Sandy Bridge-E Benchmarks Leaked: Disappointing Gaming Performance?

[Wile E]

I do not see a 2500k in those tests.

And the words "in these forums" was never once mentioned prior to this.

 

[John Doe]

I do not see a 2500k in those tests.

And the words "in these forums" was never once mentioned prior to this.

You don't have to. Point is, Sandy is faster as long as you don't load it %100. It's faster per-clock, can OC more and does it under cooler temps.

I don't need to say that either. Just because you can afford a $1000 CPU, and have use for it, doesn't mean everyone else does. 2500k has the best balance right now, and, is the CPU to get. End of story.

 

[Wile E]

You don't have to. Point is, Sandy is faster as long as you don't load it %100. It's faster per-clock, can OC more and does it under cooler temps.

I don't need to say that either. Just because you can afford a $1000 CPU, and have use for it, doesn't mean everyone else does. 2500k has the best balance right now, and, is the CPU to get. End of story.

End of whose story? Certainly not mine. The fact is, you cannot speak for everybody, and gaming performance does not represent a cpu's true potential.

 

[John Doe]

End of whose story? Certainly not mine. The fact is, you cannot speak for everybody, and gaming performance does not represent a cpu's true potential.

That's right, I can't speak for everybody. However I can speak for majority. You, on the other hand, is speaking for minority. How many enthuasiast buy their systems mainly for decoding? They certainly aren't as much as people that build for gaming.

 

[n-ster]

That's right, I can't speak for everybody. However I can speak for majority. You, on the other hand, is speaking for minority. How many enthuasiast buy their systems mainly for decoding? They certainly aren't as much as people that build for gaming.

except the majority of the 2011 market is servers?

 

[John Doe]

except the majority of the 2011 market is servers?

Except this forum or people on similar forums aren't in the corporate market.

Except you're asking me that when you know what we're actually on about.

 

[n-ster]

k fine, people who will buy 2011 will buy it for the extra features more than for the extra performance

 

[John Doe]

k fine, people who will buy 2011 will buy it for the extra features more than for the extra performance

Thing is, this platform didn't deliver those as well. It was said to, but it didn't. That's why it doesn't look great. It lacks USB 3, PCI-E 3.0 (which currently doesn't have use) and so.

http://www.tomshardware.com/reviews/core-i7-3960x-x79-performance,3026.html

 

[Over_Lord]

I don't see why anyone even needs to buy anything above Core i7 2600k for more than 90% gamers and otherwise.

No point really spending more than 300$ when it performs so admirably well, and can be overclocked to beat the crap out of additional core processors.

 

[n-ster]

Thing is, this platform didn't deliver those as well. It was said to, but it didn't. That's why it doesn't look great. It lacks USB 3, PCI-E 3.0 (which currently doesn't have use) and so.

http://www.tomshardware.com/reviews/core-i7-3960x-x79-performance,3026.html

I'd hardly trust TH to know this for sure... It is said it will be an unlockable feature in mobos maybe... and there are already 4 USB 3 ports from the controllers which isn't bad

 

[John Doe]

I don't see why anyone even needs to buy anything above Core i7 2600k for more than 90% gamers and otherwise.

No point really spending more than 300$ when it performs so admirably well, and can be overclocked to beat the crap out of additional core processors.

Yeah, the 2500k is even better for just games. If you need HT for multi-threaded work, 2700k happens to sell $10 more something. 2700k's seem to do over 5 Ghz easier than the 2600k does.

I'd hardly trust TH to know this for sure... It is said it will be an unlockable feature in mobos maybe... and there are already 4 USB 3 ports from the controllers which isn't bad

Neither do I trust TH most the time, but you can't deny that they "got" the hardware. One can say it's legitimate.

With that aside, they say they doubt mobo's will have those built-in. PCI-E expanders (like Lucid) break the original purpose of platform. You can have the same on Sandy.

 

[ramcoza]

Why 1680x1050 resolution? What will be the benchmarks(Games) at 2560x1440 resolution? will it be the same story?

 

[John Doe]

Why 1680x1050 resolution? What will be the benchmarks(Games) at 2560x1440 resolution? will it be the same story?

It's better off testing at a lower res to take GPU out of equitation. You still need the same amount of CPU power regardless of resolution.

 

[qubit]

Thread title is fine

New architecture gives same gaming framerates as old architecture = disappointing. What's hard to see?

Of course, we need to hold our breath for the official results on release day to be sure, which is why I pose it as a question.

 

[claylomax]

The question is: Can Sandy Bridge-E beat Bulldozer's "awesome" multi-GPU performance? http://hardocp.com/article/2011/11/03/amd_fx8150_multigpu_gameplay_performance_review/

 

[John Doe]

The question is: Can Sandy Bridge-E beat Bulldozer's "awesome" multi-GPU performance? http://hardocp.com/article/2011/11/03/amd_fx8150_multigpu_gameplay_performance_review/

Fail is fail I guess, no need to mention it again by going off-topic... but if I was to take it serious, even my old i7 870 can beat it. BD's performance is miserable.

 

[qubit]

Why 1680x1050 resolution? What will be the benchmarks(Games) at 2560x1440 resolution? will it be the same story?

When benching CPU framerates, one should reduce the resolution as much as possible, to remove the GPU as the bottleneck, otherwise, you just get all the results for the CPUs under test topping out and hence showing the same performance, when they actually all have different performances. If anything 1680x1050 is too high and that's why they look the same in those slides. I would test at 800x600.

Regardless, I'll say it again, wait for the official benchies tomorrow before passing judgment.

And welcome to TPU.

 

[entropy13]

When benching CPU framerates, one should reduce the resolution as much as possible, to remove the GPU as the bottleneck, otherwise, you just get all the results for the CPUs under test topping out and hence showing the same performance, when they actually all have different performances. If anything 1680x1050 is too high and that's why they look the same in those slides. I would test at 800x600.

800x600:

 

[qubit]

800x600:
http://i.imgur.com/QVEn9.png

That's hard to believe. You got a link?

If it was true, then one would simply buy the cheapest CPU and be done with it.

 

[entropy13]

That's hard to believe. You got a link?

If it was true, then one would simply buy the cheapest CPU and be done with it.

You don't know what Touhou is?

It's not exactly known for its graphics, considering the games were all made by one guy. So even if I don't have a 3960X yet, I'm sure that such results are true. AFAIK you can't do anything with the frame cap at 60fps, and even an Atom and Intel integrated graphics can play the game.

 

[erocker]

New architecture gives same gaming framerates as old architecture = disappointing. What's hard to see?

Of course, we need to hold our breath for the official results on release day to be sure, which is why I pose it as a question.

Who actually games at horribly low resolutions anymore? If they do, I don't think they are the marketing target for these chips anyways. Theoreticaly if a GPU needs two "lanes" to achieve it's maximum performance adding more "lanes" won't make any difference. Now, unless this new CPU has a GPU that can work with a discreet GPU or perhaps magic it won't increase framerates in gaming.

 

[qubit]

You don't know what Touhou is?

It's not exactly known for its graphics, considering the games were all made by one guy. So even if I don't have a 3960X yet, I'm sure that such results are true. AFAIK you can't do anything with the frame cap at 60fps, and even an Atom and Intel integrated graphics can play the game.

No, I've never heard of it before, enlighten me?

And measuring gaming performance with a softaware frame cap or vsync on is pretty stupid, isn't it? :shadedshu

Who actually games at horribly low resolutions anymore? If they do, I don't think they are the marketing target for these chips anyways. Theoreticaly if a GPU needs two "lanes" to achieve it's maximum performance adding more "lanes" won't make any difference. Now, unless this new CPU has a GPU that can work with a discreet GPU or perhaps magic it won't increase framerates in gaming.

One wouldn't game at it, but one would certainly bench at it when comparing CPU performance, to remove the GPU from the equation.

 

[erocker]

One wouldn't game at it, but one would certainly bench at it when comparing CPU performance, to remove the GPU from the equation.

Which is a useless way to test a CPU as the data provided shows no useful and/or real world benefit. Benchmarks/applications that actually use the CPU show it as being quite a bit better than SB.

 

[entropy13]

No, I've never heard of it before, enlighten me?

Just go check it's page in Wikipedia, "Touhou Project."

And measuring gaming performance with a softaware frame cap or vsync on is pretty stupid, isn't it? :shadedshu

Depends though. Comparing over 200 fps to another over 200fps is pretty stupid too.

It should have been obvious already what I was trying to point out already. There are "benchmarks" where CPU power doesn't really matter at all.

Touhou is just a very extreme and ridiculous example.

 

[John Doe]

Which is a useless way to test a CPU as the data provided shows no useful and/or real world benefit. Benchmarks/applications that actually use the CPU show it as being quite a bit better than SB.

No, he's right. Numbers are numbers. If one chip is showen to be faster (regardless of resolution), you know it'll have more potential and less chance of a CPU slowdown. Look at this chart for example, he's limited by GPU (and game) so no difference.

---------------------

Then see this. At a lower resolution, there's a big difference in frametime (in a more CPU intensive case at the same time). This is a properly done test, that one above isn't:

http://www.hardwarecanucks.com/charts/index.php?pid=70,76&tid=3