Core i7 2600K 69% faster than Zambezi-FX according to TrueCrypt benchmark leaked

[AMD_Gamer]

What do you guys think about this? this is a TrueCrypt benchmark http://static.techspot.com/articles-info/353/bench/Application_04.png

and this is a slide from AMD: http://www.abload.de/img/3ncx5.jpg

 

[kirtar]

That benchmark graph doesn't even reference zambezi. The only AMD processors I see on there are the Phenom II X6 1100T and Phenom II X4 970. Everything else is intel. Even if you reference the AMD slide, it says greater than 2x of the 1100T, which is not specific at all, this still doesn't say that much. For all you know, it could be 2.98x faster, but they can't say more than 3x if they want to stick to integers.

I suppose they did list a >1600 number on the bottom of the slide :/

 

[SonicTron]

I think the real question is what do you think about this?

 

[InternationalHat]

Full configuration at end of deck. So which fx processor was it? Also that's a specialized instruction set that's accelerated so it could be contrary to overall performance. These leaked slides really don't tell much unfortunately.

 

[nimbyfaygo]

Truecrypt 7 is utilizing the I5 and I7's AES-NI instruction set that greatly boosts performance for the cpus that support it.

Since the I5 and I7 are supporting hardware encryption/decryption and the AMD isn't, its a pretty terrible comparison.

 

[kirtar]

Full configuration at end of deck. So which fx processor was it? Also that's a specialized instruction set that's accelerated so it could be contrary to overall performance. These leaked slides really don't tell much unfortunately.

Yeah that also came to mind.

 

[kirtar]

Truecrypt 7 is utilizing the I5 and I7's AES-NI instruction set that greatly boosts performance for the cpus that support it.

Since the I5 and I7 are supporting hardware encryption/decryption and the AMD isn't, its a pretty terrible comparison.

fwiw, I believe BD will support AES-NI

 

[nimbyfaygo]

fwiw, I believe BD will support AES-NI

I don't have any idea if they will or not. I would like to think that they will though.

 

[snowysnowcones]

what a stretch of a benchmark claim

 

[shawnmramsey]

I think it looks very similar to many synthetic benches that look great, and make consumers want to purchase, but in the end will only benefit a very small percentage of the users. Not saying it is not impressive, but in my opinion it is a bit pointless for an "all around bench".

We will see what is to come, and everyone has their own application for their computer so it all matters on how you use your computer.

Interested :
Shawn

 

[savantu]

It simply shows BD's implementation of the AES-NI is suboptimal compared with SB.
But I find it distressing for AMD to show such a lousy benchmark, BD must be doing poorly if they didn't have anything better to show.

 

[griff30]

Let's see how they repeatedly compair on actual physical compairson in real world bench marks. This is bogas as the AMD chip is not officially out and any "leaked" benchmark is utter bullshit. They only release downgraded chips for developers at this point to sandbag.

 

[Funkmaster_A]

">1600" MB/s on FX, versus 2700 MB/s on an i7? If this was some random incidental thing I wouldn't care, but the fact that AMD's slide is sort-of bragging about it seems like a bad sign.

 

[ShuttleLuv]

I am making my prediction here, and it's not trolling because I love AMD just as much as INTEL and loved my Opteron, but I predict that AMD will pull ahead in one or two specialized benches, but basically take a at least a 20% beating in everything else. This new FX is going to be as disappointing perf wise as Duke Forever is gaming wise imo.

But I could be wrong and if it's faster I will bump this in congratulations and say I was wrong.

 

[Harry1969]

I am making my prediction here, and it's not trolling because I love AMD just as much as INTEL and loved my Opteron, but I predict that AMD will pull ahead in one or two specialized benches, but basically take a at least a 20% beating in everything else. This new FX is going to be as disappointing perf wise as Duke Forever is gaming wise imo.

But I could be wrong and if it's faster I will bump this in congratulations and say I was wrong.

I'm also going to make a prediction and say BD will not dissapoint. It's based on as much factual info as any prediction about BD performance which is none.

 

[insaneshell]

I dont care if it beats all the SB chips, if it maintains the trend of the last 3 years of 90% performance for 75% of the cost I'll still buy it, and still recommend it. Im not an AMD supplier because they are slower but because they offer the better price/performance and if this holds true ill stick with being an AMD supplier

 

[sc3252]

">1600" MB/s on FX, versus 2700 MB/s on an i7? If this was some random incidental thing I wouldn't care, but the fact that AMD's slide is sort-of bragging about it seems like a bad sign.

The bigger question is, how does it affect real world performance? I am just wondering if this is something like pcie x8 or x16 where the difference is minimal at best.

 

[pxc]

What do you guys think about this?

Nothing much. There are not too many applications limited by AES encryption speed. Obviously that only becomes a problem where performance might be limited by AES encryption speed and BD would be at a disadvantage.

For the vast majority of desktop/laptop users, the difference in performance is insignificant, if it's even used at all.