i7 920 vs. i7 860 WCG comparison

[Fitseries3]

our good friend 123bob over at XS has done a good comparison between a i7 920 and a i7 860 and how they crunch.

its a great read if you are wondering how they compare and how you could possibly save $ on the electric bill

http://www.xtremesystems.org/forums/showthread.php?t=235419

 

[theonedub]

After missing out on an i7 920 and seeing everyone going 1156 I was curious about this exactly. Thanks!

 

[InTeL-iNsIdE]

Thanks, will make for an interesting read

 

[PaulieG]

Great read. I'll be in this exact same situation next week, whenever my 860 comes in.

 

[[Ion]]

Thanks, this makes an i7 cruncher very appealing! I'm surprised to see that the i7 860 and the i7 920 are so close at the same speeds. If I ever go i7, I know what I'll be getting (i7 860)

 

[Fitseries3]

results......

direct from 123bob.....

OK, drum roll please....I have results....Interesting results, I might add...:yepp:

REMEMBER, an IDENTICAL COPY of 24 WUs were run on each machine.


First the mandatory screenies.

The 920 IDLE CPUz shot...

The 920 BOINC Data Screen.

Add up the hours/minutes/seconds as individual columns and you get total 85:658:702.

This is easier math than converting all to decimal hours at this point. I could not get a good cut and paste out of boincmgr to go right to excel. Had to do it manually. Arrrgh. SOMEONE IS MOST WELCOME TO CHECK MY MATH. I did this bleary eyed at 5am here...




The 860 IDLE CPUz shot. The only difference I see here is on the memory page for NB freq, whatever that is. The 860 is dual channel, the 920 is triple of course....

The 860 BOINC data screen.

Add up the hours/minutes/seconds as individual columns and you get total 85:642:697 for the 860.

Now, it gets interesting. 24 WUs were run.

For the 920, converting to decimal hours we get 96.161666 hours. Divide that by 24 and you get 4.006736 Hours/WU.

For the 860, converting to decimal hours we get 95.8936111 hours. Divide by 24 and you get 3.9955671 Hours/WU.

Note the 860 is doing WUs FASTER than the 920.....

How much? Well, .0111688 hours, or .6701322 minutes, or 40.207 seconds.

So the 860 is about 40 seconds average FASTER per WU than the 920.

I think I'd still call them even, but in no way is the 860 "crippled" compared to the 920.

So at least at the 3.7 gig data point, the 860's better power efficiency and similar score to the 920 makes it a good choice for pure crunching. If you need to go Gulftown later, then 920 might be your choice.

I'm also finding the X3440 choice INFRNL made to be interesting for farm rigs....Might have to get a dozen of those.

Bob

 

[123bob]

I'm doing another run just to be sure on the original thread at XS. I had the uncore multi a bit lower on the 920. We don't think it will make a big difference, but I want to be sure.

I paid hell to figure out how to get it to match on this MSI board. It didn't seem to be as straight forward working with the bios as I thought. I had to set the mem multi, then the needed uncore multi popped up in bios. It could just be me, but it was confusing....

@Fits, please let me know before taking a post.....If this was the only post everyone saw, and I hadn't come over to check, folks may not have gotten the "rest-of-the-story".. I don't mind you doing it, just that I know that there's another copy somewhere else, in case I have an update.

For now, I would suggest folks see the original thread at XS. I'll be updating the data with a new run in about 16 hours.

Hope everyone is doing fine here. As you can probably tell, I have not been bored lately. I'm finally getting my feet wet on the i7 stuff. Another chip we're looking at is the X3440. It's an even cheaper hyperthreaded part at 2.53gig. If it clocks well, it might be the perfect choice for crunchers on an 1156 board.

My best to TPU,
Bob

 

[Fitseries3]

thanks and sorry. people should see your thread on XS anyways to get ALL the info and not just the excerpts.

 

[123bob]

No problem Fits, it's all good.

The thread at XS also had a very good conversation on what metric to use to compare rigs. What we came up with was a way to sneak work units from one rig to another in order that both machines are crunching EXACTLY the same WUs. We are then taking the average of the completion times for 24 of them to compare. It seems to be working. The trick was to make the transfer work. I detailed that in the thread. Some folks over here may already know about how to do this, IDK.

Regards,
Bob

 

[123bob]

I put up another post showing the effect of boosting the Uncore and 1T timing on the 920 rig to exactly match the 860 rig. In a nutshell, it brought it in to be about 1% FASTER than the 860. This is not a huge amount, but a difference.

Bottom line is that they are still in the same "ballpark" for clock-to-clock performance. With the 860's better power performance, at least at 3.7 gig, it looks like a good cruncher choice to me still.

I'm going to move to 4 gig soon. We know the lynnfield 860 may be a power pig to get it there. The equation may change at that speed....

Details are on the thread.

Regards,
Bob

 

[[Ion]]

Thanks Bob, this info is very useful!