He actually benchmarked it twice. The
first result is exactly the same (save the 2nd save) due to a margin of error.
Quote:and it looks like haswell won't be a big upgrade speed-wise but it'll mostly contain improvements to power management or something like that.
What? Where did you get that from?
As far as I know, the expected increase in performance is around 10% due to the fact that it has
8 execution ports per core compared to 6.
It'll also have
twice the vector processing performance (thanks to AVX2).
Those are the confirmed features thus far afaik.
That's a great deal then. What cooler do you have (if you have one already)?
(03-17-2013, 07:36 AM)Garteal Wrote: [ -> ]He actually benchmarked it twice. The first result is exactly the same (save the 2nd save) due to a margin of error.
Quote:and it looks like haswell won't be a big upgrade speed-wise but it'll mostly contain improvements to power management or something like that.
What? Where did you get that from?
As far as I know, the expected increase in performance is around 10% due to the fact that it has 8 execution ports per core compared to 6.
It'll also have twice the vector processing performance (thanks to AVX2).
Those are the confirmed features thus far afaik.
That's a great deal then. What cooler do you have (if you have one already)?
People at gaf. They may be talking more about general game performance so maybe we could see good increases in emulation land. They're all good people though, got a case, ram and hsf there for free (thermaltake ock frio)
I've been directed there a few times during google searches. From what I've seen like most gaming forums 95% of the active users on that forum are pretentious idiots. Every time I get directed there the thread is full of people posting factually incorrect popular opinions and circlejerking each other. Never does anyone step in to correct them and if someone does they are attacked for essentially breaking the peace. And they never back up there ridiculous claims with any evidence. I wouldn't take too much stock in their "predictions" considering they are almost always based on whatever marketing nonsense is spouted at them and end up being wrong at least half the time. They may be nice and orderly but they are mostly idiots. Even the extensive moderation and difficult entry requirements doesn't change that.
Garteal Wrote:It'll also have twice the vector processing performance (thanks to AVX2).
Under perfect conditions, which is impossible in real world conditions. Not to mention only certain types of vector code will benefit.
Garteal Wrote:As far as I know, the expected increase in performance is around 10% due to the fact that it has 8 execution ports per core compared to 6.
Among other things. Many of which are listed in the image you posted. 10% is a very conservative estimate. I have my doubts that it will be that low but then again what do I know.
(03-17-2013, 03:26 PM)NaturalViolence Wrote: [ -> ]I've been directed there a few times during google searches. From what I've seen like most gaming forums 95% of the active users on that forum are pretentious idiots. Every time I get directed there the thread is full of people posting factually incorrect popular opinions and circlejerking each other. Never does anyone step in to correct them and if someone does they are attacked for essentially breaking the peace. And they never back up there ridiculous claims with any evidence. I wouldn't take too much stock in their "predictions" considering they are almost always based on whatever marketing nonsense is spouted at them and end up being wrong at least half the time. They may be nice and orderly but they are mostly idiots. Even the extensive moderation and difficult entry requirements doesn't change that.
Are you talking about neogaf? Yeah, I do see a lot of whack discussions and comments popping up there so I sort of agree with you. It kind of makes you think about it especially when registration is a tough thing in that site (mine took like a year or 2 before it got validated lol). The PC building folk though are/were really helpful so it isn't all that bad imo.
(03-17-2013, 03:26 PM)NaturalViolence Wrote: [ -> ]Among other things. Many of which are listed in the image you posted. 10% is a very conservative estimate. I have my doubts that it will be that low but then again what do I know.
Isn't ivy supposed to be 5% or more faster clock for clock than sandybridge? This benchmark doesn't really show that so I don't know. Hopefully my ordered parts arrive soon so I can start trying it out on my own.
georaldc Wrote:Isn't ivy supposed to be 5% or more faster clock for clock than sandybridge? This benchmark doesn't really show that so I don't know. Hopefully my ordered parts arrive soon so I can start trying it out on my own.
10%. But sandy bridge can OC about 10% higher on the same cooling system so if you're a serious OCer (which I gather that you are) then you are unlikely to see any performance difference between the two. However users who run their cpus at stock will see a 10% jump in performance.
(03-17-2013, 04:20 PM)georaldc Wrote: [ -> ]Isn't ivy supposed to be 5% or more faster clock for clock than sandybridge? This benchmark doesn't really show that so I don't know. Hopefully my ordered parts arrive soon so I can start trying it out on my own.
For what this bench is I think its great but I don't think you can really use it to tell if ivy is really 5% or more faster than sandy. Lots of different users with different setups contributed. There is lots of room for human error or some underlying system problem, like the apu that has recently been removed. There is discrepency between boat in view vs out of view for save 1 and save 3. Some have said it didn't make a difference. It certainly does on my system. On my 2500k @ 4.5ghz I get in save 1 84 fps with the boat on screen, right under the 2500k @4.6 that was posted, and 94 min fps once the boat leaves the screen, right under Garteal's 2500k @4.5ghz. The situation is similar on save 3, I get about a 10 fps drop once the boat comes on screen. 112 fps down to ~102. To add on to that is each user trying to screenshot the min, the max, or the average fps? I guess what I'm trying to say is there is at least a decent sized margin of error here. I haven't posted my overclocked 2500k results because I didn't really think the thread needed another one but I will if there is demand for it.
me and my friend did benchmarks like the new 3DMark and my 3570k @ 4.5ghz match his 2500k @ 4.8ghz on several occasions
his highest is 5ghz like mine and i beaten his scores slightly at 4.8ghz
Dolphin might just be one of those applications that doesn't see a significant increase in IPC from the improvements made in ivy bridge. Not every application does. Also I notice a 2500K result listed as 3.4GHz, it is likely running at 3.7 GHz with turbo boost. If you're running at stock speeds with turbo boost please remember to look up and list the clock rate that the cpu is actually running at instead of the default clock rate. You can look it up with CPUZ or HWINFO64. With turbo core cpus just list the default clock rate since the correct frequency can't be measured with software (currently, that may change in the future).
(03-18-2013, 06:37 AM)NaturalViolence Wrote: [ -> ]Also I notice a 2500K result listed as 3.4GHz, it is likely running at 3.7 GHz with turbo boost.
I believe that is my result you are talking about NV. If it is turbo boosting Higher than 3.4 than CPUz isn't reporting it. I just checked it again and like last time it Turbos to 3.4, from stock 3.3, as soon as the game starts with occasional, not enough to effect anything IMHO, spikes to 3.5.
Then list 3.5. Still that's surprisingly low. It should be turboing much higher than that.