• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

FEATURED AMD ZEN Discussion (Previous Rumor Thread)

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
I think Haswell and Haswell-E IPC are the same per clock. (might be wrong though)...

250MHz (5%, which is the Haswell to Broadwell average IPC delta ;))

More cache, higher memory bandwidth, etc should all lean haswell-e over haswell. Some real world flip flops due to platform differences from what I have seen though. The IPC improvement was also what I was trying to show, remember skylake is another 5% past that and kabylake another 5% past that, most games still aren't using 6/8 core parts so the 6700K and 7700K will still likely show better performance in games. This is unless AMD's XFR really does work as well as I hope and will just kill cores off and push the others much higher in clockspeed. We can only hope at this point that this isn't another let down, with a hypetrain dragging it by. I still feel like they are going to basically release a ivybridge-e chip performance wise, with higher clockspeeds and lower TDP. This isn't really a bad thing as that is a massive leap forward for AMD.
 
I think Kabylake to Skylake is zero IPC improvement. Might be wrong (again), though!

At some point, and above 1080p, those last gen Intel CPU's are over-powerful, and a [email protected] is already sufficient to max oout any last gen high end GPU.

As you say, in any case, it is a huge leap forward for AMD! And I am happy that finally Intel gets some competition!
 
It really depended on application anything for 1%-15% depending on what specifically we are talking about.

Looked around, and it seems that per clock, IPC gain from Skylake to Kaby lake is lower than 1%, whatever the task, as the architecture uis the same.
 
Skylake and Kaby Lake are same IPC. Any differences reported between them are due to the higher stock clocks of KL. Haswell and Haswell-E are also identical in IPC. It is trivial to show by using tasks which fit within L3 cache. Yes, ram bandwidth and cache can affect other real world performance, but it doesn't alter the core architecture. I don't count that any more I than if the hard disk slows down one system over another.

With less than a couple weeks or so until they are expected to be released, I'm hoping to do my own specific testing and see where Ryzen falls into the picture outside the mainstream stuff I'm sure the usual sites will cover before then.
 
Who cares what Intel is doing with IPC?

Top Rumor of the day....

AMD stomps Intel Price to performance at release.
 
Aside from the TLB bug and that they clocked like crap

Hey! I got 3200 MHz out of my (non BE!) Phenom X4. Couldn't get it stable over 2750 MHz, and that number went down over time, but I learned a lot with it since it was my first OC. That was the chip that brought me to OCF. That little guy is still surfing the web, checking email and streaming Netflix to this day. With a 120 GB SSD for Windows 8 the current owner sees no useful performance difference between her rig and all the shiny new toys at Fry's. LOL
 
Hey! I got 3200 MHz out of my (non BE!) Phenom X4. Couldn't get it stable over 2750 MHz, and that number went down over time, but I learned a lot with it since it was my first OC. That was the chip that brought me to OCF. That little guy is still surfing the web, checking email and streaming Netflix to this day. With a 120 GB SSD for Windows 8 the current owner sees no useful performance difference between her rig and all the shiny new toys at Fry's. LOL

I was runnin a an amd64x2 laptop with 4 gigs of ddr2 until recently (2 months ago). I just added a SSD, and it was doing sll the daily stuff flying!
 
I do.
Aside from the TLB bug and that they clocked like crap, what are you getting at?


Performance in leaks nowhere near matched performance off of the shelf or even with an overclock. The 9500 was a great chip for low costs, but it was walked all over by a Q6600, yet prerelease showed exactly what we are seeing right now. The hanger full of salt should be included with every single one of the leaks we have seen so far.
 
Performance in leaks nowhere near matched performance off of the shelf or even with an overclock. The 9500 was a great chip for low costs, but it was walked all over by a Q6600, yet prerelease showed exactly what we are seeing right now. The hanger full of salt should be included with every single one of the leaks we have seen so far.

Well, we're not talking about a Phenom in this thread, but about a $300 CPU that crushes its Intel $600 counterpart ... (Sorry Captain Scotty :D).

All leaked benchmarks show the same thing: get more performance with half or third of your INtel muney! :D

And I am happy, because I gave the blue team money enough the past 5 years!!!
 
Performance in leaks nowhere near matched performance off of the shelf or even with an overclock. The 9500 was a great chip for low costs, but it was walked all over by a Q6600, yet prerelease showed exactly what we are seeing right now. The hanger full of salt should be included with every single one of the leaks we have seen so far.
I don't believe any of the leaks.
Cautiously optimistic I'm being.
I have no doubt that AMD has a winner here, but being an Intel killer has yet to be proven to anybody.
 
Why does it have to be an Intel killer ? View it as healthy competition. That will push both companies to higher standards and new technologies :)
 
Back