets and all connecting to the northbridge. Make the interface standard, like ATX/pci/memory/ata are standards, and then upgradability is a no-brainer! How much could Dell and HP lower their costs if the case and backplane didn't need changed to upgrade the core? At the very most the PSU would need upgraded, but if the platform was designed right you'd only need to do that every 4-6 years (assuming the 2 years cycle), and even sell a cpu/HDD combo upgrade combo preinstalled with that new OS too. Shipping and storing cases and all the other crap is the expensive part of the industry, they could save millions here.

Really just two types of processors

[J.R. Random]

There are basically two models of parallelism that are used in practice. One is the Multiple Instruction Multiple Data model, in which you write threaded code with mutexes and and the like for synchronization. The other is Single Instruction Multiple Data, in which you write code that operates on vectors of data in parallel, doing pretty much the same thing on each piece of data. (There are other models of parallelism, like dataflow machines, but they don't have much traction in real life.) Multicore CPUs are MIMD machines, GPUs are SIMD machines. All those other processors -- physics processors, video processors, etc. are just SIMD machines too, which is why Nvidia and ATI could announce that their processors will do physics too, and why folding@home works so well on the new ATI cards. So I suspect that in real life there will be just two types of processors. At least I hope that is the case, because it will be a real mess if application A requires processors X, Y, and Z while application B requires processors X, Q, and T.

Re:Really just two types of processors

[Slashcrap]

There are basically two models of parallelism that are used in practice. One is the Multiple Instruction Multiple Data model, in which you write threaded code with mutexes and and the like for synchronization. The other is Single Instruction Multiple Data, in which you write code that operates on vectors of data in parallel, doing pretty much the same thing on each piece of data.

And then there's the RISC model - MISD.

This is where a long stream of single instructions each do basically fuck all to a single piece of data.

Definitely.

[Svartalf]

I remember the Amiga. I remember how much more capable and powerful they were over the other "personal" computers of the day.

It's a damn shame that Commodore couldn't market/sell their way out of a wet paper bag.

Moo

[Chacham]

and sneak peaks into the future of processors.

and sneak peaks into the past^H^H^H^Hfuture of processors.

This may have "IP" issues, it's been done before..

[Anonymous Coward]

...with the now defunct Rendition Socket X http://findarticles.com/p/articles/mi_m0EKF/is_n22 26_v44/ai_20894055 http://findarticles.com/p/articles/mi_m0CGN/is_n12 5/ai_20831295

One word

[Anonymous Coward]

Emulation

Nice and dandy but..

[JackMeyhoff]

.. what I want is the ability to install ANY number of processors from any manufacture onto a mainboard or slot. I have spare PCI slots why can I install more processors? Damn waste of space.

Re:Nice and dandy but..

[Anonymous Coward]

.. what I want is the ability to install ANY number of processors from any manufacture onto a mainboard or slot.

I have spare PCI slots why can I install more processors? Damn waste of space. 1) power draw
2) Memory Bandwidth (may be alleviated by PCI-E I haven't kept up)
3) They can't charge you for MB upgrades because they decided to change the pinout. :)

Re:Nice and dandy but..

[JackMeyhoff]

A PCI-E box full of multi core mobile chips. Your desktop probably uses more power than your fridge and freezer. With a couple of 3.2 HD's ( I dumped 3.5 long ago ), 500 to 700 Watt PSU, screen, CPUs and most deadly of all 1 or 2 GPU's I would prefer to scale OUT rather than UP.

Re:Nice and dandy but..

[JackMeyhoff]

3.2 nope I meant 3.5", those things suck quite a bit of power especially since most people run an always on power profile. Selfish, Im in the Eco Green PC fan club. Yes yes I know the metals used in them are toxic and nasty but I cut down my pwoer usage. 40% of the power in the country I am in currently is GREEN which is great but still we should use the least power possible. Low power but scaled out.

Cyclical

[nsayer]

We've seen this before. The industry is constantly cycling between specialized co-processors (what I loosely call asymetric multiprocessing) to increase performance and increasingly powerful central processors and dumb peripherals to decrease cost and bus latencies. What's old is new again.

Re:Cyclical

[TheRaven64]

This time, expect hetrogeneous multiprocessing to stay. Specialised hardware always wins on raw performance. General purpose always wins on performance per dollar. Specialised wins on performance per watt. If performance per dollar matters more than raw performance or performance per watt, then you are probably in one of the market segments that can happily make do with a five-year-old CPU.

Modular Design

[Kensai7]

I like this modular design for CPUs. A main processor paired with the co-processor handy for the task. Hey, where can I get one of those fancy AIPUs?!

Amiga love.......memories.........

[LibertineR]

Got mine on Day 1 also.

An Amiga 1000, Deluxe Paint, Flight Simulator, Amiga Basic and 2Meg of Ram = $3500.

Later got the Sidecar for DOS, and Earl Weaver Baseball. Ahhhhhhhh.

20 years later, and no hardware or software has given me such joy.

NVidia, Matrox, ATI, AMD, Intel, WTF?

The Amiga showed you how 20 years ago, and you are just now getting around to it?

Bring back multi-resolution windows, bitches!

looks like a revamped AMD 4x4

[ZirbMonkey]

If I was reading the picture on the second page correctly, it looks like AMD plans to use a "4x4" type motherboard architecture, but with the second CPU spot made for a dedicated GPU chip instead of another redundant CPU. The CPU and GPU wouldn't be on the same die in this case.

I think this would make sense to me. Right now when I upgrade my video card, I throw out the ram, GPU, and integrated circuitry of the entire package to replace everything with the new video card upgrade (which happens every 6 months for me). What if I could buy the GPU and DDR3 separately and not throw out everything each upgrade? Not only would the infrastructure be faster, but the upgrades could be cheaper since you don't need to buy the whole package every time.

This obviously only matters to the enthusiast trying to keep on the edge of Moore's Law. I like the idea, but we'll see how things turn out in reality of 2010.

Its the memory, stupid

[Anonymous Coward]

This article ignores the main issue with GPU integration -- its the memory stupid. Current high end GPU boards have an order of magnitude more memory bandwidth than the Torrenza socket provides. At least 75% of the cost of a graphics board is just the memory chips. Sure, you could put the whole lot on the motherboard, but all you're saving is the cost of a connector. As long as it makes sense for GPUs to have their own separate high-performance memory subsystem, its going to make sense to have them with separate chips on a separate board. Since the cost of memory (bandwidth and latency) has not been decreasing as fast as the cost of CPU transistors in the past, it seems unlikely to do so in the future, so this seems unlikely to change.

Re:Vista

[Joosy]

The default config used over 400MB on bootup and I find this appalling ...

Relax, it's just doing a bunch of caching. If you load a program that needs more RAM you can still use most of that 400MB. This has been discussed many places. A better question is why your sister bought a Vista laptop with just 0.5GB.

Re:Vista

[chrwei]

the 15+ seconds it takes to load any application, including IE, says otherwise. the "cache" and "free" values in task manager also say otherwise. I can deal with cache, but this is not JUST cache. I'm also talking about initial load times from a cold boot, before non-essential stuff is moved to swap and the cache is filled only with the system services and utility crap that lives in the system tray. Also, cache is important, it speeds things up a lot if you have enough of it, but if you don't and you are constantly having to flush more than 50% of the cache, then this makes things slower than not having a cache at all.

2nd part is easy: it was on sale in a store and 1GB version was not offered there, an additional 512 was ordered the same day. My point was that someone who doesn't know better would use this and feel like their old P3 with 256Meg was so much more responsive, and they would be right. Why would HP even offer such a slug? Their preload just barely fits in the 512MB, it's pointless.

Individual Processors?

[aeinstein44094]

This sounds like what Commodore did the Amiga 1000. It had a seperate proc for Graphics nameed Denise, one for sound named Paula and the other was the central chip named Agnus or Fat Agnus in the second chip set. There were a lot of benefits to this design and at the time of its release it was a superior configuration when compare to the PC or even MAC at the time of its release.

and build it inside a monitor LCD.... like..imacs

[cheekyboy]

Yeah, clone the imacs, real cheap, or maybe at min a small slot for the cpu/miniboard computer inside the monitor like the high
end display panasonics for sales displays.

let's think how many ways this NOT like amiga.

[Ayanami Rei]

1) Amiga didn't have a unified memory space with NUMA cache-coherency between all participating chips
2) Amiga had multiple processors, but they were all Commodore parts, and soldered in. We're talking about bus standards, and ISAs, and your choice of vendors and upgradability and all that stuff which is more difficult to spec-out AND get buy-in for. It's not a vendor stovepipe.

Hell, a friggin SNES has 4 coprocessors, a TurboGFX 16 had like 6, but you don't see people comparing THAT to PCs or Amigas or anything else.

Amigas weren't the only ones. Have you ever looked at an SGI workstation from the same era? Same kind of architecture, only more flexible.

PC architectures post-ISA were already surpassing the Amiga in potential. Bus interconnects, add-on cards, all that stuff was there. The software wasn't all there if you were using Microsoft though (one dynamite one was BeOS, but no one remembers that).

I don't know. You Amiga fanboys don't even know the significance of your own system in the grand scheme of 80s and 90s computing technology. You have tunnel vision.
It was important for commercial reasons, like the Commodore 64. But it wasn't technologically groundbreaking nor unique, nor better or worse than any other architecture.

*shakes head*

Re:let's think how many ways this NOT like amiga.

[blargh-dot-com]

Just a couple minor corrections:

1) Amiga didn't have a unified memory space with NUMA cache-coherency between all participating chips The Amiga did have a unified memory space. See http://en.wikipedia.org/wiki/Chip_RAM . The cache bit, not so much. :)

2) Amiga had multiple processors, but they were all Commodore parts, and soldered in. We're talking about bus standards, and ISAs, and your choice of vendors and upgradability and all that stuff which is more difficult to spec-out AND get buy-in for. It's not a vendor stovepipe. Only a small note here, on MOST Amigas the chips (actually almost all the chips) were in sockets. For example, Amiga 500 owners commonly upgraded their Fat Agnus chips to Fatter Agnus (512KB chip RAM access to 1MB chip RAM, plus a couple minor details). Etc. etc. They were all Commodore parts, however.

As for bus architectures, Zorro III was theoritical 50MB/sec (in practice could only be driven around 20MB/sec, according to http://www.thule.no/haynie/systems/amiga3k/docs/a3 000.des.txt ), and there was the CPU slot and Video slot.

It's because it doesn't take much...

[Anonymous Coward]

> AI? For porn? You have seen porn before, right?

I figured that the only reason they were able to model it with an AI was because it was so simple? :)

>> Eliza: What about your simple?

I mean, it's not like we have really powerful AIs at our disposal to begin with, or like we need them.

--
This post was not generated by a Markov chain, but it probably could've been.

Amiga my ass.

[morriscat69]

Atari 800...

....beeeches.

Re:Amiga my ass.

[aeinstein44094]

Atari may have come before the Amiga but the Amiga also used coprocs and it blew the Atari into a black hole.

Really just two types of realities.

[Anonymous Coward]

"There are other models of parallelism, like dataflow machines, but they don't have much traction in real life."

That's like saying ASICs don't have any traction in real life.

Re:Really just two types of realities.

[Anonymous Coward]

Or if that's not "traction" enough for you.

To market, to market
In 1985, Patt's research team added to Tomasulo's limited dataflow technique and suggested applying it to all chip operations, through a microarchitecture called HPS, for high-performance substrate. And by the early '90s, most of the industry began listening. What is the Pro in the Pentium Pro? An on-the-fly dataflow architecture called dynamic scheduling or out-of-order execution. Dynamic scheduling applies the benefits of dataflow processing to conventional programs. And software coding is not an obstacle, because the dataflow elements are written into the chip. Instructions flow in and out of the processor in sequential program order, but internally they are converted into a dataflow graph and executed according to the availability of data."

Re:Really just two types of realities.

[J.R. Random]

Yes, but that dataflow isn't accessible at the programmer level. To a programmer, its just a sequential processor. I'm talking about parallel architectures that are visible to the programmer.

Re:and build it inside a monitor LCD.... like..ima

[Fordiman]

I don't know about the iMacs. They weren't even SBC's, let alone SCC's.

Cynical

[Anonymous Coward]

"What's old is new again."

And Lord knows we can't have that AGAIN, now can we?

Cyclical-Man in the middle.

[Anonymous Coward]

There's an inbetween.

Re:Vista

[Anonymous Coward]

To [ab]use the car analogy, this is like selling a car where the engine can only handle pushing the car with one passenger and the radio, fan/heat/air, and lights off. Even though it has seating for 5, adding just one more person or turning on the radio causes it to take so long to get up to highway speed that you'd just rather not drive it at all, but for a $50 upgrade it'll do anything you want! Yes, I had a Dodge Shadow. It also blew head gaskets.

RIA freudian investigation

[islisis]

While AMD might be cluing us in on the next big consumer processing fetish, be sure to keep your private details hidden if you choose to go for the Torrenza. The RIA could be bustin down your door any second for suspicion of intent to speed up your multi-peer downloads...