Kickaha and Amorph couldn't moderate themselves out of a paper bag. Abdicate responsibility and succumb to idiocy. Two years of letting a member make personal attacks against others, then stepping aside when someone won't put up with it. Not only that but go ahead and shut down my posting priviledges but not the one making the attacks. Not even the common decency to abide by their warning (afer three days of absorbing personal attacks with no mods in sight), just shut my posting down and then say it might happen later if a certian line is crossed. Bullshit flag is flying, I won't abide by lying and coddling of liars who go off-site, create accounts differing in a single letter from my handle with the express purpose to decieve and then claim here that I did it. Everyone be warned, kim kap sol is a lying, deceitful poster.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Kickaha and Amorph couldn't moderate themselves out of a paper bag. Abdicate responsibility and succumb to idiocy. Two years of letting a member make personal attacks against others, then stepping aside when someone won't put up with it. Not only that but go ahead and shut down my posting priviledges but not the one making the attacks. Not even the common decency to abide by their warning (afer three days of absorbing personal attacks with no mods in sight), just shut my posting down and then say it might happen later if a certian line is crossed. Bullshit flag is flying, I won't abide by lying and coddling of liars who go off-site, create accounts differing in a single letter from my handle with the express purpose to decieve and then claim here that I did it. Everyone be warned, kim kap sol is a lying, deceitful poster.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Kickaha and Amorph couldn't moderate themselves out of a paper bag. Abdicate responsibility and succumb to idiocy. Two years of letting a member make personal attacks against others, then stepping aside when someone won't put up with it. Not only that but go ahead and shut down my posting priviledges but not the one making the attacks. Not even the common decency to abide by their warning (afer three days of absorbing personal attacks with no mods in sight), just shut my posting down and then say it might happen later if a certian line is crossed. Bullshit flag is flying, I won't abide by lying and coddling of liars who go off-site, create accounts differing in a single letter from my handle with the express purpose to decieve and then claim here that I did it. Everyone be warned, kim kap sol is a lying, deceitful poster.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Personally I could care less about the high end, I want to see a PM back at the $1499 or $1599 price, or better yet a replacement for the G4's that they have discontinued starting at $1299. That will do more for Apple's market share than a high end tower or another overpriced iMac.
Cheaper prices= more marketshare is a myth. Take AMD for instance. The "perception" is, with the Athlon64 and Opteron kicking butt, AMD is on a roll but sadly their Q3 earnings tell otherwise. Despite enthusiasm amongst users AMD only turned a 30 million profit...half that of Apple who is expected to vacate the hardware market because they can't compete (so says Rob Enderle). Apple is working for their investors. I want to see more marketshare as well but I don't think it's pricing that dictates it but rather a bevy of many different things.
The high end systems are where the profit is at. In fact I hope Apple's Quad system is $3999. Let the folks deep pockets fund and subsidize those of us without the same.
About the interconnect, it seems that the architecture will differ from the power 4. The power 4 has three independants L2 cache, and a huge L3 controller (which is in charge also of communication between various power 4 chips).
The PPC 970 MP seems to have one dedicated L2 cache per core. Like air sluf said the interconnect will be at full clock speed (there is no reason to do otherwise), but the wide of the bus is unknown (64 bits, 128 bits ? ...). I guess that the communication between the two cores, will be de facto a communications between the two L2 cache via a controller.
Finally, the G5 has higher clock speeds than what AMD is putting on the market, and you dont hear anyone whining about them not hitting 3ghz. What is that, they have 3200+ cpus, oh, thats the _MADE UP_ number that they use for speed.
1. AMD never promised 3GHz.
2. AMD is the current king of the hill in performance. Their Opterons kill everything in the low-end space -- G5s, P4's, Xeons, etc... No one is complaining about the low clockspeed, because unlike the G5, they are actually faster than the competition.
3. It's not a made-up number. Benchmarks clearly show, especially with the Athlon64 line, that it corresponds well with Intel.
Quote:
t's not a secret that mac os X, is much more optmized for MP than X86 PC.
Cheaper prices= more marketshare is a myth. Take AMD for instance. The "perception" is, with the Athlon64 and Opteron kicking butt, AMD is on a roll but sadly their Q3 earnings tell otherwise. Despite enthusiasm amongst users AMD only turned a 30 million profit...half that of Apple who is expected to vacate the hardware market because they can't compete (so says Rob Enderle). Apple is working for their investors. I want to see more marketshare as well but I don't think it's pricing that dictates it but rather a bevy of many different things.
Apples and oranges.
AMD doesn't sell systems. Apple does. AMD may offer great bang for buck to DIY'ers and small assemblers, but they don't have the ability to produce enough volume to get 1st tier manufacturers and keep them. Those design wins go to intel 9 out of 10 times. And since most people buy their systems from major manufacturers rather than small ones, and even fewer build their own, Intel keeps on mopping the floor with AMD, regardless of the performance or price. I'm willing to bet that Intel even beats AMD on price when it comes to delivering 100's of thousands of parts for large production runs. I'd be surprised in fact, if overall, it wasn't cheaper for companies like Dell, HP, and IBM to turn to Intel for the CPU needs, than to AMD. AMD only retails lower. Intel very agressively cuts prices to major manufactures. You and I can't get those prices, nor can some relatively low volume builder, liek an Alienware, but the guys with the sales (the guys that matter), they're doing just fine buying Intel.
Does anyone have an idea when the PowerPC chips based off of the Power5 series will appear? My guess is that this core would be much better suited to multicore. I think that I and others expected that those chips would be around soon, as in the end of the year time frame. Could this rumor have mistaken the pending release of the 980, as in the end of year or early '05 for a 970MP. I would think that a 980MP would be easier to do because I thought that IBM was going to make a multicore PowerPC based off of the Power5 series wheather Apple was going to use it or not. Maybe someone else will remember these details, maybe those chips were originally for XBox or IBM servers. Other things that I remember is that the 980 was supposed to be for low power operation, a cooler core chip. What better to use as dual core than a core that is designed to run cooler.
Quote:
Originally posted by Powerdoc
Concerning the watt/heat issue :
1) The first version of the Athlon 0,13 produced nearly the same heat than the latest generation of Athlon 0,18. Then they introduced some tweaks (and if my memory is correct add a layer) droppind down watt consumption (by a 30 % factor).
The 970 Fx is not a huge benefit in watt consumption compared to the original PPC 970 : we can expect a more efficient design especially if they go in the new 10 layers process.
2) one problem of the PPC 970 FX is the small die : a 2,5 ghz 970 FX approximatively produce the same heat than a 2 ghz PPC 970 but on a much smaller die : thus Apple was obliged to use watercooling to extract heat from a smaller surface of the chip.
The PPC 970 have a larger die, thus heat on the new process will be less concentrated, and thus less difficult to evacuate.
If you consider point 1 and 2, it seems that a dual 970 MP Apple computer is a serious option.
I think we should see again two version of the powermac : the single MP powermac and the dual MP powermac. (two single models, and one dual)
Concerning the speed issue :
Due to the larger Cache (One MB of cache per core) and due to the high speed interconnection of the cores, a single computer 2,5 Ghz 970 MP will be faster than a dual 970 Fx. But (there is always a but), as the MP version is more pipelined, and dispite the improved branch predict unit, a pipeline penalty may occur with this new version.
We will have to wait for the first benchmarks (who will be based upon a new mobo with faster RAM IMO) to know this point. However, this new design will reduce costs, and will allow higher clock speeds.
Does anyone have an idea when the PowerPC chips based off of the Power5 series will appear? My guess is that this core would be much better suited to multicore.
The POWER4 and POWER5 are very similar in the respect of multi core. And they are very well suited for it too. IBM probably had a multi core variant of the 970 in mind all along so the 970MP won't be anything like a hack with just two cores put together for good measure.
Quote:
Originally posted by Brendon
I would think that a 980MP would be easier to do because I thought that IBM was going to make a multicore PowerPC based off of the Power5 series wheather Apple was going to use it or not.
IBM is not relying on Apple for the future of its PowerPC line in general or 970 in perticular. IBM will use them themselves (and I hope in Linux workstations too, not only servers) and besides that we have the likes of Microsoft who'll become a substantially larger customer for these kinds of high performance CPUs than Apple will ever be.
A light version of the POWER5 _is_ in the works, that's certain, but we can oly guess in what time frame. I expect it to come in all kinds of flavours from single core, to mobile, dual core and even more. The Xenon-processor in Xbox NeXT will be tripple core, and it'll be based on POWER5 since it is also hyperthreaded, something that is a new feature of the POWER5.
I wouldn't be surprised if a single core version om GR-UL (aka 975/980/Power5 light) showed up around the same time frame as 970MP.
Perhaps if I was focused on the products but I'm not. I'm strictly concerned with Profits thus the comparison is "Apples to Apples".
Quote:
I wouldn't be surprised if a single core version om GR-UL (aka 975/980/Power5 light) showed up around the same time frame as 970MP.
Neither would I. I think Apple has some options here on how they want to handle their lineup. I wouldn't mind seeing the P5 derivative ship in 2006 with all the "trimmings"
Internal enhancements
SMT
Dual cores with SMT on each core
beefier Altivec
Ondie Memory Controller
65nm fab
That's worth waiting 18 months for isn't it? The 970MP could take us to 3.2-3.4Ghz in the meantime.
The Xenon-processor in Xbox NeXT will be tripple core, and it'll be based on POWER5 since it is also hyperthreaded, something that is a new feature of the POWER5.
I didn't try, and quote you twice on purpose Henriok it just turned out that waybecause I wanted to reference these two. <img src="/images/smilies/1smile.gif" border="0" alt="" title="aSmile" class="inlineimg" /> <br />
<br />
<br />
<div class="quote-container"><span>Quote:</span>
<div class="quote-block">
<i>Originally posted by Henriok </i><br />
<b>The competition offers quad systems _today_ and have for some time, but Apple is not doing anything about it. If cooling a single core Prescott is hard, then I ccan't imagine how hard it will be to cool a dual core, let alone two or even four dual core chips. </b>
</div>
</div>
<br>The thing about those Quad Opteron, and Quad XEON boxes is that the actual real performance gains on things like 3D, or anything we would find useful are limited. It's great in servers, but only the Sparc from SUN, and it's 4 way SIMD unit in it's processors seems to pack any punch in territory where we could see some performance to justify the additional price of 4 processors. Intel, and AMD have not the same effect. And SUN can not seem to sell squat outside of the military for intelligence research. IBM on the other hand is another animal all together. They are capable, but as I was saying Apple didn't need to make a quad single core motherboard. Who would they have marketed it to? They don't have a big enough share of the enterprise community, and they are just laying the groundwork in super-computing where IBM already is well known. I think there was no point in making a quad single core motherboard wit 4 physical chip sockets for 4 physical chips.<br />
<br />
<div class="quote-container"><span>Quote:</span>
<div class="quote-block">
<i>Originally posted by Henriok </i><br />
<br />
IBM is not relying on Apple for the future of its PowerPC line in general or 970 in perticular. IBM will use them themselves (and I hope in Linux workstations too, not only servers) and besides that we have <b>the likes of Microsoft who'll become a substantially larger customer for these kinds of high performance CPUs than Apple will ever be.<br />
</b>
</div>
</div>
<br>That remains to be seen. Sure M$ will sell a butt load of XBOX's in the US, (and it will probably stop there again) but the team of IBM, and Apple has not shown us if they are actually going to come out, and play with the big boys, and make some more competitive products so Apple can have a true workstation class computer. This also remains to be seen, but once they do Apple will gain ground as will IBM. <br />
<br />
<div class="quote-container"><span>Quote:</span>
<div class="quote-block">
<i>Originally posted by Aphelion </i><br />
<b><br /> I'm tired of Apple trying to attain parity with Wintel (and AMD) systems. Even if you summon up the MHz-Myth and call the 2.5 970FX a "3200+" a-la AMD factoring (some would say "2999+"), it still is too close to call. The "worlds fastest personal computer" should be an unassailable claim.<br />
<br />
Apple needs to set a performance bar that clearly exceeds that of the rest of the personal computer industry. And they need to do it by a wide margin in every category, Not just Photoshop bake-off's, but SPEC scores, and every other conceivable benchmark. Cross platform applications must run significantly faster on the top end Macintosh, even those that are not well optimized for the Mac.<br />
<br />
To limit dual core chips to cheaper duals (or simply more margin) for the flagship towers would be unthinkable to me. A chance to dominate the high end for once must not be missed.<br />
<br />
<br />
... </b>
</div>
</div>
<br><br />
My sentiments are quite similar. They should come out and set the bar much higher anyway, and not lie about the actual speeds. I already know AMD kicks the crap out of a G5 in everything (even photoshop) If the actual performance was a good as WWDC 2003 said it was it would have been great, but it was false information, and bunch of BS false marketing.
It's not totally unwarranted. It's based on this anonymous white paper post that was making headlines in certain circles around the internet.
Quote:
Xenon Hardware Overview
By Pete Isensee, Development Lead, Xbox Advanced Technology Group
This documentation is an early release of the final documentation, which may be changed substantially prior to final commercial release, and is confidential and proprietary information of MS Corporation. It is disclosed pursuant to a nondisclosure agreement between the recipient and MS.
?Xenon? is the code name for the successor to the Xbox® game console from MS. Xenon is expected to launch in 2005. This white paper is designed to provide a brief overview of the primary hardware features of the console from a game developer?s standpoint.
Caveats
In some cases, sizes, speeds, and other details of the Xenon console have not been finalized. Values not yet finalized are identified with a ?+? sign, indicating that the numbers may be larger than indicated here. At the time of this writing, the final console is many months from entering production. Based on our experience with Xbox, it?s likely that some of this information will change slightly for the final console.
For additional information on various hardware components, see the other relevant white papers.
Hardware Goals
Xenon was designed with the following goals in mind:
?Focus on innovation in silicon, particularly features that game developers need. Although all Xenon hardware components are technologically advanced, the hardware engineering effort has concentrated on digital performance in the CPU and GPU.
?Maximize general purpose processing performance rather than fixed-function hardware. This focus on general purpose processing puts the power into the Xenon software libraries and tools. Rather than being hamstrung by particular hardware designs, software libraries can support the latest and most efficient techniques.
?Eliminate the performance issues of the past. On Xbox, the primary bottlenecks were memory and CPU bandwidth. Xenon does not have these limitations.
Basic Hardware Specifications
Xenon is powered by a 3.5+ GHz IBM PowerPC processor and a 500+ MHz ATI graphics processor. Xenon has 256+ MB of unified memory. Xenon runs a custom operating system based on MS® Windows NT®, similar to the Xbox operating system. The graphics interface is a superset of MS® Direct3D® version 9.0.
CPU
The Xenon CPU is a custom processor based on PowerPC technology. The CPU includes three independent processors (cores) on a single die. Each core runs at 3.5+ GHz. The Xenon CPU can issue two instructions per clock cycle per core. At peak performance, Xenon can issue 21 billion instructions per second.
The Xenon CPU was designed by IBM in close consultation with the Xbox team, leading to a number of revolutionary additions, including a dot product instruction for extremely fast vector math and custom security features built directly into the silicon to prevent piracy and hacking.
Each core has two symmetric hardware threads (SMT), for a total of six hardware threads available to games. Not only does the Xenon CPU include the standard set of PowerPC integer and floating-point registers (one set per hardware thread), the Xenon CPU also includes 128 vector (VMX) registers per hardware thread. This astounding number of registers can drastically improve the speed of common mathematical operations.
Each of the three cores includes a 32-KB L1 instruction cache and a 32-KB L1 data cache. The three cores share a 1-MB L2 cache. The L2 cache can be locked down in segments to improve performance. The L2 cache also has the very unusual feature of being directly readable from the GPU, which allows the GPU to consume geometry and texture data from L2 and main memory simultaneously.
Xenon CPU instructions are exposed to games through compiler intrinsics, allowing developers to access the power of the chip using C language notation.
GPU
The Xenon GPU is a custom 500+ MHz graphics processor from ATI. The shader core has 48 Arithmetic Logic Units (ALUs) that can execute 64 simultaneous threads on groups of 64 vertices or pixels. ALUs are automatically and dynamically assigned to either pixel or vertex processing depending on load. The ALUs can each perform one vector and one scalar operation per clock cycle, for a total of 96 shader operations per clock cycle. Texture loads can be done in parallel to ALU operations. At peak performance, the GPU can issue 48 billion shader operations per second.
The GPU has a peak pixel fill rate of 4+ gigapixels/sec (16 gigasamples/sec with 4× antialiasing). The peak vertex rate is 500+ million vertices/sec. The peak triangle rate is 500+ million triangles/sec. The interesting point about all of these values is that they?re not just theoretical?they are attainable with nontrivial shaders.
Xenon is designed for high-definition output. Included directly on the GPU die is 10+ MB of fast embedded dynamic RAM (EDRAM). A 720p frame buffer fits very nicely here. Larger frame buffers are also possible because of hardware-accelerated_ partitioning and predicated rendering that has little cost other than additional vertex processing. Along with the extremely fast EDRAM, the GPU also includes hardware instructions for alpha blending, z-test, and antialiasing.
The Xenon graphics architecture is a unique design that implements a superset of Direct3D version 9.0. It includes a number of important extensions, including additional compressed texture formats and a flexible tessellation engine. Xenon not only supports high-level shading language (HLSL) model 3.0 for vertex and pixel shaders but also includes advanced shader features well beyond model 3.0. For instance, shaders use 32-bit IEEE floating-point math throughout. Vertex shaders can fetch from textures, and pixel shaders can fetch from vertex streams. Xenon shaders also have the unique ability to directly access main memory, allowing techniques that have never before been possible.
As with Xbox, Xenon will support precompiled push buffers (?command buffers? in Xenon terminology), but to a much greater extent than the Xbox console does. The Xbox team is exposing and documenting the command buffer format so that games are able to harness the GPU much more effectively.
In addition to an extremely powerful GPU, Xenon also includes a very high-quality resize filter. This filter allows consumers to choose whatever output mode they desire. Xenon automatically scales the game?s output buffer to the consumer-chosen resolution.
Memory and Bandwidth
Xenon has 256+ MB of unified memory, equally accessible to both the GPU and CPU. The main memory controller resides on the GPU (the same as in the Xbox architecture). It has 22.4+ GB/sec aggregate bandwidth to RAM, distributed between reads and writes. Aggregate means that the bandwidth may be used for all reading or all writing or any combination of the two. Translated into game performance, the GPU can consume a 512×512×32-bpp texture in only 47 microseconds.
The front side bus (FSB) bandwidth peak is 10.8 GB/sec for reads and 10.8 GB/sec for writes, over 20 times faster than for Xbox. Note that the 22.4+ GB/sec main memory bandwidth is shared between the CPU and GPU. If, for example, the CPU is using 2 GB/sec for reading and 1 GB/sec for writing on the FSB, the GPU has 19.4+ GB/sec available for accessing RAM.
Eight pixels (where each pixel is color plus z = 8 bytes) can be sent to the EDRAM every GPU clock cycle, for an EDRAM write bandwidth of 32 GB/sec. Each of these pixels can be expanded through multisampling to 4 samples, for up to 32 multisampled pixel samples per clock cycle. With alpha blending, z-test, and z-write enabled, this is equivalent to having 256 GB/sec of effective bandwidth! The important thing is that frame buffer bandwidth will never slow down the Xenon GPU.
Audio
The Xenon CPU is a superb processor for audio, particularly with its massive mathematical horsepower and vector register set. The Xenon CPU can process and encode hundreds of audio channels with sophisticated per-voice and global effects, all while using a fraction of the power of a single CPU core.
The Xenon system south bridge also contains a key hardware component for audio?XMA decompression. XMA is the native Xenon compressed audio format, based on the WMA Pro architecture. XMA provides sound quality higher than ADPCM at even better compression ratios, typically 6:1?12:1. The south bridge contains a full silicon implementation of the XMA decompression algorithm, including support for multichannel XMA sources. XMA is processed by the south bridge into standard PCM format in RAM. All other sound processing (sample rate conversion, filtering, effects, mixing, and multispeaker encoding) happens on the Xenon CPU.
The lowest-level Xenon audio software layer is XAudio, a new API designed for optimal digital signal processing. The Xbox Audio Creation Tool (XACT) API from Xbox is also supported, along with new features such as conditional events, improved parameter control, and a more flexible 3D audio model.
Input/Output
As with Xbox, Xenon is designed to be a multiplayer console. It has built-in networking support including an Ethernet 10/100-BaseT port. It supports up to four controllers. From an audio/video standpoint, Xenon will support all the same formats as Xbox, including multiple high-definition formats up through 1080i, plus VGA output.
In order to provide greater flexibility and support a wider variety of attached devices, the Xenon console includes standard USB 2.0 ports. This feature allows the console to potentially host storage devices, cameras, microphones, and other devices.
Storage
The Xenon console is designed around a larger world view of storage than Xbox was. Games will have access to a variety of storage devices, including connected devices (memory units, USB storage) and remote devices (networked PCs, Xbox Live?). At the time of this writing, the decision to include a built-in hard disk in every Xenon console has not been made. If a hard disk is not included in every console, it will certainly be available as an integrated add-on component.
Xenon supports up to two attached memory units (MUs). MUs are connected directly to the console, not to controllers as on Xbox. The initial size of the MUs is 64 MB, although larger MUs may be available in the future. MU throughput is expected to be around 8 MB/sec for reads and 1 MB/sec for writes.
The Xenon game disc drive is a 12× DVD, with an expected outer edge throughput of 16+ MB/sec. Latency is expected to be in the neighborhood of 100 ms. The media format will be similar to Xbox, with approximately 6 GB of usable space on the disk. As on Xbox, media will be stored on a single side in two 3 GB layers.
Industrial Design
The Xenon industrial design process is well under way, but the final look of the box has not been determined. The Xenon console will be smaller than the Xbox console.
The standard Xenon controller will have a look and feel similar to the Xbox controller. The primary changes are the removal of the Black and White buttons and the addition of shoulder buttons. The triggers, thumbsticks, D-pad, and primary buttons are essentially unchanged. The controller will support vibration.
Xenon Development Kit
The Xenon development environment follows the same model as for Xbox. Game development occurs on the PC. The resulting executable image is loaded by the Xenon development kit and remotely debugged on the PC. MS® Visual Studio® version 7.1 continues as the development environment for Xenon.
The Xenon compiler is based on a custom PowerPC back end and the latest MS® Visual C++® front end. The back end uses technology developed at MS for Windows NT on PowerPC. The Xenon software group includes a dedicated team of compiler engineers updating the compiler to support Xenon-specific CPU extensions. This team is also heavily focused on optimization work.
The Xenon development kit will include accurate DVD emulation technology to allow developers to very precisely gauge the effects of the retail console disc drive.
Miscellaneous Xenon Hardware Notes
Some additional notes:
?Xenon is a big-endian system. Both the CPU and GPU process memory in big-endian mode. Games ported from little-endian systems such as the Xbox or PC need to account for this in their game asset pipeline.
?Tapping into the power of the CPU is a daunting task. Writing multithreaded game engines is not trivial. Xenon system software is designed to take advantage of this processing power wherever possible. The Xbox Advanced Technology Group (ATG) is also exploring a variety of techniques for offloading graphics work to the CPU.
?People often ask if Xenon can be backward compatible with Xbox. Although the architecture of the two consoles is quite different, Xenon has the processing power to emulate Xbox. Whether Xenon will be backward compatible involves a variety of factors, not the least of which is the massive development and testing effort required to allow Xbox games run on Xenon.
... "as I was saying Apple didn't need to make a quad single core motherboard. Who would they have marketed it to? ... I think there was no point in making a quad single core motherboard wit 4 physical chip sockets for 4 physical chips."...
A quad Macintosh, whether single or dual core in design, would be welcomed by anyone using Apple's high end video and imaging applications. Logic users might put it to good use as well.
I was most likely early in calling for a quad Mac (around the release of the MERSI compliant Motorola 7410), but it was the capability of the 7410 in a MP environment that made me call for a Wintel killer.
As in all things computer, it's the software that makes things possible, and in 2000 OSX was new, and most likely not quite ready for four processor systems. And applications that could fully use such a system were not mainstream, if they were available at all.
But that was then and this is now. Developers have had a chance to use threading in their applications. Even After Effects can set up rendering on the second processor.
More importantly, Apple could no doubt optimize it's own applications for use by a quad system.
As to who would buy it... BUILD IT AND THEY WILL COME!
The GPU AGP/PCI-E bandwidth does not impinge on the FSB bandwidth, they are on completely separate busses. Those separate busses will compete for RAM bandwidth at times, which as Powerdoc notes is the limiting bottleneck for now.
This is the second time I've seen this stated, that is that the GPU bandwidth does not impinge on the FSB bandwidth. Now I went looking for my JPEG showing the layout of the busses and of course didn't find it, it any event isn't this a bit of an over simplification?
The first thing that comes to mind is that all FSB traffic goes through the North bridge, with traffic directed to the correct auxilary bus be it PCI, AGP, memory or other hypertransport busses. I would be very surprised if Apple did not talk to the AGP card directly (well via a hop though the bridge) from the processor from time to time. Even if it doesn't happen much now it is very likely to happen more in the future.
I do not see this as the same thing as competeing for RAM bandwidth though that could impact FSB, depending on how much buffering is going on in the bridge chip. Just about any DMA activity that takes place on the bridge is likely to impact transfers to/from memory via the FSB. Besides just how does the GPU talk to memory?
Quote:
Also the bus needs to run at an integer multiple of the core speed. This will drive the tradeoffs on what bus speeds are physically available and practically usable. Sure 1.5Ghz is 3.0/2 but with RAM limitations a 1.5Ghz bus may not be required to handle the max RAM throughput. This may even mean the 1.25Ghz in current machines could be underutilized as it may be driven more by the core speed correspondence rather than bus bandwidth utilization.
Well we do not know how accurate the information is. A 1 GHz FSB may be perfect for a processor with an integrated memory controller. Though I believe many mis one aspect of an IMC that many mis is that of increased FSB usage due to the GPU haveing to get to memory through the FSB.
I tend to think that the big limitation on FSB speed right now is that of PCB technology. As to the bus being underutilized now, that may very well be the case for many applications. But one has to remember the nature of this bus, that is the bandwidth is really half of the stated spec in any one direction. This changes the perspective a bit, it may be far easier to hit transfer limits than one may first realize.
Quote:
e.g.: 2.5/3 ~= .833Ghz which is very close to 800Mhz which has been shown to be "maxable" in bandwidth utilization with non-exotic memory configurations in the Wintel world. Making 2.5/2 a much better choice as it avoids known bandwidth issues at a competitive price point. So where is the cost/throughput tradeoff for faster busses and core speeds? It may be much higher than Apple can market now at a 1.5Ghz bus speed, but not significantly impacted performance-wise at 1.0Ghz. All just informed speculation, but illustrates that bus speed choices are not simple "must be faster or kill performance" decisions.
Yep many decisions go into selecting things such a bus speeds. However transfer rates eventually have to go up to improve performance. As others have pointed out there are options beyond clock rate, one of which would be wider buses. For the speculated chip to have any future though I suspect that bandwidth may be more than that implied by clock rate. It is a given that in a year or two the bnadwidths that we are all so happy about now will be old technology.
The other very real possibility is that the info is just wrong. That is either misquoted or confused by missing information.
Apple didn't need to make a quad single core motherboard. Who would they have marketed it to? They don't have a big enough share of the enterprise community, and they are just laying the groundwork in super-computing where IBM already is well known. I think there was no point in making a quad single core motherboard wit 4 physical chip sockets for 4 physical chips.
I hate being quoted partially when what I said was all supposed to be in one breath. To edit out the middle removes the validity of what I was referring to from what was being said in the previous statement.
What I was referring to was whom AMD, and the Intel quad motherboards were mostly addressing in the marketplace - which I had referenced previously.
That aside I do think wasting the extra space inside a G5 case would have been premature at that time. People may have been interested, but the amount of users needed to succeed at it were not paying as close attention to Apple as a possibility for such hardware. Most people wont come over from the other side unless Apple proves itself with a few consecutive showings of absolute dedication to provide continuous solutions in such a demanding highend marketplace. If Apple were to put fourth efforts to get these users they have to remain committed, but to do that they could not just jump right in with both feet, and say look I'm here. The were smart in taking it gradually. Now most people see them as a possibility as long as they add the key elements needed to do the work in their fields. Such as 3D. I know a lot of people that think Apple had a winner with their OS, and the G5 is an exciting processor with potential, and a strong looking future with IBM. Most 3D users think Apple just needs to be able to provide a Professional level 3D video card, and the G5 just needs to get a over that hump of being one step behind the playing field.
The biggest reason Apple needs to prove it's committed to this is the software switch. It's big chunk of change to move platforms. Most providers will license you a new Mac version if you do, but there is also sometimes a fee to do it, and those fees add up. Nevertheless, Apple is at limbo right now with everybody watching. It's either do it, or not.
Kickaha and Amorph couldn't moderate themselves out of a paper bag. Abdicate responsibility and succumb to idiocy. Two years of letting a member make personal attacks against others, then stepping aside when someone won't put up with it. Not only that but go ahead and shut down my posting priviledges but not the one making the attacks. Not even the common decency to abide by their warning (afer three days of absorbing personal attacks with no mods in sight), just shut my posting down and then say it might happen later if a certian line is crossed. Bullshit flag is flying, I won't abide by lying and coddling of liars who go off-site, create accounts differing in a single letter from my handle with the express purpose to decieve and then claim here that I did it. Everyone be warned, kim kap sol is a lying, deceitful poster.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Perhaps if I was focused on the products but I'm not. I'm strictly concerned with Profits thus the comparison is "Apples to Apples".
Nope, Apples and apricots perhaps, but still two different types of fruit.
You're focused on prices and marketshare. You're analogy is faulty because in the majority of sales Intel beats AMD on price. AMD only retails lower to build your own enthusiasts and small volume manufacturers. Intel is cheaper everywhere where there is any sort of volume. Dell doesn't contract with Intel just to pay more for CPU's or motherboards. Big vendors use Intel because the economies are better -- one stop shopping for CPU, Gfx, Mobo, I/O. Hence, your price argument falls flat.
AMD's/Intel's price to you is no indication of their ability to price out orders for IBM/HP/Dell/Sony. Wanna guess who wins on price in that company?
Onlooker, completely agree with what you said. When the G5 hit, I think most were amazed. I was still recovering from shock a whole week later. I thought, 'finally, Apple are going bleeding edge and staying there...'.
The last updates didn't repeat that emphasis... Disppointing. Even if you take IBM's failure to make 3 gig, (a number I would have thought they would have achieved...or even, at least, get closer to...), you have to ask why a top spec card is bundled in the top spec PowerMac. Still no monitor bundle. No PCI Express. But considering it aint going to mean jot until the Nividi/Ati cards ship in significant volume, have games that can take advantage of it and a rev b of PCI Express that starts to show what this 'next gen' 'AGP' can do. Reviews are decidly muted on what impact PCI Express has right now.
In short, a dual 2.5 G5 and Ati Radeon 9800xt and a Nvidia 6800 dual dual DVI card...are as good as the PC world has right now, eh?
So we can afford to wait for half a year for our bug tweaked and patched versions of Doom 3 and Half Life 2, eh?
Hmmm.
But Onlooker might say that's not the point.
Where's the Wildcat Pro 3D card with dual 256 texture units. Or dual 512 texture units. Where's the token 3D card with a $2000 price on its head? Y'know. Play ball. Look like you're serious about 3D. Couldn't Apple at least license development with creative labs or nividia to have a Quadro Uber on the Apple store? Share some of the development costs? Build to order? Just how different is a Mac graphics card from a PC one anyhow? Endian and all.
At least Maya Unlimited is coming. That's something, right? The mothership is moving in the right direction.
Maybe when the dual 970MP machines hit with Maya Unlimited with an Ati Fire Pro, Quadro or Wildcat available as options we can say the Mac has finally arrived as a Workstation.
I'd like to see it.
I'd also like to see Apple buy Softimage 4.0 and yank the PC version.
That's the one piece missing from Apple's uber software line-up. Heavy duty 3D app.
What I don't understand is why they didn't take the opportunity to buy Maya when Alias Wavey became 'Alias' and was sold for...$26 million (? How much did it go for?)
Pocket change for Apple. Would such an aquisition have driven Newtek, Maxxon etc away? I wouldn't have thought so as 3D users are quite loyal to their package.
Lightwave benches on Apple's page look good with the dual 2.5 handing a dual 3.2 Xeon its ASS.
Hmmm.
If Think Secret says the MP is coming...it's probably more than 50/50 it's coming. There aren't many doubters out there.
I guess I'm a little confused by the arrival of the Power 5 from IBM which seems to hint at a truly scaleable architecture for Powerbooks to PowerMacs. There's rumours of the 975 being right on schedule! Even thought the Fx was late! what does this mean? Ur...I don't know.
Is the 975 (which showed itself on an IBM page in the orient...saw it myself...) going into a new class of Workstation? With the FX derived MP going into convential PowerMacs and Powertuned into Powerbooks and iMacs?
975 Workstation.
970FXMP PowerMacs.
Single Core MP iMacs/PowerBooks.
The FX throws out that much heat then maybe the MP has a better layout to dissapate the heat..? The FX as it stands doesn't seem like the chip to go into PowerBooks and iMacs despite the much ballyhoo about its supposedly better heat temperatures. That doesn't seem to have worked out in practice re: the 2.5 liquid cooled Macs.
If 975 chips are on schedule and ready now...then why weren't they announced? Ramping? IBM and Apple swallowing Fx costs? ie FX development was too far along to back out of without waste, expense etc...and the 975 was still too far from ramp to use?
WILL the 975 arrive in single core Macs and the MP go into workstation macs?
I'm confused by the apparent 'parallel' nature of the new kid of the rumour block, the 'MP' arriving out of nowhere to seemingly surplant all those '975' rumours. We've even seen the 975 on IBM hosted web pages! With nary a mention of this 'MP'.
Unless the MP IS the 975 MP..?
Was the 970 single core rushed to market in light of Motorola's problems?
Maybe the 970 MP is what WAS intended but it needed 0.09 to make it happen? Less of a rush, more time to develop, better chip layout, better process and tech's. Hence the 'swift' transition to 0.09. To get the 'MP' closer. Maybe the 'MP' is the real 'place holder' until the road map for Apple and IBM was established. ie it will keep Apple competitive with single and dual core Wintelians allowing for the road map to come on stream aka the 975 Power 5 derived chip.
Confused? I am. But it's a nice kind of confusion to have...it's just the imminent rumour talk of the 975...maybe the rumour sites are suffering from premature specification and the hidden face/newcomer has come out from under cover of darkness to be revealed as the fx successor..?
This horse has been beat. No need to argue over a flawed argument. I believe that Ford makes a truck that has an engine in it that is made by another company. Forget the fact that Ford makes engines, and look at this argument. I agree that Apple lowering prices has a cost. There are many ways to expand market share. Apple should pursue each path as best they can, that means lowering prices when possible and practical, no need to cut off the nose to spite the face.
iPods & iTunes for windows = a few switchers and more money
Apple stores = a few switchers and more money
Super computer mind share = a few switchers and more money
Buying professional software companies ala Shake and others = a few switchers and more money
Airport extreme type solutions = a few switchers and more money
There are others that I have missed but you get the picture, these "fews" add up, lower prices would also help but the $64,000 question still remains HOW low. Too low and Apple just sold the company for greater market share. NO the way there is slow and steady. Bit by bit.
While this argument is flawed, I agree totally.
Quote:
Originally posted by hmurchison
Cheaper prices= more marketshare is a myth. Take AMD for instance. The "perception" is, with the Athlon64 and Opteron kicking butt, AMD is on a roll but sadly their Q3 earnings tell otherwise. Despite enthusiasm amongst users AMD only turned a 30 million profit...half that of Apple who is expected to vacate the hardware market because they can't compete (so says Rob Enderle). Apple is working for their investors. I want to see more marketshare as well but I don't think it's pricing that dictates it but rather a bevy of many different things.
The high end systems are where the profit is at. In fact I hope Apple's Quad system is $3999. Let the folks deep pockets fund and subsidize those of us without the same.
Comments
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Personally I could care less about the high end, I want to see a PM back at the $1499 or $1599 price, or better yet a replacement for the G4's that they have discontinued starting at $1299. That will do more for Apple's market share than a high end tower or another overpriced iMac.
Cheaper prices= more marketshare is a myth. Take AMD for instance. The "perception" is, with the Athlon64 and Opteron kicking butt, AMD is on a roll but sadly their Q3 earnings tell otherwise. Despite enthusiasm amongst users AMD only turned a 30 million profit...half that of Apple who is expected to vacate the hardware market because they can't compete (so says Rob Enderle). Apple is working for their investors. I want to see more marketshare as well but I don't think it's pricing that dictates it but rather a bevy of many different things.
The high end systems are where the profit is at. In fact I hope Apple's Quad system is $3999. Let the folks deep pockets fund and subsidize those of us without the same.
The PPC 970 MP seems to have one dedicated L2 cache per core. Like air sluf said the interconnect will be at full clock speed (there is no reason to do otherwise), but the wide of the bus is unknown (64 bits, 128 bits ? ...). I guess that the communication between the two cores, will be de facto a communications between the two L2 cache via a controller.
Originally posted by mmmpie
Finally, the G5 has higher clock speeds than what AMD is putting on the market, and you dont hear anyone whining about them not hitting 3ghz. What is that, they have 3200+ cpus, oh, thats the _MADE UP_ number that they use for speed.
1. AMD never promised 3GHz.
2. AMD is the current king of the hill in performance. Their Opterons kill everything in the low-end space -- G5s, P4's, Xeons, etc... No one is complaining about the low clockspeed, because unlike the G5, they are actually faster than the competition.
3. It's not a made-up number. Benchmarks clearly show, especially with the Athlon64 line, that it corresponds well with Intel.
t's not a secret that mac os X, is much more optmized for MP than X86 PC.
Not true.
Originally posted by hmurchison
Cheaper prices= more marketshare is a myth. Take AMD for instance. The "perception" is, with the Athlon64 and Opteron kicking butt, AMD is on a roll but sadly their Q3 earnings tell otherwise. Despite enthusiasm amongst users AMD only turned a 30 million profit...half that of Apple who is expected to vacate the hardware market because they can't compete (so says Rob Enderle). Apple is working for their investors. I want to see more marketshare as well but I don't think it's pricing that dictates it but rather a bevy of many different things.
Apples and oranges.
AMD doesn't sell systems. Apple does. AMD may offer great bang for buck to DIY'ers and small assemblers, but they don't have the ability to produce enough volume to get 1st tier manufacturers and keep them. Those design wins go to intel 9 out of 10 times. And since most people buy their systems from major manufacturers rather than small ones, and even fewer build their own, Intel keeps on mopping the floor with AMD, regardless of the performance or price. I'm willing to bet that Intel even beats AMD on price when it comes to delivering 100's of thousands of parts for large production runs. I'd be surprised in fact, if overall, it wasn't cheaper for companies like Dell, HP, and IBM to turn to Intel for the CPU needs, than to AMD. AMD only retails lower. Intel very agressively cuts prices to major manufactures. You and I can't get those prices, nor can some relatively low volume builder, liek an Alienware, but the guys with the sales (the guys that matter), they're doing just fine buying Intel.
Originally posted by Powerdoc
Concerning the watt/heat issue :
1) The first version of the Athlon 0,13 produced nearly the same heat than the latest generation of Athlon 0,18. Then they introduced some tweaks (and if my memory is correct add a layer) droppind down watt consumption (by a 30 % factor).
The 970 Fx is not a huge benefit in watt consumption compared to the original PPC 970 : we can expect a more efficient design especially if they go in the new 10 layers process.
2) one problem of the PPC 970 FX is the small die : a 2,5 ghz 970 FX approximatively produce the same heat than a 2 ghz PPC 970 but on a much smaller die : thus Apple was obliged to use watercooling to extract heat from a smaller surface of the chip.
The PPC 970 have a larger die, thus heat on the new process will be less concentrated, and thus less difficult to evacuate.
If you consider point 1 and 2, it seems that a dual 970 MP Apple computer is a serious option.
I think we should see again two version of the powermac : the single MP powermac and the dual MP powermac. (two single models, and one dual)
Concerning the speed issue :
Due to the larger Cache (One MB of cache per core) and due to the high speed interconnection of the cores, a single computer 2,5 Ghz 970 MP will be faster than a dual 970 Fx. But (there is always a but), as the MP version is more pipelined, and dispite the improved branch predict unit, a pipeline penalty may occur with this new version.
We will have to wait for the first benchmarks (who will be based upon a new mobo with faster RAM IMO) to know this point. However, this new design will reduce costs, and will allow higher clock speeds.
Originally posted by Brendon
Does anyone have an idea when the PowerPC chips based off of the Power5 series will appear? My guess is that this core would be much better suited to multicore.
The POWER4 and POWER5 are very similar in the respect of multi core. And they are very well suited for it too. IBM probably had a multi core variant of the 970 in mind all along so the 970MP won't be anything like a hack with just two cores put together for good measure.
Originally posted by Brendon
I would think that a 980MP would be easier to do because I thought that IBM was going to make a multicore PowerPC based off of the Power5 series wheather Apple was going to use it or not.
IBM is not relying on Apple for the future of its PowerPC line in general or 970 in perticular. IBM will use them themselves (and I hope in Linux workstations too, not only servers) and besides that we have the likes of Microsoft who'll become a substantially larger customer for these kinds of high performance CPUs than Apple will ever be.
A light version of the POWER5 _is_ in the works, that's certain, but we can oly guess in what time frame. I expect it to come in all kinds of flavours from single core, to mobile, dual core and even more. The Xenon-processor in Xbox NeXT will be tripple core, and it'll be based on POWER5 since it is also hyperthreaded, something that is a new feature of the POWER5.
I wouldn't be surprised if a single core version om GR-UL (aka 975/980/Power5 light) showed up around the same time frame as 970MP.
Apples and oranges.
Perhaps if I was focused on the products but I'm not. I'm strictly concerned with Profits thus the comparison is "Apples to Apples".
I wouldn't be surprised if a single core version om GR-UL (aka 975/980/Power5 light) showed up around the same time frame as 970MP.
Neither would I. I think Apple has some options here on how they want to handle their lineup. I wouldn't mind seeing the P5 derivative ship in 2006 with all the "trimmings"
Internal enhancements
SMT
Dual cores with SMT on each core
beefier Altivec
Ondie Memory Controller
65nm fab
That's worth waiting 18 months for isn't it? The 970MP could take us to 3.2-3.4Ghz in the meantime.
Originally posted by Henriok
The Xenon-processor in Xbox NeXT will be tripple core, and it'll be based on POWER5 since it is also hyperthreaded, something that is a new feature of the POWER5.
This is an unwarranted assumption.
<br />
<br />
<div class="quote-container"><span>Quote:</span>
<div class="quote-block">
<i>Originally posted by Henriok </i><br />
<b>The competition offers quad systems _today_ and have for some time, but Apple is not doing anything about it. If cooling a single core Prescott is hard, then I ccan't imagine how hard it will be to cool a dual core, let alone two or even four dual core chips. </b>
</div>
</div>
<br>The thing about those Quad Opteron, and Quad XEON boxes is that the actual real performance gains on things like 3D, or anything we would find useful are limited. It's great in servers, but only the Sparc from SUN, and it's 4 way SIMD unit in it's processors seems to pack any punch in territory where we could see some performance to justify the additional price of 4 processors. Intel, and AMD have not the same effect. And SUN can not seem to sell squat outside of the military for intelligence research. IBM on the other hand is another animal all together. They are capable, but as I was saying Apple didn't need to make a quad single core motherboard. Who would they have marketed it to? They don't have a big enough share of the enterprise community, and they are just laying the groundwork in super-computing where IBM already is well known. I think there was no point in making a quad single core motherboard wit 4 physical chip sockets for 4 physical chips.<br />
<br />
<div class="quote-container"><span>Quote:</span>
<div class="quote-block">
<i>Originally posted by Henriok </i><br />
<br />
IBM is not relying on Apple for the future of its PowerPC line in general or 970 in perticular. IBM will use them themselves (and I hope in Linux workstations too, not only servers) and besides that we have <b>the likes of Microsoft who'll become a substantially larger customer for these kinds of high performance CPUs than Apple will ever be.<br />
</b>
</div>
</div>
<br>That remains to be seen. Sure M$ will sell a butt load of XBOX's in the US, (and it will probably stop there again) but the team of IBM, and Apple has not shown us if they are actually going to come out, and play with the big boys, and make some more competitive products so Apple can have a true workstation class computer. This also remains to be seen, but once they do Apple will gain ground as will IBM. <br />
<br />
<div class="quote-container"><span>Quote:</span>
<div class="quote-block">
<i>Originally posted by Aphelion </i><br />
<b><br />
I'm tired of Apple trying to attain parity with Wintel (and AMD) systems. Even if you summon up the MHz-Myth and call the 2.5 970FX a "3200+" a-la AMD factoring (some would say "2999+"), it still is too close to call. The "worlds fastest personal computer" should be an unassailable claim.<br />
<br />
Apple needs to set a performance bar that clearly exceeds that of the rest of the personal computer industry. And they need to do it by a wide margin in every category, Not just Photoshop bake-off's, but SPEC scores, and every other conceivable benchmark. Cross platform applications must run significantly faster on the top end Macintosh, even those that are not well optimized for the Mac.<br />
<br />
To limit dual core chips to cheaper duals (or simply more margin) for the flagship towers would be unthinkable to me. A chance to dominate the high end for once must not be missed.<br />
<br />
<br />
... </b>
</div>
</div>
<br><br />
My sentiments are quite similar. They should come out and set the bar much higher anyway, and not lie about the actual speeds. I already know AMD kicks the crap out of a G5 in everything (even photoshop) If the actual performance was a good as WWDC 2003 said it was it would have been great, but it was false information, and bunch of BS false marketing.
Originally posted by Programmer
This is an unwarranted assumption.
It's not totally unwarranted. It's based on this anonymous white paper post that was making headlines in certain circles around the internet.
Xenon Hardware Overview
By Pete Isensee, Development Lead, Xbox Advanced Technology Group
This documentation is an early release of the final documentation, which may be changed substantially prior to final commercial release, and is confidential and proprietary information of MS Corporation. It is disclosed pursuant to a nondisclosure agreement between the recipient and MS.
?Xenon? is the code name for the successor to the Xbox® game console from MS. Xenon is expected to launch in 2005. This white paper is designed to provide a brief overview of the primary hardware features of the console from a game developer?s standpoint.
Caveats
In some cases, sizes, speeds, and other details of the Xenon console have not been finalized. Values not yet finalized are identified with a ?+? sign, indicating that the numbers may be larger than indicated here. At the time of this writing, the final console is many months from entering production. Based on our experience with Xbox, it?s likely that some of this information will change slightly for the final console.
For additional information on various hardware components, see the other relevant white papers.
Hardware Goals
Xenon was designed with the following goals in mind:
?Focus on innovation in silicon, particularly features that game developers need. Although all Xenon hardware components are technologically advanced, the hardware engineering effort has concentrated on digital performance in the CPU and GPU.
?Maximize general purpose processing performance rather than fixed-function hardware. This focus on general purpose processing puts the power into the Xenon software libraries and tools. Rather than being hamstrung by particular hardware designs, software libraries can support the latest and most efficient techniques.
?Eliminate the performance issues of the past. On Xbox, the primary bottlenecks were memory and CPU bandwidth. Xenon does not have these limitations.
Basic Hardware Specifications
Xenon is powered by a 3.5+ GHz IBM PowerPC processor and a 500+ MHz ATI graphics processor. Xenon has 256+ MB of unified memory. Xenon runs a custom operating system based on MS® Windows NT®, similar to the Xbox operating system. The graphics interface is a superset of MS® Direct3D® version 9.0.
CPU
The Xenon CPU is a custom processor based on PowerPC technology. The CPU includes three independent processors (cores) on a single die. Each core runs at 3.5+ GHz. The Xenon CPU can issue two instructions per clock cycle per core. At peak performance, Xenon can issue 21 billion instructions per second.
The Xenon CPU was designed by IBM in close consultation with the Xbox team, leading to a number of revolutionary additions, including a dot product instruction for extremely fast vector math and custom security features built directly into the silicon to prevent piracy and hacking.
Each core has two symmetric hardware threads (SMT), for a total of six hardware threads available to games. Not only does the Xenon CPU include the standard set of PowerPC integer and floating-point registers (one set per hardware thread), the Xenon CPU also includes 128 vector (VMX) registers per hardware thread. This astounding number of registers can drastically improve the speed of common mathematical operations.
Each of the three cores includes a 32-KB L1 instruction cache and a 32-KB L1 data cache. The three cores share a 1-MB L2 cache. The L2 cache can be locked down in segments to improve performance. The L2 cache also has the very unusual feature of being directly readable from the GPU, which allows the GPU to consume geometry and texture data from L2 and main memory simultaneously.
Xenon CPU instructions are exposed to games through compiler intrinsics, allowing developers to access the power of the chip using C language notation.
GPU
The Xenon GPU is a custom 500+ MHz graphics processor from ATI. The shader core has 48 Arithmetic Logic Units (ALUs) that can execute 64 simultaneous threads on groups of 64 vertices or pixels. ALUs are automatically and dynamically assigned to either pixel or vertex processing depending on load. The ALUs can each perform one vector and one scalar operation per clock cycle, for a total of 96 shader operations per clock cycle. Texture loads can be done in parallel to ALU operations. At peak performance, the GPU can issue 48 billion shader operations per second.
The GPU has a peak pixel fill rate of 4+ gigapixels/sec (16 gigasamples/sec with 4× antialiasing). The peak vertex rate is 500+ million vertices/sec. The peak triangle rate is 500+ million triangles/sec. The interesting point about all of these values is that they?re not just theoretical?they are attainable with nontrivial shaders.
Xenon is designed for high-definition output. Included directly on the GPU die is 10+ MB of fast embedded dynamic RAM (EDRAM). A 720p frame buffer fits very nicely here. Larger frame buffers are also possible because of hardware-accelerated_ partitioning and predicated rendering that has little cost other than additional vertex processing. Along with the extremely fast EDRAM, the GPU also includes hardware instructions for alpha blending, z-test, and antialiasing.
The Xenon graphics architecture is a unique design that implements a superset of Direct3D version 9.0. It includes a number of important extensions, including additional compressed texture formats and a flexible tessellation engine. Xenon not only supports high-level shading language (HLSL) model 3.0 for vertex and pixel shaders but also includes advanced shader features well beyond model 3.0. For instance, shaders use 32-bit IEEE floating-point math throughout. Vertex shaders can fetch from textures, and pixel shaders can fetch from vertex streams. Xenon shaders also have the unique ability to directly access main memory, allowing techniques that have never before been possible.
As with Xbox, Xenon will support precompiled push buffers (?command buffers? in Xenon terminology), but to a much greater extent than the Xbox console does. The Xbox team is exposing and documenting the command buffer format so that games are able to harness the GPU much more effectively.
In addition to an extremely powerful GPU, Xenon also includes a very high-quality resize filter. This filter allows consumers to choose whatever output mode they desire. Xenon automatically scales the game?s output buffer to the consumer-chosen resolution.
Memory and Bandwidth
Xenon has 256+ MB of unified memory, equally accessible to both the GPU and CPU. The main memory controller resides on the GPU (the same as in the Xbox architecture). It has 22.4+ GB/sec aggregate bandwidth to RAM, distributed between reads and writes. Aggregate means that the bandwidth may be used for all reading or all writing or any combination of the two. Translated into game performance, the GPU can consume a 512×512×32-bpp texture in only 47 microseconds.
The front side bus (FSB) bandwidth peak is 10.8 GB/sec for reads and 10.8 GB/sec for writes, over 20 times faster than for Xbox. Note that the 22.4+ GB/sec main memory bandwidth is shared between the CPU and GPU. If, for example, the CPU is using 2 GB/sec for reading and 1 GB/sec for writing on the FSB, the GPU has 19.4+ GB/sec available for accessing RAM.
Eight pixels (where each pixel is color plus z = 8 bytes) can be sent to the EDRAM every GPU clock cycle, for an EDRAM write bandwidth of 32 GB/sec. Each of these pixels can be expanded through multisampling to 4 samples, for up to 32 multisampled pixel samples per clock cycle. With alpha blending, z-test, and z-write enabled, this is equivalent to having 256 GB/sec of effective bandwidth! The important thing is that frame buffer bandwidth will never slow down the Xenon GPU.
Audio
The Xenon CPU is a superb processor for audio, particularly with its massive mathematical horsepower and vector register set. The Xenon CPU can process and encode hundreds of audio channels with sophisticated per-voice and global effects, all while using a fraction of the power of a single CPU core.
The Xenon system south bridge also contains a key hardware component for audio?XMA decompression. XMA is the native Xenon compressed audio format, based on the WMA Pro architecture. XMA provides sound quality higher than ADPCM at even better compression ratios, typically 6:1?12:1. The south bridge contains a full silicon implementation of the XMA decompression algorithm, including support for multichannel XMA sources. XMA is processed by the south bridge into standard PCM format in RAM. All other sound processing (sample rate conversion, filtering, effects, mixing, and multispeaker encoding) happens on the Xenon CPU.
The lowest-level Xenon audio software layer is XAudio, a new API designed for optimal digital signal processing. The Xbox Audio Creation Tool (XACT) API from Xbox is also supported, along with new features such as conditional events, improved parameter control, and a more flexible 3D audio model.
Input/Output
As with Xbox, Xenon is designed to be a multiplayer console. It has built-in networking support including an Ethernet 10/100-BaseT port. It supports up to four controllers. From an audio/video standpoint, Xenon will support all the same formats as Xbox, including multiple high-definition formats up through 1080i, plus VGA output.
In order to provide greater flexibility and support a wider variety of attached devices, the Xenon console includes standard USB 2.0 ports. This feature allows the console to potentially host storage devices, cameras, microphones, and other devices.
Storage
The Xenon console is designed around a larger world view of storage than Xbox was. Games will have access to a variety of storage devices, including connected devices (memory units, USB storage) and remote devices (networked PCs, Xbox Live?). At the time of this writing, the decision to include a built-in hard disk in every Xenon console has not been made. If a hard disk is not included in every console, it will certainly be available as an integrated add-on component.
Xenon supports up to two attached memory units (MUs). MUs are connected directly to the console, not to controllers as on Xbox. The initial size of the MUs is 64 MB, although larger MUs may be available in the future. MU throughput is expected to be around 8 MB/sec for reads and 1 MB/sec for writes.
The Xenon game disc drive is a 12× DVD, with an expected outer edge throughput of 16+ MB/sec. Latency is expected to be in the neighborhood of 100 ms. The media format will be similar to Xbox, with approximately 6 GB of usable space on the disk. As on Xbox, media will be stored on a single side in two 3 GB layers.
Industrial Design
The Xenon industrial design process is well under way, but the final look of the box has not been determined. The Xenon console will be smaller than the Xbox console.
The standard Xenon controller will have a look and feel similar to the Xbox controller. The primary changes are the removal of the Black and White buttons and the addition of shoulder buttons. The triggers, thumbsticks, D-pad, and primary buttons are essentially unchanged. The controller will support vibration.
Xenon Development Kit
The Xenon development environment follows the same model as for Xbox. Game development occurs on the PC. The resulting executable image is loaded by the Xenon development kit and remotely debugged on the PC. MS® Visual Studio® version 7.1 continues as the development environment for Xenon.
The Xenon compiler is based on a custom PowerPC back end and the latest MS® Visual C++® front end. The back end uses technology developed at MS for Windows NT on PowerPC. The Xenon software group includes a dedicated team of compiler engineers updating the compiler to support Xenon-specific CPU extensions. This team is also heavily focused on optimization work.
The Xenon development kit will include accurate DVD emulation technology to allow developers to very precisely gauge the effects of the retail console disc drive.
Miscellaneous Xenon Hardware Notes
Some additional notes:
?Xenon is a big-endian system. Both the CPU and GPU process memory in big-endian mode. Games ported from little-endian systems such as the Xbox or PC need to account for this in their game asset pipeline.
?Tapping into the power of the CPU is a daunting task. Writing multithreaded game engines is not trivial. Xenon system software is designed to take advantage of this processing power wherever possible. The Xbox Advanced Technology Group (ATG) is also exploring a variety of techniques for offloading graphics work to the CPU.
?People often ask if Xenon can be backward compatible with Xbox. Although the architecture of the two consoles is quite different, Xenon has the processing power to emulate Xbox. Whether Xenon will be backward compatible involves a variety of factors, not the least of which is the massive development and testing effort required to allow Xbox games run on Xenon.
... "as I was saying Apple didn't need to make a quad single core motherboard. Who would they have marketed it to? ... I think there was no point in making a quad single core motherboard wit 4 physical chip sockets for 4 physical chips."...
A quad Macintosh, whether single or dual core in design, would be welcomed by anyone using Apple's high end video and imaging applications. Logic users might put it to good use as well.
I was most likely early in calling for a quad Mac (around the release of the MERSI compliant Motorola 7410), but it was the capability of the 7410 in a MP environment that made me call for a Wintel killer.
As in all things computer, it's the software that makes things possible, and in 2000 OSX was new, and most likely not quite ready for four processor systems. And applications that could fully use such a system were not mainstream, if they were available at all.
But that was then and this is now. Developers have had a chance to use threading in their applications. Even After Effects can set up rendering on the second processor.
More importantly, Apple could no doubt optimize it's own applications for use by a quad system.
As to who would buy it... BUILD IT AND THEY WILL COME!
...
Originally posted by AirSluf
The GPU AGP/PCI-E bandwidth does not impinge on the FSB bandwidth, they are on completely separate busses. Those separate busses will compete for RAM bandwidth at times, which as Powerdoc notes is the limiting bottleneck for now.
This is the second time I've seen this stated, that is that the GPU bandwidth does not impinge on the FSB bandwidth. Now I went looking for my JPEG showing the layout of the busses and of course didn't find it, it any event isn't this a bit of an over simplification?
The first thing that comes to mind is that all FSB traffic goes through the North bridge, with traffic directed to the correct auxilary bus be it PCI, AGP, memory or other hypertransport busses. I would be very surprised if Apple did not talk to the AGP card directly (well via a hop though the bridge) from the processor from time to time. Even if it doesn't happen much now it is very likely to happen more in the future.
I do not see this as the same thing as competeing for RAM bandwidth though that could impact FSB, depending on how much buffering is going on in the bridge chip. Just about any DMA activity that takes place on the bridge is likely to impact transfers to/from memory via the FSB. Besides just how does the GPU talk to memory?
Also the bus needs to run at an integer multiple of the core speed. This will drive the tradeoffs on what bus speeds are physically available and practically usable. Sure 1.5Ghz is 3.0/2 but with RAM limitations a 1.5Ghz bus may not be required to handle the max RAM throughput. This may even mean the 1.25Ghz in current machines could be underutilized as it may be driven more by the core speed correspondence rather than bus bandwidth utilization.
Well we do not know how accurate the information is. A 1 GHz FSB may be perfect for a processor with an integrated memory controller. Though I believe many mis one aspect of an IMC that many mis is that of increased FSB usage due to the GPU haveing to get to memory through the FSB.
I tend to think that the big limitation on FSB speed right now is that of PCB technology. As to the bus being underutilized now, that may very well be the case for many applications. But one has to remember the nature of this bus, that is the bandwidth is really half of the stated spec in any one direction. This changes the perspective a bit, it may be far easier to hit transfer limits than one may first realize.
e.g.: 2.5/3 ~= .833Ghz which is very close to 800Mhz which has been shown to be "maxable" in bandwidth utilization with non-exotic memory configurations in the Wintel world. Making 2.5/2 a much better choice as it avoids known bandwidth issues at a competitive price point. So where is the cost/throughput tradeoff for faster busses and core speeds? It may be much higher than Apple can market now at a 1.5Ghz bus speed, but not significantly impacted performance-wise at 1.0Ghz. All just informed speculation, but illustrates that bus speed choices are not simple "must be faster or kill performance" decisions.
Yep many decisions go into selecting things such a bus speeds. However transfer rates eventually have to go up to improve performance. As others have pointed out there are options beyond clock rate, one of which would be wider buses. For the speculated chip to have any future though I suspect that bandwidth may be more than that implied by clock rate. It is a given that in a year or two the bnadwidths that we are all so happy about now will be old technology.
The other very real possibility is that the info is just wrong. That is either misquoted or confused by missing information.
Dave
Apple didn't need to make a quad single core motherboard. Who would they have marketed it to? They don't have a big enough share of the enterprise community, and they are just laying the groundwork in super-computing where IBM already is well known. I think there was no point in making a quad single core motherboard wit 4 physical chip sockets for 4 physical chips.
I hate being quoted partially when what I said was all supposed to be in one breath. To edit out the middle removes the validity of what I was referring to from what was being said in the previous statement.
What I was referring to was whom AMD, and the Intel quad motherboards were mostly addressing in the marketplace - which I had referenced previously.
That aside I do think wasting the extra space inside a G5 case would have been premature at that time. People may have been interested, but the amount of users needed to succeed at it were not paying as close attention to Apple as a possibility for such hardware. Most people wont come over from the other side unless Apple proves itself with a few consecutive showings of absolute dedication to provide continuous solutions in such a demanding highend marketplace. If Apple were to put fourth efforts to get these users they have to remain committed, but to do that they could not just jump right in with both feet, and say look I'm here. The were smart in taking it gradually. Now most people see them as a possibility as long as they add the key elements needed to do the work in their fields. Such as 3D. I know a lot of people that think Apple had a winner with their OS, and the G5 is an exciting processor with potential, and a strong looking future with IBM. Most 3D users think Apple just needs to be able to provide a Professional level 3D video card, and the G5 just needs to get a over that hump of being one step behind the playing field.
The biggest reason Apple needs to prove it's committed to this is the software switch. It's big chunk of change to move platforms. Most providers will license you a new Mac version if you do, but there is also sometimes a fee to do it, and those fees add up. Nevertheless, Apple is at limbo right now with everybody watching. It's either do it, or not.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Originally posted by hmurchison
Perhaps if I was focused on the products but I'm not. I'm strictly concerned with Profits thus the comparison is "Apples to Apples".
Nope, Apples and apricots perhaps, but still two different types of fruit.
You're focused on prices and marketshare. You're analogy is faulty because in the majority of sales Intel beats AMD on price. AMD only retails lower to build your own enthusiasts and small volume manufacturers. Intel is cheaper everywhere where there is any sort of volume. Dell doesn't contract with Intel just to pay more for CPU's or motherboards. Big vendors use Intel because the economies are better -- one stop shopping for CPU, Gfx, Mobo, I/O. Hence, your price argument falls flat.
AMD's/Intel's price to you is no indication of their ability to price out orders for IBM/HP/Dell/Sony. Wanna guess who wins on price in that company?
2 Socket 970MP(Quad processing)
1GB of DDR2 500
250 SATA HD 16MB buffer and NCQ on SATA II bus
Nvidia GPU 256MB PCIe
Superdrive (16x DVD-R, 4x DL)
802.11g built in
2 FW 400 and 800 bus each.
4 USB2
Drool.
That's what I want too.
Onlooker, completely agree with what you said. When the G5 hit, I think most were amazed. I was still recovering from shock a whole week later. I thought, 'finally, Apple are going bleeding edge and staying there...'.
The last updates didn't repeat that emphasis... Disppointing. Even if you take IBM's failure to make 3 gig, (a number I would have thought they would have achieved...or even, at least, get closer to...), you have to ask why a top spec card is bundled in the top spec PowerMac. Still no monitor bundle. No PCI Express. But considering it aint going to mean jot until the Nividi/Ati cards ship in significant volume, have games that can take advantage of it and a rev b of PCI Express that starts to show what this 'next gen' 'AGP' can do. Reviews are decidly muted on what impact PCI Express has right now.
In short, a dual 2.5 G5 and Ati Radeon 9800xt and a Nvidia 6800 dual dual DVI card...are as good as the PC world has right now, eh?
So we can afford to wait for half a year for our bug tweaked and patched versions of Doom 3 and Half Life 2, eh?
Hmmm.
But Onlooker might say that's not the point.
Where's the Wildcat Pro 3D card with dual 256 texture units. Or dual 512 texture units. Where's the token 3D card with a $2000 price on its head? Y'know. Play ball. Look like you're serious about 3D. Couldn't Apple at least license development with creative labs or nividia to have a Quadro Uber on the Apple store? Share some of the development costs? Build to order? Just how different is a Mac graphics card from a PC one anyhow? Endian and all.
At least Maya Unlimited is coming. That's something, right? The mothership is moving in the right direction.
Maybe when the dual 970MP machines hit with Maya Unlimited with an Ati Fire Pro, Quadro or Wildcat available as options we can say the Mac has finally arrived as a Workstation.
I'd like to see it.
I'd also like to see Apple buy Softimage 4.0 and yank the PC version.
That's the one piece missing from Apple's uber software line-up. Heavy duty 3D app.
What I don't understand is why they didn't take the opportunity to buy Maya when Alias Wavey became 'Alias' and was sold for...$26 million (? How much did it go for?)
Pocket change for Apple. Would such an aquisition have driven Newtek, Maxxon etc away? I wouldn't have thought so as 3D users are quite loyal to their package.
Lightwave benches on Apple's page look good with the dual 2.5 handing a dual 3.2 Xeon its ASS.
Hmmm.
If Think Secret says the MP is coming...it's probably more than 50/50 it's coming. There aren't many doubters out there.
I guess I'm a little confused by the arrival of the Power 5 from IBM which seems to hint at a truly scaleable architecture for Powerbooks to PowerMacs. There's rumours of the 975 being right on schedule! Even thought the Fx was late! what does this mean? Ur...I don't know.
Is the 975 (which showed itself on an IBM page in the orient...saw it myself...) going into a new class of Workstation? With the FX derived MP going into convential PowerMacs and Powertuned into Powerbooks and iMacs?
975 Workstation.
970FXMP PowerMacs.
Single Core MP iMacs/PowerBooks.
The FX throws out that much heat then maybe the MP has a better layout to dissapate the heat..? The FX as it stands doesn't seem like the chip to go into PowerBooks and iMacs despite the much ballyhoo about its supposedly better heat temperatures. That doesn't seem to have worked out in practice re: the 2.5 liquid cooled Macs.
If 975 chips are on schedule and ready now...then why weren't they announced? Ramping? IBM and Apple swallowing Fx costs? ie FX development was too far along to back out of without waste, expense etc...and the 975 was still too far from ramp to use?
WILL the 975 arrive in single core Macs and the MP go into workstation macs?
I'm confused by the apparent 'parallel' nature of the new kid of the rumour block, the 'MP' arriving out of nowhere to seemingly surplant all those '975' rumours. We've even seen the 975 on IBM hosted web pages! With nary a mention of this 'MP'.
Unless the MP IS the 975 MP..?
Was the 970 single core rushed to market in light of Motorola's problems?
Maybe the 970 MP is what WAS intended but it needed 0.09 to make it happen? Less of a rush, more time to develop, better chip layout, better process and tech's. Hence the 'swift' transition to 0.09. To get the 'MP' closer. Maybe the 'MP' is the real 'place holder' until the road map for Apple and IBM was established. ie it will keep Apple competitive with single and dual core Wintelians allowing for the road map to come on stream aka the 975 Power 5 derived chip.
Confused? I am. But it's a nice kind of confusion to have...it's just the imminent rumour talk of the 975...maybe the rumour sites are suffering from premature specification and the hidden face/newcomer has come out from under cover of darkness to be revealed as the fx successor..?
Lemon Bon Bon
iPods & iTunes for windows = a few switchers and more money
Apple stores = a few switchers and more money
Super computer mind share = a few switchers and more money
Buying professional software companies ala Shake and others = a few switchers and more money
Airport extreme type solutions = a few switchers and more money
There are others that I have missed but you get the picture, these "fews" add up, lower prices would also help but the $64,000 question still remains HOW low. Too low and Apple just sold the company for greater market share. NO the way there is slow and steady. Bit by bit.
While this argument is flawed, I agree totally.
Originally posted by hmurchison
Cheaper prices= more marketshare is a myth. Take AMD for instance. The "perception" is, with the Athlon64 and Opteron kicking butt, AMD is on a roll but sadly their Q3 earnings tell otherwise. Despite enthusiasm amongst users AMD only turned a 30 million profit...half that of Apple who is expected to vacate the hardware market because they can't compete (so says Rob Enderle). Apple is working for their investors. I want to see more marketshare as well but I don't think it's pricing that dictates it but rather a bevy of many different things.
The high end systems are where the profit is at. In fact I hope Apple's Quad system is $3999. Let the folks deep pockets fund and subsidize those of us without the same.