2500K vs 1100T

This is true mate, however we do it all the time with graphics cards. Different number of cores, memory bandwidth and overclock ability and we make a decision between two products, be it nvidia, amd or intracompany. Also its not that hard to compare, the video does that.

The point here is that they would be the two i would imagine that would be considered if a current person was building a gaming rig no?

I think what is important here is that yes, using mid range cards there will be little difference in gamin, but in a couple of years, the person that doesn't want to spend money on a new computer every couple of years, can upgrade the GPU without bottling the CPU, the 1100T is being bottled on what may be a mid range gpu in a few years. The Intel system wont be as much, plus you have an upgrade path - and we haven't even discussed overclocking yet ;). Happy to discuss this as I think it should factor in peoples minds.
 
Last edited:
I hate the word bottleneck.

In games that rely heavily on CPU power (ie GTA IV, ARMA series, etc) you will have overall lower performance if the program or game in question is not coded for more than 2/3/4 cores. That's not a CPU bottlenecking a GPU, it's a CPU bottleneck. The GPU has nothing to do with it. In games that aren't as intensive on the CPU you should have near equal performance.

PS How do you like your G500? Do you use the weights?
 
Well the GPUs are so powerful to remove that as the bottleneck to get closer to properly testing and isolating the CPU.

Yeah the 500 is good, i use light weights.
 
I liked that video.It was a good idea to remove the GPU bottleneck by using 2 GTX 590.
And as spynoodle said that with a mid range GPU with Fear 3 the difference was pretty low.
But I would have also(not instead) liked to see other Games benchmarked using a mid-range GPU thus simulating a practical computer.

On a side note Fear 3 is poorly coded.



PS:I never got this saying "Food for thought" shouldnt it be "Food for mind(or something)".
 
spynoodle said:
Guys, you might kinda be overreacting a bit. Maybe bigfella is just subscribed to this dude on youtube. I find it to be a pretty interesting comparison, actually. You don't notice much of a difference between CPUs when using a mid-range GPU, but with a GTX 590 the difference really shows. It's food for thought.
Click to expand...

The video card made 0 difference, he managed to muck up the drivers on the AMD system, as he mentioned in the video and the description. There is no way a single HD6950 should be beating SLIed 590's

bigfellla said:
Wow talk about thread hijacking and trolling. If it were me, i probably wouldve had three infractions ;) :)

There are plenty of posts about 2600K vs 2500K, and I believe the 2600K is a better wafer, has its benefits and I can afford it, so i went with it


Exactly, thank you.

Back on topic, I am interested in talking about the architectural differences between AMD's K10 and SB. The 1100T is the flagship product for AMD currently, and arguably the 2500K is the same for Intel in terms of gaming. So this imho is a valid topic.
Click to expand...

Because the mod/admin team are so hard on you, right ;)

I disagree that the 1100T is the flagship for gaming. As has been stated to you several times before, games do not utilise all of the cores, mainly because they are either ported from consoles or also designed for consoles, so rather than fully rewrite to be completely multi-threaded for either quad core processors or up to 8 threads (quad + hyperthreading) they only utilise 3, because of the Xbox having a tri-core CPU.

All that the video and any other comparison using games shows is a comparison of clock-for-clock performance, not the chip when you are comparing a current gen quad to a 3 generation old hex and only 3 cores are used on each.

Go and find some true multi-threaded tests with both chips and then look at the results. I am not saying that the Thuban will beat SB in all of the tests, but you won't see what are extremely skewed results that don't show the true performance of the CPU's.

I see 0 reason for any mainstream user, as in one that doesn't perform specialist tasks regularly that will utilise all of the cores, to get a hex core CPU, so that makes the Thuban chips there for a very niche market, for the gimmick of 6 cores and for marketing.

I would, if I had the same amount of money for a Thuban chip, take an Intel of the same price, because as someone that doesn't sit with multi-threaded apps all of the time, the Intel would give better performance, there isn't any more comparison to be made until developers stop being lazy
 
Ok Aastii, which two cpus would be apples and apples? As both CPUs have 2 cores available (with 2 - 4 spare) for the game. The 1100T is the flagship product as it is the best available right? Why not compare it.

Secondly, we are comparing games in their current state. Regardless of limitations. Also, many people are interested in how the CPU will perform in future when the GPU is less of the limiting factor. This is clearly the interest in this topic.

If not, please provide a AMD chip we can compare it with. You're point that there cannot be a comparison may be techincally valid, but in the real world when people are buying a set up, they want to know what will be faster. If you are an enthusiast then clearly this is a valid argument, even with the threading limitations. The funny thing is that even in multithreaded applications (video editing, etc) the 1100T still doesn't shine. http://www.bit-tech.net/hardware/cpus/2011/01/03/intel-sandy-bridge-review/8 Why is that?

But this is not what this thread is about. Its about the differences in real world comparisons in gaming where you (as much as possible) test the cpu. Thats the interesting bit for me at least. But yeah, Aastii, your points are (as usual) valid and interesting.
 
Last edited:
I would compare it to a 955 or 965. Same clock, same number of cores (even if 1 or 2 may be redundant), but as with this test, there would be 0 surprises, clearly the Intel would win because of the newer architecture.

The only people that would find the results of the video surprising are those that don't know much at all about current technology, and those that don't and do a little searching will find in a relatively short amount of time that a hex core is completely pointless if gaming is the primary purpose.

I understand that it was done to try and create similar price points from each retailer and compare their "top of the line" CPUs, but it would have been easy to guess before even seeing the results what the outcome would be
 
Yeah, you're probably right, but at the end of the day, thats true about almost everything technical.
 
The whole point is AMD need to wakeup and make a CPU can beats Intel SB or even IVY when it comes out.
Amd is $6.50 and Intel is $21. that is the whole point. Should have bought more that fat cat instead the loussy AMD.
Cheers.
 
Aastii said:
The video cards perform the same regardless of which manufacturer of CPU you have. AMD/AMD, Intel/AMD, Intel/Nvidia, Intel/AMD, all the same
Click to expand...

AMD/Nvidia...Sorry, not the kind to just let typos go. Haha.

To be fair, yeah the phenom ii x4 is a lot closer to the actual infrastructure of the i5, but still, it kind of goes to show how fair intel has come compared to AMD (granted, AMD is always a few months behind intel, but I like it that way). I wouldn't see how comparing it to the 965 would make any sort of difference. The clock speed is hardly different, and even though the 1100T has more cores, in the bench tests that linus is running, it's not like those extra cores are even being utilized, so he might as well have used a 955/965.
 
paulcheung said:
The whole point is AMD need to wakeup and make a CPU can beats Intel SB or even IVY when it comes out.
Amd is $6.50 and Intel is $21. that is the whole point. Should have bought more that fat cat instead the loussy AMD.
Cheers.
Click to expand...

That's not their strategy at the moment.

They seem to be hung up on being better value for money and not being the performance kings. We'll see how that changes with Bulldozer.
 
Troncoso said:
AMD/Nvidia...Sorry, not the kind to just let typos go. Haha.

To be fair, yeah the phenom ii x4 is a lot closer to the actual infrastructure of the i5, but still, it kind of goes to show how fair intel has come compared to AMD (granted, AMD is always a few months behind intel, but I like it that way). I wouldn't see how comparing it to the 965 would make any sort of difference. The clock speed is hardly different, and even though the 1100T has more cores, in the bench tests that linus is running, it's not like those extra cores are even being utilized, so he might as well have used a 955/965.
Click to expand...

AMD aren't a few months behind, they are a few years behind. Phenom II is essentially a tweaked Athlon 64, they haven't brought out a brand new architecture for years.

and what you said is exactly the point - it wouldn't make any difference because you aren't using the full potential of the Thuban CPU. It is all well and good saying let's look at gaming performance, but when the chip isn't really designed with gaming in mind because games aren't truely completely mult-threaded, who cares, it proves absolutely nothing and is a completely stupid and pointless test
 
Its not pointless. It just doesn't agree with your point of view.

The point is, where a GPU isn't the limiting factor. Essentially what you are saying until games can use 6 cores we cant compare. thats ridiculous. Its totally a fair comparison. If i want a new cpu and motherboard, my choices are 2500K or 1100T being the best for gaming. And the comparison is clear.
 
bigfellla said:
Its not pointless. It just doesn't agree with your point of view.

The point is, where a GPU isn't the limiting factor. Essentially what you are saying until games can use 6 cores we cant compare. thats ridiculous. Its totally a fair comparison. If i want a new cpu and motherboard, my choices are 2500K or 1100T being the best for gaming. And the comparison is clear.
Click to expand...

I am not saying to compare for gaming is stupid, I am saying it is a completely pointless test to pit a hex core, which only has 3 cores being used, against a quad and say 2 extra cores and it is still not beating it. Well go for full multi-threaded apps, which the Thuban is designed for, then go and look. It is common knowledge games won't use 6 cores or even 4 (for the most part), it is common knowledge it is the same core as all other Phenom II's, therefore same performance per core, so you may as well just look at the results you already have for your 955/965.

It isn't my point of view, it is a test that serves 0 purpose. If you didn't know what the outcome would be before watching the video, you haven't been keeping up with technology. It isn't a suprise, it isn't "omg, look how great the Intel chips are, even with 2 extra cores the AMD is getting destroyed!!" because the extra cores aren't getting touched.

If you are getting the hex core for gaming, you are buying into the gimmick and spending more just because it says "extra cores, therefore better". The information is there to explain why, the information is there to predict the results before even doing the benches so is a useless test
 
I will agree that it's kinda stupid that the guy was pairing dual GTX 590s with two mid-range CPUs. I'm sure the difference between the CPUs would've proved to be much smaller if he had used, say a single GTX 570. Granted, the video proves that, in theory, the 2500k provides a better upgrade path GPU-wise before you start to see a bottleneck.

...of course, if you buy a bulldozer-upgradable motherboard, then you won't have that problem. There's a decent debate on both sides of this argument IMO.
 
You must log in or register to reply here.
Back
Top