• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Ryzen Sandy Bridge IPC

I agree with you moun.

i have a 4770k / 1080 ti on a 165hz gysnc montor. i like to game as near to 165hz @ 1440p. At present there's not much of an upgrade path it seems - which is a good thing seems as though im saving to buy a house :p
 
muon said:
Just goes to show that unless you are using a GTX 1080Ti at 1080p, a lot of old processors are good enough when overclocked, especially if it has 8 threads.
Click to expand...
Yep, which is why the whole "but an i7-7700K is worth an extra £200+ over an R5 1600 cos it's 20% faster at max clocks" argument is rather silly for 99.9% of people.
 
DragonQ said:
Yep, which is why the whole "but an i7-7700K is worth an extra £200+ over an R5 1600 cos it's 20% faster at max clocks" argument is rather silly for 99.9% of people.
Click to expand...

But 5 people out of a thousand want to run 1080p at 144hz! So Ryzen must be crap.
 
I found the 5.625ghz SB vid whilst I was at work and didn't want to post without seeing if it had swearies or not. I will post it up once my daughter is in bed and I have had a chance to check for naughty language....

Getting to that level though is extremely rare, and if comparing then make sure you compare against.the new chips that are OC'd.
 
I did some tests from my Sandy to Ryzen. Things like SuperPi showed a modest IPC increase (~15%), whereas Cinebench and CPU-Z were much higher, more like 25%+.

Reading through this thread, its very apparent that many people don't understand what IPC - Instructions Per Clock actually means (they miss out the PER CLOCK part.) Due to Ryzen's architecture, its largely irrelevant as its so variable.

Either way, despite my Sandy always showing decent frame-rates, my Ryzen build feels far faster - all games now feel silky smooth and i don't regret it one bit.
 
EdButler said:
I did some tests from my Sandy to Ryzen. Things like SuperPi showed a modest IPC increase (~15%), whereas Cinebench and CPU-Z were much higher, more like 25%+.

Reading through this thread, its very apparent that many people don't understand what IPC - Instructions Per Clock actually means (they miss out the PER CLOCK part.) Due to Ryzen's architecture, its largely irrelevant as its so variable.

Either way, despite my Sandy always showing decent frame-rates, my Ryzen build feels far faster - all games now feel silky smooth and i don't regret it one bit.
Click to expand...

In Cinebench Ryzen has a higher IPC than Haswell, because AMD SMT is better than Intel it even scores slightly higher per clock than Kabylake in MT.
 
@EdButler I don't think anyone is doubting that Ryzen has higher IPC than Sandy, the difference should be fairly large.
But moving past IPC, the issue is that Sandy Bridge can still have faster single threaded performance because it can reach much higher clocks than Ryzen can, so that IPC advantage Ryzen has gets overtaken by the clock advantage of K Sandy's.
 
AndreiD said:
@EdButler I don't think anyone is doubting that Ryzen has higher IPC than Sandy, the difference should be fairly large.
But moving past IPC, the issue is that Sandy Bridge can still have faster single threaded performance because it can reach much higher clocks than Ryzen can, so that IPC advantage Ryzen has gets overtaken by the clock advantage of K Sandy's.
Click to expand...

Can be in some situations. In others Ryzen can beat Kaby.
 
AndreiD said:
@EdButler I don't think anyone is doubting that Ryzen has higher IPC than Sandy, the difference should be fairly large.
But moving past IPC, the issue is that Sandy Bridge can still have faster single threaded performance because it can reach much higher clocks than Ryzen can, so that IPC advantage Ryzen has gets overtaken by the clock advantage of K Sandy's.
Click to expand...

Same goes for Ivy Bridge and Haswell given Sandy can also clock higher than those.
 
humbug said:
In Cinebench Ryzen has a higher IPC than Haswell, because AMD SMT is better than Intel it even scores slightly higher per clock than Kabylake in MT.
Click to expand...

SMT broke the call of duty ww2 beta the other day, was a stuttering mess.
Turned it off and all was well. It also performs better in BF1 with SMT turned off, no idea what causes this but generally games work better with it turned off.
 
You system is always breaking so its hardly surprising you have yet more problems with it ^^^^

image.png


Sandy > Ivy = 3%
Ivy > Haswell = 9%
Haswell > SkyLake / KabyLake 8%

So from Sandy to Haswell = 12%
Sandy to Kabylake = 20%
 
humbug said:
You system is always breaking so its hardly surprising you have yet more problems with it ^^^^

image.png


Sandy > Ivy = 3%
Ivy > Haswell = 9%
Haswell > SkyLake / KabyLake 8%

So from Sandy to Haswell = 12%
Sandy to Kabylake = 20%
Click to expand...

Doesn't this put the Intel's IPC has only increased due to ddr4 argument to bed?
6700k is faster than the 4770k even with ddr3.
Granted it's not much but this shows the increase does not come from dd4 alone.
 
@humbug you ignored the entire article page I linked and you extrapolated IPC in Cinebench as being the norm, is it really that difficult to read to the bottom of the page where they actually give you a % jump in IPC based on several workloads?

Literally at the bottom of the page:
Sandy Bridge to Ivy Bridge: Average ~5.8% Up
Ivy Bridge to Haswell: Average ~11.2% Up
Haswell to Broadwell: Average ~3.3% Up
Broadwell to Skylake (DDR3): Average ~2.4% Up
Broadwell to Skylake (DDR4): Average ~2.7% Up
Haswell to Skylake (DDR3): Average ~5.7% Up
 
Typical humbug response, if something doesn't meet his expectations then it is broken.
If you had one you would see for yourself, some games don't like SMT.
 
You must log in or register to reply here.
Back
Top Bottom