Message boards :
Number crunching :
Core i7 running Seti
Message board moderation
Author | Message |
---|---|
Francois Piednoel Send message Joined: 14 Jun 00 Posts: 898 Credit: 5,969,361 RAC: 0 |
welcome to Core i7 965 into the Top 40 computers. here Still climbing! the press has many articles for your reading. help yourself. my friend movieman did run on WCG click here, if you like it iced and overclocked. enjoy! and get ready to crunch, it will be available in store the 17th, in quantity :) Who? |
Luke Send message Joined: 31 Dec 06 Posts: 2546 Credit: 817,560 RAC: 0 |
Wow! What can we expect for the 920 & 940? - Luke. |
Paul D Harris Send message Joined: 1 Dec 99 Posts: 1122 Credit: 33,600,005 RAC: 0 |
Say Who? Why don't you overclock yours? http://setiathome.berkeley.edu/results.php?hostid=4613769 I am sure you can do better than that. I can with my Q6600 G0. |
Francois Piednoel Send message Joined: 14 Jun 00 Posts: 898 Credit: 5,969,361 RAC: 0 |
a ruff estimate is around 25 to 30% faster than Core 2 Quad same frequency. Then, if you overclock, you 'll beat skulltrail ... especially with fast memory. I am running at 1866Mhz DDRIII, at 4.0Ghz :) |
Francois Piednoel Send message Joined: 14 Jun 00 Posts: 898 Credit: 5,969,361 RAC: 0 |
Say Who? if you look carefully, on longer units, it is almost as fast as the Top MACs. remember, it does calculate 8 at a time and it is a single processor. |
Paul D Harris Send message Joined: 1 Dec 99 Posts: 1122 Credit: 33,600,005 RAC: 0 |
Say Who? I saw a lot of 17 credit WU done in about 1200 seconds. You can do better I am sure. My Q6600 does 17 credit WU in about 800 seconds. |
Luke Send message Joined: 31 Dec 06 Posts: 2546 Credit: 817,560 RAC: 0 |
Here a just 3 of the few thousand news stories and reviews on the Core i7.... Intel Core i7 reviews arrive - and yes, it's fast! Intel Core i7 965 Extreme Edition Review Intel Core i7 - Nehalem arrives and FSB departs The latter two just 1 hour old, and the first is just 7 minutes old according to Google News! The Tick has become a Tock.... - Luke. |
Francois Piednoel Send message Joined: 14 Jun 00 Posts: 898 Credit: 5,969,361 RAC: 0 |
|
Francois Piednoel Send message Joined: 14 Jun 00 Posts: 898 Credit: 5,969,361 RAC: 0 |
Say Who? you are doing 4 in 800 seconds, I am doing 8 in 1200 seconds ... in average, Core i7 if much better. you are pushing one out in average ever 200 seconds, Core i7 is pushing one every 150 seconds :) only the dual MACs do better, and on 45 credits, take a look, it equal the MACs with a single processor. |
Paul D Harris Send message Joined: 1 Dec 99 Posts: 1122 Credit: 33,600,005 RAC: 0 |
Yeah more credits in 1200 seconds. |
Francois Piednoel Send message Joined: 14 Jun 00 Posts: 898 Credit: 5,969,361 RAC: 0 |
Yeah more credits in 1200 seconds. exactly, in my computers, one is running stock frequency, the other one at 4.0Ghz, both Core i7 965. |
Paul D Harris Send message Joined: 1 Dec 99 Posts: 1122 Credit: 33,600,005 RAC: 0 |
Hey Who? Are you runing the Intel "Smackover" motherboard with the x58 chip? |
Francois Piednoel Send message Joined: 14 Jun 00 Posts: 898 Credit: 5,969,361 RAC: 0 |
yes, i use smackover. |
Sirius B Send message Joined: 26 Dec 00 Posts: 24879 Credit: 3,081,182 RAC: 7 |
Methinks it's time to move over to Intel....... Francois, will the integrated graphics (2009) hit the performance? |
Francois Piednoel Send message Joined: 14 Jun 00 Posts: 898 Credit: 5,969,361 RAC: 0 |
Methinks it's time to move over to Intel....... I am doing all i can to get good integrated Performance in the GPU. The increase in bandwidth of Core 17 gives it a fair change to be impressive too. |
archae86 Send message Joined: 31 Aug 99 Posts: 909 Credit: 1,582,816 RAC: 0 |
I dusted off Fred W's data vac about three days ago when I noticed crunch3r's post in the new computer thread pointing out two Nehalem SETI hosts. One of those hosts has been running a 64-bit science ap of build 42, the other a 32-bit ap of build 41. Which does better differs by Angle Range: On the possibly false assumption that the hosts were not changed over the period represented, there appears to be a systematic difference. At VLAR, Neh_1 gets lower times, while at VHAR Neh_2 gets lower times. In the crucial central region the points are packed too tightly to make a fair conclusion in this representation. An expanded view suggests strongly that Neh_1 has the advantage from AR 0.432 through 0.447 at the least. Together with the VLAR advantage, this suggest that with the current ap versions, the 64 bit ap from build 42 beats the 32 bit ap from build 41 on these Nehalem hosts. In these views, using straight reported times, the Nehalem hosts generally get longer times than a 2.79 GHz Conroe-class Q6600 Quad and particularly a 2.83 GHz Penryn-class Q9550 Quad. The 64-bit build 42 ap wins at VLAR, but overall this view suggests inferiority in reported time, as has been discussed in this thread. The computer listing on the SETI site shows these as 8-CPU processors. Nehalem is reported to have a modernized hyperthreading capability. I think it very likely that we are looking at the fruit of hyperthreaded operation, so for comparison of Quad-core to Quad-core, it is proper to divide the reported Nehalem CPU times in half. The next two graphs use a secondary y axis to provide this comparison. In the VLAR range the Nehalem results are triumphant, far superior to even the Q9550 results. In the VHAR range the Nehalem advantage is considerably less, but still very substantial (by far more than the rather small probable clock rate difference of approximately 3.2 GHz vs. 2.8 GHz). The portion of the crucially important mid-AR range for which we have comparison also shows utterly compelling Nehalem advantage: Overall, I should mention that the Q6600, E6600, and Q9550 data shown were all recorded with the processors running 3 Einstein and 1 SETI result. So far as I know, the Nehalem results are probably from running 8 simultaneous SETI jobs. Whether or not that is the case, I have strong reason to believe, based on actual measurement on an earlier science ap, that the Q6600 SETI times would have degraded signficantly with three simutaneous SETI jobs, and considerably worse at 4. I expect that the larger cache of the Q9550 reduces this effect, but have not made observations. I expect that Nehalem will degrade less, given the considerable effort and expense that has gone into its considerably different memory access architecture. In summary, my graphs almost certainly understate the (already very large) advantage of a fully SETI-loaded Nehalem over a fully SETI-loaded Q6600. Key questions: 1. what actual clock frequency are the Nehalem hosts running (ap reporting appears to be unreliable, though the central tendency may be toward a slight overclock). 2. were the Nehalem results generated running hyperthreaded? 3. how many SETI aps were actually running simutaneously when these were generated? 4. what fraction of wall clock hours since these two hosts were signed up to SETI have they actually been running SETI? The current RAC seems lower than my optimistic assessment of their capabilities if runnning 8 aps 24x7, especially for the longer-running host 4603078 (Neh_2). |
zoom3+1=4 Send message Joined: 30 Nov 03 Posts: 65773 Credit: 55,293,173 RAC: 49 |
I dusted off Fred W's data vac about three days ago when I noticed crunch3r's post in the new computer thread pointing out two Nehalem SETI hosts. One of those hosts has been running a 64-bit science ap of build 42, the other a 32-bit ap of build 41. Which does better differs by Angle Range: You need o make It more obvious as to which of the two Neh_'s are which, As I had to dig to find out which is which and I'm still not so sure. The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's |
archae86 Send message Joined: 31 Aug 99 Posts: 909 Credit: 1,582,816 RAC: 0 |
I think I see one little flaw in Your very pretty graphs, Which of the two Neh(1 or 2) is 64 bit? Pink or Orange? JokerCPoC (and any others responding), please edit your reply, even if already posted, to remove duplication of all the graphs. Thanks. As to your question, I did document that answer both by providing links to the hosts and in the subtitle. The one I called Neh_1 and consistently displayed as orange diamonds was running 64-bit build 42, and the one I called Neh_2 was running 32-bit build 41. |
Fred W Send message Joined: 13 Jun 99 Posts: 2524 Credit: 11,954,210 RAC: 0 |
The graphs say it all - Neh_1 is 64 bit, Neh_2 is 32 bit. F. [edit]And another nice piece of work by the stats-master[/edit] |
zoom3+1=4 Send message Joined: 30 Nov 03 Posts: 65773 Credit: 55,293,173 RAC: 49 |
I think I see one little flaw in Your very pretty graphs, Which of the two Neh(1 or 2) is 64 bit? Pink or Orange? Well It's not that obvious, I mean Neh_1 and Neh_2 could have been called: Neh_64 Neh_32 More obvious isn't It? The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.