Just swapped my 1090T for a 2500K

[bgt]

Using the 2500k as my main computer as of now but I am not entirely happy with the multitasking speed. Small programms fly.
In trying to see the differences between these 2 CPU's I am using both with the same setup in software.
The 2500k boots very fast and unpacks a single item truly fast.......but when heavy multitasking is involved it looses out to the 1090T sometimes to a degree that I think its stalling. For a 4core cpu its mighty fast though. Normal feel is however the 1090T isn't really fast in anything but also not really slow in anything while the 2500K can be. The 2500K perf. is sometimes worse than the 1090T and sometimes better/faster. The 1090T is more balanced.
However power draw from the wall is a difference.
idle(surfing) load(Winrar perf.test)
1090T 66W 115W 2750 perf. Winrar
2500K 43W 85W 3100 perf.

Playing XXHighend music programm now(through optical SPdiff with external Highend DAC=enjoying tremendously), typing here, moving a file from 1 disk to another, doing a backup and an extraction and the CPU starts sweating.

Will keep on testing and look for differences





Asus P8H67-M LE/i5 2500K/8GbCrucial1333Mhz/C300-64GbSSD+1Tb/Win64HP
Asus M4A88TD-M EVO/PhII1090T/8GbCrucial1333Mhz/M4-64GbSSD+1Tb/Win64HP

 

[richierich1212]

Same here. I feel the difference too.

 

[sangyup81]

are you running them both on stock? how are temps?

 

[bgt]

Yep, both on stock. Temp is between 37-50C for the 2500K and 1090T.

 

[BD231]

More cores FTW?

 

[podspi]

For heavy office multitasking, more cores (as long as those cores are reasonably quick) is better than fewer faster cores.

My X6 setup is easily the best experience I've ever had. I still think a quad-core is more than enough for most people, though.

 

[toyota]

For heavy office multitasking, more cores (as long as those cores are reasonably quick) is better than fewer faster cores.

My X6 setup is easily the best experience I've ever had. I still think a quad-core is more than enough for most people, though.

even a stock 2500k easily beats the 1100 X6 oced to 4.2 in multi tasking. http://www.bit-tech.net/hardware/cpus/2011/01/03/intel-sandy-bridge-review/9

 

[GaiaHunter]

even a stock 2500k easily beats the 1100 X6 oced to 4.2 in multi tasking. http://www.bit-tech.net/hardware/cpus/2011/01/03/intel-sandy-bridge-review/9

Maybe it is similar to you snappier feeling?

 

[toyota]

Maybe it is similar to you snappier feeling?

lol, maybe

 

[Phynaz]

Playing XXHighend music programm now(through optical SPdiff with external Highend DAC=enjoying tremendously), typing here, moving a file from 1 disk to another, doing a backup and an extraction and the CPU starts sweating.

CPU isn't sweating, you're I/O (disk) bound at this point.

Nothing listed above is going to tax any quad core.

 

[tweakboy]

Most of you have never even used 50 percent on their CPU's . a constant 50 percent ? Games avg or video or audio ..

 

[podspi]

even a stock 2500k easily beats the 1100 X6 oced to 4.2 in multi tasking. http://www.bit-tech.net/hardware/cpus/2011/01/03/intel-sandy-bridge-review/9

I've seen your other thread, I know you know what I'm talking about

I'm not trying to imply that the X6 has a higher level of throughput than the 2500K, I'm implying that there are two different (and important) types of speed:

There is computation speed and interface speed. My example is fairly ridiculous, and also simple, in the hopes that I get my point across, I'll give a real world example as well:

Suppose we had one SB core, no HT, clocked at something rediculous, say 50ghz.


Further suppose we started using that core to transcode an HD video that is long enough to take a while, and the priority of that thread is set to real-time. If I want to do anything else with that computer, I'm pretty much screwed, because there are no more resources available. The interface will "feel" slow.


Contrast this to say, a 1ghz dual-core Bobcat, one-thread converting the video, and another free to do whatever. In this case, even though the Bobcat is slower in every single way, it will still "feel" faster.


Now, that isn't a practical example, and as others have pointed out, most people are likely to be IO constrained in those types of situations. I have an SSD and 8gb of memory. At one point I was analyzing a large dataset (30gb+) with some custom code I myself had written. I spawned 5 threads, leaving the last core idle. Subjectively, the machine operated as fast as my C2D @ 1.5ghz. IMHO, that is worth something.


More recently, I was playing Terraria, Minecraft, and compiling CyanogenMod in a Linux VM at the same time. Yes, not a likely use-case for many people, and perhaps a 2500K would work just as well. But my statement still stands, best machine I ever purchased

 

[bgt]

even a stock 2500k easily beats the 1100 X6 oced to 4.2 in multi tasking. http://www.bit-tech.net/hardware/cpus/2011/01/03/intel-sandy-bridge-review/9

Sorry to say but real life experience when u are using both machines tells something different. Numbers are numbers. There is no substitute for real time usage and the 1090T feels snappier when doing a lot of stuff.

 

[bgt]

CPU isn't sweating, you're I/O (disk) bound at this point.

Nothing listed above is going to tax any quad core.

With an SSD? The 1090T has no problem with this. The 2500k starts stutterting a bit when doing other things now.

 

[Arkaign]

lol, maybe

hehehe

 

[Rhoxed]

1090T 66W 115W 2200 perf. Winrar
2500K 43W 85W 3100 perf.

I this in winrar benchmark?

I score right at 3k with a 1090T @ 3.5ghz, 4.0ghz I get well over 3k

EDIT: ram speed might have an effect, i am running at 1700mhz

 

[bgt]

I this in winrar benchmark?

I score right at 3k with a 1090T @ 3.5ghz, 4.0ghz I get well over 3k

EDIT: ram speed might have an effect, i am running at 1700mhz

Yep, everything at stock.

 

[Arkaign]

Looks like you've got the NB clock up pretty good as well That seems to have a respectable impact on some things.

 

[bgt]

How to OC on this board? I see that the P67 can only make use of the multiplier function. Got the wrong board for this.

 

[Arkaign]

How to OC on this board? I see that the P67 can only make use of the multiplier function. Got the wrong board for this.

No, you have no problems. The whole idea of the K variant SBs is that you get an unlocked multi. Regardless of how you get there, running that sucker at 4.5ghz or so is gonna be a rocketship. You kind of have to forget the entire idea of bus overclocking the 2500 and 2600k chips.

http://www.bit-tech.net/hardware/cpus/2011/01/07/how-to-overclock-the-intel-core-i5-2500k/1

 

[bgt]

My motherboard, the Asus P8H67-M LE, shows 33x being the highest multi.
So no 4,5Ghz

 

[bgt]

Picture quality improved a lot with an extra 6450 graphic card. There is something with the IGP thats not right. You just cannot get the colors right. This is much better with the 4200IGP. So now with the 6450 its ok. Energy consumption increased a bit to 55W idling/surfing.

 

[dma0991]

My motherboard, the Asus P8H67-M LE, shows 33x being the highest multi.
So no 4,5Ghz

There is no unlocked multiplier on a H67 board so you can't go above 33x100 unless in a Turbo Boost scenario. There are no unlocked multiplier for H61 and H67 boards while the chipsets with unlocked multiplier are exclusively for P67 and Z68 motherboards only.

 

[bgt]

Yes, pity. Anyway..a reason to swap to the 2500 for me was energy consumption. Don't know how much it will go up when on 4.5Ghz?

 

[dma0991]

I would prefer to buy a Core i5 2500K for its overclocking potential rather than saving electricity. The additional 20W power consumption of the 1090T is small and to me the ROI in power savings is not worth changing. It's a desktop so extra power consumption doesn't matter a lot but it's a different case if it were a laptop or any mobile device IMHO.

It depends on the settings you have at 4.5GHz but some do get a Core i5 2500K + H67 just because they want an IGP that is better than the HD2000 and do not need a IGP that is better than the HD3000.