sblantipodi
2[H]4U
- Joined
- Aug 29, 2010
- Messages
- 3,765
Single GPU.
is there a link where [H] shows a big bottleneck with R290 using PCI Exp 2.0 8x?
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Single GPU.
is there a link where [H] shows a big bottleneck with R290 using PCI Exp 2.0 8x?
I dunno
I was pointing out that the 290X card is single GPU.
290 series has no crossfie connector so it pushes all data through PCI-express
SLI should be much less depend on pci-ex bandwidth.
Do, what I just did - ordered Z77 motherboard. I got the native USB 3.0, more SATA controllers, better internet chip. And I'm able to keep my 2500k
Do, what I just did - ordered Z77 motherboard. I got the native USB 3.0, more SATA controllers, better internet chip. And I'm able to keep my 2500k
at this rate intel will fade in obsolescence...
LOL what are you talking about?...
ARM is gaining on them fast. If they don't get their act together, they will be irrelevant. It's all about the rapid gains in speed of AMR vs the turtle pace of Intel.
yes bud.. keep dreaming.. thats not gonna happen ARM its RISC.. they are FAR AWAY from actually truly desktop processors.. if you talking about the mobile market then you could be right.. maybe! but do not talk shit about ARM in a intel 2011 sandy bridge thread without any knowledge of what are you talking about, claiming something without any clear answer or any solid info. only with a "at this rate intel will fade in obsolescence..." being actually intel haswell mobile chips the king of the performance in the mobile market.. O_O
The CISC vs RISC debate? That's your argument? That's like saying China is not a large economy because its people make $500/month. There are 1.3 BILLION of them! More RISC vs powerful CISC... Anyway, one is moving fast, the other is scratching his balls. Time is not on Intel's side.
Last time i checked Intel was getting a lot of design wins with their new family of atoms for android shortly after they got serious about market.
When they bring 14nm Atom chips it's ARM who will be losing market share.
3 years on and my 2500K is still relevant and I don't find myself even close to wanting more CPU power.
Honestly though, I'm not sure what's better in terms of awesome for the time, my 2500K or the Q6600 it replaced. I feel spoiled, having those two chips in a row.
Agreed, that is all I have replaced over the past 3 years too.Yeah, almost feeling guilty. Going into year 3 with my current(sig) rig and see no real reason to upgrade anything but the GPU..maybe..
Yeah, almost feeling guilty. Going into year 3 with my current(sig) rig and see no real reason to upgrade anything but the GPU..maybe..
+1
Very happy with my 2600k. Amazed that I have made it this long with no need to upgrade, I remember in one 16 month period I upgraded my CPU 3 times! On water no less. At this point I'm starting to miss it.
+1
Very happy with my 2600k. Amazed that I have made it this long with no need to upgrade, I remember in one 16 month period I upgraded my CPU 3 times! On water no less. At this point I'm starting to miss it.
Sandy will probably go down as the most long lived chip from Intel ever at this rate. Im extremely happy I jumped on this bandwagon.
If the 2nd Gen 2500/2600K's are still this viable.. I imagine my 3770K will do just, if not better, as well. Glad I ponied up the cash.
I've been wanting to upgrade, but there's really no way to justify retiring the ol' 2500k. It's still cranking along at 4.6GHz for several years now. Next upgrade for me will likely be new Maxwell video cards later this year. I don't see a new processor anytime in the future unless someone comes along and gets Intel off their ass. I don't see any change to them milking these marginally faster CPUs for all they're worth... which is always what happens when there's no competition.
Curious if this is normal or if this chip is damaged...
I recently bought a used i7-2600k and I just got it in my machine for a light OCing and noticed that "core1" is not utilized fully or at all at time in Prime95. Is this normal? Has anyone experienced this? Using 8 threads.
Using CoreTemp to measure temps and activity.
EDIT: Actually only does it at really high OCs.. i didnt realize it could throttle only 1 core at a time. When I back down multiplier one notch I see 100% utilization for a minute then backs down to mid 50%s. Indicating only one thread in that core. At stock I get none of this, its 100% all the time.
it can happen in a single thread.. or in multiple threads you should see that in one of the 8 little windows of P05FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
its normal.. that its because the Prime95 working has stopped working due to errors in the overclocking and making it unstable, you need to bump the voltage a little more.. even a minimal 0.005V bump can help with those instability issues..
This its an example of the kind of instability errors you can find (in able to replicate this anytime this its why i choose this): it can happen in a single thread.. or in multiple threads you should see that in one of the 8 little windows of P05
Debating on increasing voltage... I like to do "green overclocking" where you only bump up voltage one notch over stock and get the highest stable OC. Its looking like its stable at 4.5GHz.. might try to give it some more voltage just to see how far it can go but usually Im happy with low voltage OCing for 24/7 usage.
thats the best way to overclock in fact the minimal voltage needed its what help to keep the chip healthy longer time.. you might want to try the use Offset voltage for that task, im using that way of overclocking since my first sandy, low voltage when idle, low voltage at regular loads and good voltage when the chip fully needed i found that the best for daily use specially if you are able to even dominate a good negative offset because you will be using even less power at idle and low powered task like browsing etc..