Is the Xenon better than current PC dual-cores?

The Xenon runs faster than most Dual-cores that I have seen (unless they are overclocking) and it has 3 cores as opposed to 2. Shouldn't the Xenon be leaps ahead of Core 2 Duos? I know the GPU has been surpassed by the newest video cards for PCs, but the CPU of the 360 looks like something still ahead of PCs. Is the Xenon better than PC CPUs?

the cell is better then core duodeadmeat59

No, clock-speed and number of cores are not genuine indicators of performance. The Core 2 Duo, running at a slower clock speed, with fewer cores, still outperforms the Xenon by a wide margin. There are a large number of factors that determine processor power.

Short answer - architecture

Medium answer- work per-clock cycle

Long answer - no

[QUOTE="deadmeat59"]the cell is better then core duoVandalvideo

No, clock-speed and number of cores are not genuine indicators of performance. The Core 2 Duo, running at a slower clock speed, with fewer cores, still outperforms the Xenon by a wide margin.subrosian

I see. Interesting.

[QUOTE="deadmeat59"] at raw computing cell has 7 spe at 3.2 ghz it can do 2.2 taraflops . but for games it has yet to show it can pass even the 360.Vandalvideo

Fanboy. I own a Sony Toaster and it's eventually going to make superior toast:x Just you wait...

No, clock-speed and number of cores are not genuine indicators of performance. The Core 2 Duo, running at a slower clock speed, with fewer cores, still outperforms the Xenon by a wide margin. There are a large number of factors that determine processor power.

Short answer - architecture

Medium answer- work per-clock cycle

Long answer - no

subrosian

what does architecture entail, and how is that a difference between the Xenon and the Core 2 Duo?

Does architecture include cache? What are standard memory caches on PCs versus the Xenon's cache?

Being that it's different archetecture, consoles run in-order processing, while PC's have out-of-order processing, it's hard to compare. Both work in different ways, as you can't run artificial benchmarks on consoles for comparison, but if I had to guess, the Xenon is already pretty dated. It's cores don't have the power of each core on a quade core on PC, and are also utilized differently. For the most part, developers are able to utilize the xenon more for it's gaming potential, whereas as PC processors aren't neccessarily used all the way, therefore you have to have a more powerful PC to make up the difference, so naturally, I would imagine there's much more horsepower in a C2D 6800 than a Xenon. While the Xenon is a decent chip for what it is, it's a few years old at this point, and compared to new C2D's and Quade Cores from both AMD and Intel, there's not much of a match. Furthermore, Intel and AMD's new chips are much more effecient with their 65nm and 45nm manufacturing process, using far less power than they did a generation ago.

For certain math calculations I know the PS3's chip is extremely powerful, but when it comes to other operations, it falls short. To play it safe, I would never ever bet against Intel or AMD's latest chips, as they keep gettng more and more powerful, and smaller and smaller.

[QUOTE="deadmeat59"] at raw computing cell has 7 spe at 3.2 ghz it can do 2.2 taraflops . but for games it has yet to show it can pass even the 360.Vandalvideo

Please, don't turn this into another drawn out Cell arguement. :cry:I only want to know about PC vs Xenon.

Since video game systems are starting to get compared more to PCs somebody needs to release some benchmark software that can be used to benchmark your game console.

And list its comparions to computers running a Intel or AMD CPU.

what does architecture entail, and how is that a difference between the Xenon and the Core 2 Duo?

Does architecture include cache? What are standard memory caches on PCs versus the Xenon's cache?

Mordred19

Think of architecture, just like you would with a highway. A highway architected to be a Inter-state highway will have less exits per mile, if you were to compare it to a regular road. So for the purpose going from Boston to New York, it would be best to take a interstate highway. However, if you wanted to get from one town to a neighboring town, you would take the backroads.

Computer architecture is sorta like that, except its on a bigger scale. Instead of representing just one highway or road, it more closely represents an entire transportation system. For example, New England's transportation system versus California transportion system. Which is better for commercial traffic? Which is better for commuter traffic? Which is the best for quick short distance drive? long distance drives? etc.

In the same sense. Just because New England has more roads, and more exits per mile of road -- California may still be better / for efficient.

I don't know if this helps, but I'm trying to abstract it.

[QUOTE="subrosian"]

No, clock-speed and number of cores are not genuine indicators of performance. The Core 2 Duo, running at a slower clock speed, with fewer cores, still outperforms the Xenon by a wide margin. There are a large number of factors that determine processor power.

Short answer - architecture

Medium answer- work per-clock cycle

Long answer - no

Mordred19

what does architecture entail, and how is that a difference between the Xenon and the Core 2 Duo?

Does architecture include cache? What are standard memory caches on PCs versus the Xenon's cache?

its more about things like how BOTH the Cell and Xenon use In-order processing. That CAN be useful, particularly for raw computation as it reduces frills like reordering the computations after they are completed, but during games, Out-of-order is much better. Also, for multi-tasking like most PC's do, Out-of-order is much better and thats why out of order processors are the standard. Also, the there are a number of other factors that are too complicated to completely explainwithout going on a long rant that also contribute. Out-of-order processingis a major contibutor though.

[QUOTE="Mordred19"]

what does architecture entail, and how is that a difference between the Xenon and the Core 2 Duo?

Does architecture include cache? What are standard memory caches on PCs versus the Xenon's cache?

rimnet00

Think of architecture, just like you would with a highway. A highway architected to be a Inter-state highway will have less exits per mile, if you were to compare it to a regular road. So for the purpose going from Boston to New York, it would be best to take a interstate highway. However, if you wanted to get from one town to a neighboring town, you would take the backroads.

Computer architecture is sorta like that, except its on a bigger scale. Instead of representing just one highway or road, it more closely represents an entire transportation system. For example, New England's transportation system versus California transportion system. Which is better for commercial traffic? Which is better for commuter traffic? Which is the best for quick short distance drive? long distance drives? etc.

In the same sense. Just because New England has more roads, and more exits per mile of road -- California may still be better / for efficient.

I don't know if this helps, but I'm trying to abstract it.

Okay, I think I see what you mean.

[QUOTE="subrosian"]

No, clock-speed and number of cores are not genuine indicators of performance. The Core 2 Duo, running at a slower clock speed, with fewer cores, still outperforms the Xenon by a wide margin. There are a large number of factors that determine processor power.

Short answer - architecture

Medium answer- work per-clock cycle

Long answer - no

Mordred19

what does architecture entail, and how is that a difference between the Xenon and the Core 2 Duo?

Does architecture include cache? What are standard memory caches on PCs versus the Xenon's cache?

Architecture includes a lot of things, cache (which is just a fancy way of saying "super fast memory attached to the processor"), internal bandwidth (how many lanes is the "highway" connecting the different parts of the processor? how fast are those lanes? how do they handle traffic?), external bandwidth, et cetera.

Architecture also includes how basic commands "work". If you think of a task for a processor to perform, say "add 2 + 2 and output the result", it has to break those down into binary operations. In the real world, that means breaking that down something as simple as "add 2 plus 2" into lots of little operations like "take the value "2" and store it in cache", "move the value in the cache into the ALU (the CPU's calculator)".

So, how powerful a processor is depends also on how well it performs those tasks.

On top of that (not to make this more complicated), a processor receives requests to perform many tasks at once, often with one task depending on the result of another task (think of a math problem like "2 + (2 * 4)" you need to calculate that 2 * 4 = 8 before you can add it to the value of 2). So, a processor's "architecture" needs to have an efficient scheduler - basically the processor's air traffic controller. Not only does it need to make sure the tasks happen in the right order, it needs to *anticipate* upcoming tasks, and make sure as much of the processor's "power" is in use at once.

-

I know it's a bit complicated, but basically "clock speed" is simply something used to "sync" the tasks, actually the same way we use clocks. If you told your friend to meet you at Starbucks at 7pm to work on some homework with you, you have a far better chance of actually meeting up to work than if you say "I'll meet up to work on the homework at Starbucks whenever". The "clock" in a CPU works the same way - if your friend were to wear a watch that "ticks" twice as often as yours, does that give him more time in the day? Of course not.

Cores are a different issue. What we do with "cores" is create distinct, seperate processors, group them together, and use some form of scheduling (either software or hardware) to divide work amongst them. It actually isn't too much different than the traditional CPU in that we think in the same way - efficient scheduling, and efficiency matter. The only thing that's changed is that there is now a software-side to it - the people building the games have to think about how to use those seperate cores efficiently.

However, the "power" of the total processor still depends on how good each core is, so, a Core 2 Duo, which has two very-efficient, very-powerful cores in the CPU is stronger than the Xenon, which has three older, less-powerful, less-efficient cores.

[QUOTE="Mordred19"][QUOTE="subrosian"]

No, clock-speed and number of cores are not genuine indicators of performance. The Core 2 Duo, running at a slower clock speed, with fewer cores, still outperforms the Xenon by a wide margin. There are a large number of factors that determine processor power.

Short answer - architecture

Medium answer- work per-clock cycle

Long answer - no

subrosian

what does architecture entail, and how is that a difference between the Xenon and the Core 2 Duo?

Does architecture include cache? What are standard memory caches on PCs versus the Xenon's cache?

Architecture includes a lot of things, cache (which is just a fancy way of saying "super fast memory attached to the processor"), internal bandwidth (how many lanes is the "highway" connecting the different parts of the processor? how fast are those lanes? how do they handle traffic?), external bandwidth, et cetera.

Architecture also includes how basic commands "work". If you think of a task for a processor to perform, say "add 2 + 2 and output the result", it has to break those down into binary operations. In the real world, that means breaking that down something as simple as "add 2 plus 2" into lots of little operations like "take the value "2" and store it in cache", "move the value in the cache into the ALU (the CPU's calculator)".

So, how powerful a processor is depends also on how well it performs those tasks.

On top of that (not to make this more complicated), a processor receives requests to perform many tasks at once, often with one task depending on the result of another task (think of a math problem like "2 + (2 * 4)" you need to calculate that 2 * 4 = 8 before you can add it to the value of 2). So, a processor's "architecture" needs to have an efficient scheduler - basically the processor's air traffic controller. Not only does it need to make sure the tasks happen in the right order, it needs to *anticipate* upcoming tasks, and make sure as much of the processor's "power" is in use at once.

-

I know it's a bit complicated, but basically "clock speed" is simply something used to "sync" the tasks, actually the same way we use clocks. If you told your friend to meet you at Starbucks at 7pm to work on some homework with you, you have a far better chance of actually meeting up to work than if you say "I'll meet up to work on the homework at Starbucks whenever". The "clock" in a CPU works the same way - if your friend were to wear a watch that "ticks" twice as often as yours, does that give him more time in the day? Of course not.

Cores are a different issue. What we do with "cores" is create distinct, seperate processors, group them together, and use some form of scheduling (either software or hardware) to divide work amongst them. It actually isn't too much different than the traditional CPU in that we think in the same way - efficient scheduling, and efficiency matter. The only thing that's changed is that there is now a software-side to it - the people building the games have to think about how to use those seperate cores efficiently.

However, the "power" of the total processor still depends on how good each core is, so, a Core 2 Duo, which has two very-efficient, very-powerful cores in the CPU is stronger than the Xenon, which has three older, less-powerful, less-efficient cores.

Well, you definetly cleared things up for me. Thanks. :D

The Xenon runs faster than most Dual-cores that I have seen (unless they are overclocking) and it has 3 cores as opposed to 2. Shouldn't the Xenon be leaps ahead of Core 2 Duos? I know the GPU has been surpassed by the newest video cards for PCs, but the CPU of the 360 looks like something still ahead of PCs. Is the Xenon better than PC CPUs?

Mordred19

[QUOTE="Mordred19"]

The Xenon runs faster than most Dual-cores that I have seen (unless they are overclocking) and it has 3 cores as opposed to 2. Shouldn't the Xenon be leaps ahead of Core 2 Duos? I know the GPU has been surpassed by the newest video cards for PCs, but the CPU of the 360 looks like something still ahead of PCs. Is the Xenon better than PC CPUs?

LibertySaint

you have no idea what you are talking about....a core 2 duo, is far more powerfull than a xenon.

[QUOTE="Mordred19"]

The Xenon runs faster than most Dual-cores that I have seen (unless they are overclocking) and it has 3 cores as opposed to 2. Shouldn't the Xenon be leaps ahead of Core 2 Duos? I know the GPU has been surpassed by the newest video cards for PCs, but the CPU of the 360 looks like something still ahead of PCs. Is the Xenon better than PC CPUs?

LibertySaint

Please stop spreading misinformation, read this page for further info.

[QUOTE="LibertySaint"][QUOTE="Mordred19"]

The Xenon runs faster than most Dual-cores that I have seen (unless they are overclocking) and it has 3 cores as opposed to 2. Shouldn't the Xenon be leaps ahead of Core 2 Duos? I know the GPU has been surpassed by the newest video cards for PCs, but the CPU of the 360 looks like something still ahead of PCs. Is the Xenon better than PC CPUs?

mismajor99

Please stop spreading misinformation, read this page for further info.

how is what i said any diffrent, i just said specs instead of posting techinal information on cpus....lol

No, not better than even the poorest of Pentium Ds.RichterBelmont7

It's not that weak :|

[QUOTE="RichterBelmont7"]No, not better than even the poorest of Pentium Ds.thirstychainsaw

It's not that weak :|

i believe it was sarcasmn because pent d's are like the lowest of anything...heck they have hard time with i-tunes.

[QUOTE="Vandalvideo"][QUOTE="deadmeat59"]the cell is better then core duodeadmeat59

The Cell is a Single-precision floating point beast..but its not as good as the Core 2 Duo, even if it excels beyond all else in a couple of areas.

[QUOTE="Mordred19"]

The Xenon runs faster than most Dual-cores that I have seen (unless they are overclocking) and it has 3 cores as opposed to 2. Shouldn't the Xenon be leaps ahead of Core 2 Duos? I know the GPU has been surpassed by the newest video cards for PCs, but the CPU of the 360 looks like something still ahead of PCs. Is the Xenon better than PC CPUs?

LibertySaint

If the details were the same, as is you are utterly wrong. Intel's Netburst signaled the end of the era where you could just CPUs by clockspeed.

[QUOTE="RichterBelmont7"]No, not better than even the poorest of Pentium Ds.thirstychainsaw

It's not that weak :|

It is, I'm sure its stronger than them in some ways but as far as the average kind of power you can get over the entire range of tasks its weaker.

Xenon and Core 2 are designed for very different purposes, run very different types of code, have very different design constraints, and of course come complete with a very different price tag. In this respect, how do you go about saying one is "better" than the other? What does "better" even mean in this case? Better for games? Okay, which games? What are these games doing? How were they coded? You can see how quickly this sort of thing breaks down, I hope.


Teufelhuhn

For some 360 owners more quickly than others... *coughs*

... oh and I agree, it's weird to see people talk about "the cell" as though it's meant to be compared to a general purpose PC processor...

Xenon and Core 2 are designed for very different purposes, run very different types of code, have very different design constraints, and of course come complete with a very different price tag. In this respect, how do you go about saying one is "better" than the other? What does "better" even mean in this case? Better for games? Okay, which games? What are these games doing? How were they coded? You can see how quickly this sort of thing breaks down, I hope.


Teufelhuhn

Yeah, I nowsee it's not a clear-cut comparison.

[QUOTE="Vandalvideo"][QUOTE="deadmeat59"] at raw computing cell has 7 spe at 3.2 ghz it can do 2.2 taraflops . but for games it has yet to show it can pass even the 360.rimnet00

Fanboy. I own a Sony Toaster and it's eventually going to make superior toast:x Just you wait...

I remember sony saying that one day your toaster will be able to make your PS3 more powerful haha, the sad thing is, im not lying.

ugggh. Zomg teh more cores down not equal more powerful, nor does TEH HIGH CLOCK SPEEDS. Its all about architecture.Vandalvideo

[QUOTE="Vandalvideo"][QUOTE="deadmeat59"]the cell is better then core duodeadmeat59