Anandtech useas oudated bechmark softwares on pupose(?)

[Phynaz]

Thats gross man, what about the athlonxp series?
Why dont you overclcok the 1100T too?

An overclocked 2500K will handily beat an overclocked 1100T in absolutely anything you wish to do.

Reading comprehension failure

 

[Rifter]

Sorry to break it to you but AMD is just not competitive right now. They are doing all they can with the phenom II to keep up with 4 year old intel CPU's. I hope BD changes that.

BTW your post makes no sense whatsoever, if its not multi threaded then HT is DOING NOTHING ANYWAYS. Intel just has a very large single threaded advantage, this is one of the reasons i upgraded from a phenom II to a i7..

 

[dac7nco]

Reading comprehension failure

Posting from Assbackistan.

Daimon

 

[Arkaign]

Hyperthreading is Symmetric Multi Threading. HT is Intels' brand name for SMT.

What did you think it is? And how did you think it helps in single threaded applications?

It's a little confusing. HT is a form of SMT, but SMT is weaker than SMP, and SMP doesn't require HT, it just requires a multi-core CPU. A non-HT Intel dual, quad, hex, etc, will still have SMP. At least that's my understanding, I'm not nearly as technical as a few on here.

HT seems to help with loading the pipeline with more threads to avoid wasted CPU cycles. Though there is the drawback that some apps actually slow down a bit due to being loaded onto the virtual HT cores, I think this is due to the process scheduler and varying degrees of optimization for the HT/SMT/SMP considerations.

 

[podspi]

I doubt anybody would be able to come up with an example where, with the same benchmark, Anand's results are significantly different than everybody elses' (unless there is some other factor, like IO in a Sysmark benchmark, for example).


Benchmark selection and updating benchmarks is difficult, because while newer benchmarks are likely to be more applicable to how new CPUs are going to be used, they are also incomparable to older benchmarks. I do think that this is an issue, and personally I think Anand should (perhaps every year?) do a survey for readers asking them what programs they use, and what benchmarks they'd like seen done. Every year, a few benchmarks can be added/replaced, and old benchmarks removed. This way we don't have an entire refresh (so we can still compare somewhat to the past) but we also keep benchmarks "fresh".

Also, I personally believe that this obsession some people have with "benchmarks favoring one arch over another" is nonsense. People need to come to terms with the fact that Intel has the majority of the marketshare, and thus many programs, benchmarks and retail programs may be specifically optimized for Intel processors. While I don't believe Intel should be allowed to (and they no longer are allowed to) cripple AMD processors via their compiler, benchmarks, imho, should be chosen on what people actually use, regardless of what they've been optimized for.

 

[dac7nco]

I do think that this is an issue, and personally I think Anand should (perhaps every year?) do a survey for readers asking them what programs they use, and what benchmarks they'd like seen done.

That is the best idea I've heard in quite some time.

Daimon

 

[Phynaz]

OMG, a test where AMD does exceedingly poorly at, THROW IT OUT!!!

Come on man! You know Spintel colluded with M$ to make Excel run slower on AMD CPUs.

If you don't admit it then you are nothing but a paid shill!!

 

[MagnusTheBrewer]

I doubt anybody would be able to come up with an example where, with the same benchmark, Anand's results are significantly different than everybody elses' (unless there is some other factor, like IO in a Sysmark benchmark, for example).


Benchmark selection and updating benchmarks is difficult, because while newer benchmarks are likely to be more applicable to how new CPUs are going to be used, they are also incomparable to older benchmarks. I do think that this is an issue, and personally I think Anand should (perhaps every year?) do a survey for readers asking them what programs they use, and what benchmarks they'd like seen done. Every year, a few benchmarks can be added/replaced, and old benchmarks removed. This way we don't have an entire refresh (so we can still compare somewhat to the past) but we also keep benchmarks "fresh".

Also, I personally believe that this obsession some people have with "benchmarks favoring one arch over another" is nonsense. People need to come to terms with the fact that Intel has the majority of the marketshare, and thus many programs, benchmarks and retail programs may be specifically optimized for Intel processors. While I don't believe Intel should be allowed to (and they no longer are allowed to) cripple AMD processors via their compiler, benchmarks, imho, should be chosen on what people actually use, regardless of what they've been optimized for.

One fairly recent review that definitely showed discrepancies with many other sites (albeit regarding a GPU) was the review of the HD 6790.

 

[OCGuy]

Come on man! You know Spintel colluded with M$ to make Excel run slower on AMD CPUs.

If you don't admit it then you are nothing but a paid shill!!

I seriously think some of these guys sit around on AMDzone and say "Hey, let's go raid Anandtech with conspiracy garbage, just to troll it's members!"

I understand that the rumour of H2 retail availability of BD is probably getting the blood boiling, but no need to re-hash old conspiracy arguments.


If they stop being fed, they will go away.

 

[losbellos]

Reading comprehension failure

My mistake however I disagree, 2500k is just about the same as the x6

 

[cloudzero]

My mistake however I disagree, 2500k is just about the same as the x6

Sweet. Do some benchmarks and show them.

 

[aigomorla]

say it with me people...

its for scaling.

we here on the forum love to say... x is better then y.

The only way we can logically prove x is better then y is to run them on the same applications.

If you run X on one set of software and Y on the other, how does that help us in x vs y?

 

[darkswordsman17]

This.

It's simply a matter of convenience for aiding in the performance comparisons to systems tested in the past.



What's so surprising?

Are you going to be like this when Llano is launched and all the on-die GPU comparisons are made to sandy-bridge? "ZOMFG, Llano, at much lower clockspeed to SB's IGP, trounces the SB IGP! This so called "GPU benchmark" is so totally over-optimized! For shame AT, for shame!".

That's true, but I would personally like to also see more current benchmarks in addition to them. Look at how they've broken up the years in the GPUs on Bench, why can't they do that with CPUs as well?

I'm not claiming bias, and this isn't just about Anandtech, but I've felt CPU benchmarking has gotten stale the last few years. I'd like to see in depth heavy multitasking benchmarks. Seemed like that was barely touched ever, many sites sorta started to do it back when they first were going dual core and the like, but that was with WinXP. You don't even see web browsing benchmarks any more and there's been some huge changes in that the last few years. I know that it can often go from a CPU comparison to a software comparison, but I do wish we had more real world testing.

 

[Elixer]

say it with me people...

its for scaling.

we here on the forum love to say... x is better then y.

The only way we can logically prove x is better then y is to run them on the same applications.

If you run X on one set of software and Y on the other, how does that help us in x vs y?

If the tests are flawed, then that don't prove anything.
If that is the case, then the logical thing to do is to use another program, and then BOTH things should be retested.

 

[Cogman]

If the tests are flawed, then that don't prove anything.
If that is the case, then the logical thing to do is to use another program, and then BOTH things should be retested.

Tests from the past 5->10 years generally are not flawed. Synthetic benchmarks will, at most, shift which instructions it finds most important. That does not in any way make the test flawed.

IF someone proves that one of the old tests is flawed THEN we should replace or ignore them. Until then, the old tests provide some really good info on how components compare to each other.

When anand updates his benchmarks, more than likely he won't recheck the old stuff. He might do the previous generation and that is it. You guys don't seem to realize how much time running a full benchmarking suite takes.

 

[996GT2]

My mistake however I disagree, 2500k is just about the same as the x6

Lol you must be smoking some good stuff:

2500K already destroys the 1100T when both are at stock speeds, and the 2500K can overclock much, much higher. When was the last time you saw an 1100T @ 4.8 GHz on air cooling? That's an overclock that pretty much any 2500K can achieve on air cooling.

 

[darkswordsman17]

Lol you must be smoking some good stuff:

2500K already destroys the 1100T when both are at stock speeds, and the 2500K can overclock much, much higher. When was the last time you saw an 1100T @ 4.8 GHz on air cooling? That's an overclock that pretty much any 2500K can achieve on air cooling.

Hey, if you cherry pick a couple of those, then they are just about the same. The Phenom even beats it in the HD Encode x264 2nd Pass test. ()

 

[skipsneeky2]

Lol i was reading that chart up above and i must have been living under a rock.

Theres now a amd x4 980? wow impressive stock clocks but im sure even still it cant even come close to a 2500k at the same price tag.

Still a good option for those still on a amd dual core and wanting that last minute upgrade before ivy bridge perhaps.

 

[Accord99]

Do you happen to know for example that cinebench 10 is unable to perform for example ambient occlusion pass multithreaded,so this is why intel processors are in advantage since hyperthreading can help in it a lot. Of course this doesnt happen in cinebench 11,5....

Hyperthreading has a bigger benefit in Cinebench 11.5 than it does in 10 for Sandy Bridge processors.

 

[Ben90]

You don't even see web browsing benchmarks any more and there's been some huge changes in that the last few years.

 

[Bill Brasky]

LOL. Well played, sir. :thumbsup:

 

[darkswordsman17]

Touche...

I basically always have multiple browsing windows open with between 5-100+ tabs open, often with video playing and other heavy flash usage, or other relatively strenuous (there's a lot of content creation and photo editing tools that are web-based) processing going on. I would like to know how CPUs compare when handling load like that while they manage other programs that are much more processor intensive (CAD, video/audio editing, even games).

 

[aigomorla]

If the tests are flawed, then that don't prove anything.
If that is the case, then the logical thing to do is to use another program, and then BOTH things should be retested.

lol... your right...

and we tend to not realize they are flawed until we compiled a ton of data.

But some of them arent, and some do allow us to scale.


In all honesty Elixer, a testers life is a serious BIYATCH (me having been there). You think the free stuff is worth it sometimes... other times, you wonder why u got yourself in this big hole u dug yourself.

 

[Cerb]

Touche...

I basically always have multiple browsing windows open with between 5-100+ tabs open, often with video playing and other heavy flash usage, or other relatively strenuous (there's a lot of content creation and photo editing tools that are web-based) processing going on. I would like to know how CPUs compare when handling load like that while they manage other programs that are much more processor intensive (CAD, video/audio editing, even games).

That is very difficult to benchmark. General productivity benchmarks will tend to be the closest to what you are looking for.

 

[mosox]

Not only Anand uses outdated software but for games they don't even show the full HD results and they don't state what video card and settings they used in the comparison bench.

And in the Sandy Bridge review Anand used low resolutions (even 1024 x 768) and no AA/AF in order to make the Phenom II look bad compared to Intel.

No matter that nobody runs a fast CPU and video card at those settings.

Lots and lots of gamers on a budget are buying Intel's expensive stuff and weak video cards because they're being misled by Anand's benches who are trotted on all the HW forums instead of buying cheaper AMD CPUs and mobos and good video cards for the same money.

But the benches are not made for them.