• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Which will come first? Or at all?

D

DMoshington

Weaksauce
Joined
Apr 16, 2009
Messages
94
Actual 8 core, retail purchaseble I7 processors
or
Dual socket i7 motherboards? Is there even a possibility for that?

My workstation is a Dual quad core (2 x QX9775) on a d5400xs intel board.

I'm of the belief that an 8way i7 or dual quad i7s (or even better, dual 8way i7s...:cool: ) would be faster..

Which can I hope for first?

Will they both be out eventually?
 
DMoshington said:
Xeon, Not i7

I've already got almost equivelant of this.
Click to expand...

Xeon, i7, same thing, just like with the C2 chips, they are still Nehalem based dual socket 1366 chips. Only thing is that these boards are not gaming boards and so do not have 2+ full speed PCI-E 2.0 slots, but then again, neither does the Skulltrail.
 
No performance / compatibility issues in o/s and application usage?
The current 4way - Xeon's are "i7" based, not "core2quad" based?
 
DMoshington said:
No performance / compatibility issues in o/s and application usage?
The current 4way - Xeon's are "i7" based, not "core2quad" based?
Click to expand...
There are some Xeon's that are Core 2 based, and there are some i7(Nehalem) based.
http://en.wikipedia.org/wiki/List_of_Intel_Xeon_microprocessors#Nehalem_based_Xeons

I think Intel's roadmap last I saw it was 8 core(16thread) cpu's in workstation class Xeon CPUs only. I don't think that will be phasing down to consumer/enthusiast line for this generation. See link:
http://en.wikipedia.org/wiki/Intel_Nehalem_(microarchitecture)#45_nm_processor_architecture
(not much yet is known about the 32nm Westmere refresh, but I doubt anything major different other than process shrink and maybe cache increase + bug fixes and/or new SSE instructions)

The Core i7, consumer/enthusiast line, likely has SMP disabled like for their previous generation chips, so dual socket i7 motherboards will not be possible.
 
Last edited:
I think I'm going to consider http://www.newegg.com/Product/Product.aspx?Item=N82E16813182186 The supermicro. It's the closest thing on the list to a "desktop" board. It's got dual PCI-e x16s, 3 pcis, onboard 8 channel audio (I don't need much sound, but I'd like SOME) and has 12 memory slots. So, with the AVAILABLE memory that's easy to find (ddr3 1333 @ 2GB/stick) it's easy to get to 24gb.. and "in theory" you could get, someday, some 4gb or 8gb sticks.
Dual x5560s and 12 sticks of DDR3-1333 @ 7-7-7-20 should do nicely?

As it's a server board, I can't seem to see if it needs special ram....
 
Last edited:
DMoshington said:
I think I'm going to consider http://www.newegg.com/Product/Product.aspx?Item=N82E16813182186 The supermicro. It's the closest thing on the list to a "desktop" board. It's got dual PCI-e x16s, 3 pcis, onboard 8 channel audio (I don't need much sound, but I'd like SOME) and has 12 memory slots. So, with the AVAILABLE memory that's easy to find (ddr3 1333 @ 2GB/stick) it's easy to get to 24gb.. and "in theory" you could get, someday, some 4gb or 8gb sticks.
Dual x5560s and 12 sticks of DDR3-1333 @ 7-7-7-20 should do nicely?

As it's a server board, I can't seem to see if it needs special ram....
Click to expand...
It supports the ECC memory, but I don't know if it *requires* it. Hopefully it's optional but I would try to confirm.
 
The xeon 3500 and 5500 xeons are almost identical to core i7s. You'll probably never see core branded dual socket motherboards.
Beckton (8 core nehalem) will be 4+ socket xeon only, as well.
We'll see gulftown (6 core s1366, replacing bloomfield) in q1 2010. Dont expect anything more than 6 cores for consumers until sandy bridge, around Q4'2010 - Q1'2011.
 
Last edited:
InvisiBill said:
Can you run KribiBench on your system to get some numbers? Use the "Sponge Explode" and "City" model tests to compare with http://www.guru3d.com/article/intel-core-i7-920-and-965-review/15. The i7 920 shows about a 40% gain over the QX9770. I'd be interested in seeing what 2xQX9775 puts out.
Click to expand...

I don't know if i'm running the test right, but is it supposed to only take a few seconds? Guess I'm used to the sort of stuff from futuremark that runs for 20 minutes.

SpongeExplode gave me 18.22522
City gave me 16.39112

I've got to be running this wrong, right? This benchmark only takes like 5 seconds?
 
Last edited:
DMoshington said:
I don't know if i'm running the test right, but is it supposed to only take a few seconds? Guess I'm used to the sort of stuff from futuremark that runs for 20 minutes.

SpongeExplode gave me 18.22522
City gave me 16.39112
Click to expand...

if it means anything..
Squadron gave me just under 6 and Ultra gave me just over
 
hmmm...

No way my Q8200 is faster than a Core I7...

though it is running at 3.1Ghz with memory at 1063.

that I7 is running at 3.6ghz

I got

sponge explode: average frame rate : 16.3757 fps
City: average frame rate : 16.1117 fps
 
soccrstar said:
hmmm...

No way my Q8200 is faster than a Core I7...

though it is running at 3.1Ghz with memory at 1063.

that I7 is running at 3.6ghz

I got

sponge explode: average frame rate : 16.3757 fps
City: average frame rate : 16.1117 fps
Click to expand...

Just means that you and I must be running the test wrong.
Your number seems to high, maybe mine is too low.

There's sure as hell no way your q8200 is going to come close to dual qx9775 on an app that supposedly "loves SMP"
 
DMoshington said:
I don't know if i'm running the test right, but is it supposed to only take a few seconds? Guess I'm used to the sort of stuff from futuremark that runs for 20 minutes.

SpongeExplode gave me 18.22522
City gave me 16.39112

I've got to be running this wrong, right? This benchmark only takes like 5 seconds?
Click to expand...


soccrstar said:
hmmm...

No way my Q8200 is faster than a Core I7...

though it is running at 3.1Ghz with memory at 1063.

that I7 is running at 3.6ghz

I got

sponge explode: average frame rate : 16.3757 fps
City: average frame rate : 16.1117 fps
Click to expand...

They got 11s and 12s on the i7 with HT on, 9ish with HT off (which is in line with what I got too). I'd say that 16-18 for 2xQX9775 seems about right, assuming it scales really well with SMP like they say. The single QX9770 got ~8.5.

The Q8200 at 16 seems really high though. I don't know how it could score almost double what the QX9770 got (and similar to the 2xQX9775).
 
Just for fun and to make everyone appreciate their hardware just a little bit more, my 2 x 2.66 GHz Pentium 4-based Xeon box gets

5.14 fps in Sponge Explode
4.51 fps in City
 
DMoshington said:
Just means that you and I must be running the test wrong.
Your number seems to high, maybe mine is too low.

There's sure as hell no way your q8200 is going to come close to dual qx9775 on an app that supposedly "loves SMP"
Click to expand...

InvisiBill said:
They got 11s and 12s on the i7 with HT on, 9ish with HT off (which is in line with what I got too). I'd say that 16-18 for 2xQX9775 seems about right, assuming it scales really well with SMP like they say. The single QX9770 got ~8.5.

The Q8200 at 16 seems really high though. I don't know how it could score almost double what the QX9770 got (and similar to the 2xQX9775).
Click to expand...

Maybe I am running this wrong as well, but I got,

City: 17.8938
Spongeexplo.d: 17.1339

And my rig is down clocked right now from the 3.8 in my sig to 3.4.

*EDIT*

Just retested at 3.8

City.d: 18.1689
Spongeexplo.d: 18.2769
 
Last edited by a moderator:
I'm just launching Kribi, opening city.d or spongeexplo.d, then hitting the Test button. I'm not really sure how you can do it wrong, but I'm not sure how you could get such high numbers either. *shrug*

106.955 M polygons
2 directional lights
city.d / flat [ Kribi 1.1.9 ]
average frame rate : 11.3145 fps

19.2001 M polygons
2 directional lights
spongeexplo.d / flat [ Kribi 1.1.9 ]
average frame rate : 10.4719 fps

i7 920, completely stock settings, HT disabled. When I tested before, it was a little lower. Could be due to dnetc slowing things down even though it runs at 0 priority, or something to do with my new hard drive and fresh Windows install.

I just noticed that my link was messed up all along. Maybe some people are getting a different version from another source?
 
I am getting mine from www.adeptdevelopment.com. I am also running the 1.1.9v.

Just did the test again to get some caps.

icf57i.jpg


if6.jpg


*shrugs*
 
You must log in or register to reply here.
Back
Top