Core i5 2500K & Core i7 2600K processor review [Guru3D.com]

If peoples want it they can ... there's around a lot of motherboards who support it with this chipset, gigabyte UD7- Asus Workstation, Maximus IV, some MSI.... just choose what fit for you .

 

The article references the board dropping to x8/x8 when running multi-GPU. I'm assuming the same occurs if you populate the second x16 slot w/ a non-GPU card? I run an PCIe x8 Highpoint RR4320 RAID card in the 2nd PCIe x16 slot.

 

A little niggle about the article. You make this statement:

The reviewer also states that Intel's OEM cooler is used in the review. Asus states in the product manual (actually in an addendum insert) that best cooling is achieved by using a downward blowing CPU cooler. Installation of the assistant fan is recommended for those not using downward blowing CPU coolers (ie tower coolers, water coolers, etc).

I wonder if the "pointless feature" statement would have been made if the reviewer weren't using the stock cooler w/ its downward blowing fan - I'd argue that most users of this board will not be using the stock cooler, or an aftermarket cooler with downward blowing fan. In summary, Asus agrees that the assistant fan is unnecessary for in configs such as used in this review. But the reviewer offered a blanket statement that may or may not be true for users of other cooling methods - something that he didn't bother to verify.

 

If there's not an option in the BIOS to set those slots manualy to x16/x1 or x8/x8, there probably will be such an option added in a BIOS update. But to answer your question: The system doesn't know if the card being sloted is a graphics card or a RAID card or an (insert card type here) card. All it can tell is how many PCIe lanes the card is trying to use.

But it's a moot point. If you put that PCIe x8 card in the second slot, you're going to have to give it 8 lanes of bandwidth anyways, and that means the system has to set the two slots to x8/x8, period. There's no getting around that with this chipset.
Solution? Either go with a x58-based board, or wait for the Waimea Bay chipset to launch. That's the only way you're gonna get more than 16 lanes of PCIe bandwidth on an Intel-based board.

 

http://www.guru3d.com/article/core-i5-2500k-and-core-i7-2600k-review/10

The title says it's the i5 2600k. That should be i7 2600k right?

 

Considering the base clock has been dropped to 100Mhz I would like to see a testing of the i870 vs i2600K at the same overall speed. ie i870 @ 133Mhz base versus i2600K downclocked to i870 speeds via multiplier.

Then let's see what the scores are eh? I find that not much is said regarding this castrating of increasing base clock speed.

I'll stick with the i870 clocked to 4.0Ghz with a base clock at 200Mhz rather than a 2600k at 4.5Ghz on a 103Mhz base clock. (I've played with ten and I've not found ONE make it past 105Mhz without generating errors in the stabilty testing.

i870 is faster than i2600k clocked to 4.3ghz when multiplexing raw audio and video into a finished product via Adobe Premier.

I'm not impressed with this review that tows the Intel line. :biggun: :bang:

 

No, it was proven it was gigabytes quality control problem

 

Sure... Well, I guess you're right and every hardware website and end-user is wrong; the 2600k is inferior to the 870.

How about we have a simple test, it takes my 2500k 1:48 sec to do Super PI to 8m places.

 

I'm pretty sure X79 + LGA 2011 CPU's are going to trounce the 2600k. Lets wait till March :nerd:

 

Why wait till march? They should be out december-january. :eyebrows:

 

When you can leverage the two extra cores in a meaningful way, absolutely.

I doubt real-world per-core performance will see any notable gains though. Time will tell, obviously.

 

your right, I am now a believer. that extra pin 1156 vs 1155 is the dealbreaker. anybody selling a 1156 so I can upgrade this inferior 1155?

 

What's with old thread resurrections these days..

 

The information is still valid. The technology of both platforms is yet to be obsolete.

It stands on it's own.

 

It's new tech. I don't believe this new tech is a replacement for the older tech 'yet'.

I will be ready to purchase when they tick the baby. If I was going to purchase now, I would go with i870 as I am not certain Intel won't play pin games again.

But they are the only fast runner on the block.

This sandybridge is not the performance increaser that some are saying. Not yet it isn't.

:wanker: :biggun:

 

I was being sarcastic. SB is faster than 1156 no doubt

 

So you'd choose a slower processor on an obsolete socket over one with an unknown future (past Ivy Bridge anyway)?

That makes absolutely no sense.

Over here an i7 870 and an i7 2600K is the same price and the latter is most definitely faster.

 

I disagree. The tock is no faster than the tick unless you like playing with benchmarks and at the same speed, there is a delimit to the 100Mhz base clock regardless of the overclock.

Multiplexing raw video and audio in realtime with Adobe Premier shows the 100Mhz Base Clock castrates the flow.

Remember the i486 DX 50 vs i486 DX2 50

The first ran on a 50Mhz bus and the second was a castrated multiplied 25Mhz bus.

The second was far slower. Then the DX2 66Mhz came out with a 33Mhz bus and a 2x multiplier. It also was slower than the DX50. The DX50 ran hot so they moved to the multiplier to increase speed of the CPU and not be hampered by the restrictions of the bus. When DX4 100 (25 x 4) came out it finally processed faster at the lower bus level regardless of the overclock.

http://registrywinner.net/pcsupport.../9732-Which-is-faster--a-DX-50-or-DX2-66.html

So, you play with your 2600K and tell yourself what a speed demon it is. When the data drops to the base clock in heavy multiplexing, my 200MHz Base Clock does not drop frames. At the same speed with the 100Mhz Base Clock, the 2600K drops frames in preview.

I await the tick.

 

You may do that of course, I merely explained why you're wrong to do so.

I'm fairly sure my Motorola 68060 from the mid-90s would offer decent performance as well - if it could run at 4.5GHz as opposed to 50MHz.

Unfortunately it can't.

You BCLK claims are nonsense, the S1156 has the clock locked against external buses while S1155 hasn't. That's why you're able to overclock the BLCK to such a great extent in the first place.

But don't take my word on performance, take any review comparison that you'd like.

 

Tock has seldom been faster than tick. Your Motorola comparison is long before the intel tick/tock and is not Intel anyway. Stay on track in comparisons as this one is quite tangential to the discussion.

If you don't know that then you are far more ignorant of Intel CPU history.

You can continue to argue that this tock is much faster than the tick.


You won't be the first to argue this side. But the evidence is there that the tock is seldom faster than the tick.

http://en.wikipedia.org/wiki/Intel_Tick-Tock