"scientific tasks" is the most vague thing I have ever heard in my life. In terms of CPU instructions that could be literally anything. What is it that a CPU does that isn't a "scientific task"? Painting a work of art? cutting the grass?
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.
Colleagues of mine were at a Dell EMC event last week at Shepperton Studios hosted by SNS and there was a lot of pushing EPYC products from DELL.. ive not yet noticed on the Dell Premier Portal much in the way of Ryzen in Business Clients yet though... been keeping my eyes peeled... However my Boss has said a few days ago, our Central IT (I work for a worldwide FTSE 500 Manufacturing company) has started pushing why arent we looking at AMD / Zen / EPYC etc.. so i can imagine before long we will see a shift.. Money men make the decisions after all.
"scientific tasks" is the most vague thing I have ever heard in my life. In terms of CPU instructions that could be literally anything. What is it that a CPU does that isn't a "scientific task"? Painting a work of art? cutting the grass?
13% seems really low gains for a move from "12nm" to a 7nm node.
About 6 years worth of Intel generations... Not sure how you correlate process to IPC.
A 13% gain with even a minimal boost in clockspeed will result in a significant performance uplift and put AMD a long way in front of Intel.
Word on the street though is that minimal boosts in clock speed are about all it will manage over current Ryzen CPUs. I suppose if we get a 3700x @ 4.5ghz it will do well enough.
About 6 years worth of Intel generations... Not sure how you correlate process to IPC.
A 13% gain with even a minimal boost in clockspeed will result in a significant performance uplift and put AMD a long way in front of Intel.
Intel has really not been trying though - really shouldn't be advocating pace at their rate. The difference between TSMC's 7nm which by all reports is shaping up decently and the 12nm process Ryzen CPUs are currently on is vast - more than the jump from Intel's 22nm to 14nm+++++++++ or whatever variant it is now. Which should give huge scope for pipeline efficiency especially in terms of reduced latency which would help with prediction misses, etc.
Intel is trying very hard. You can't compare nodes from different firms. TSMC is focused on a mobile CPU production for some time, obviously AMD will influence have huge a influence at TSMC but that remains to be seen. You don't need a new process to improve your chip design. It literally has no relevance to this topic.
Pretty much everything you said other than the bit in bold is wrong.
To you maybe. Clearly you know Intel are trying hard and you can't compare nodes, so everything you typed is wrong...
Intel have done very little over the past ten years.
Also, if daddy Intel hasn't been able to improve IPC beyond 5% since 2014, how can AMD do it?.
Of course. Except it's not really much of a lower base. 154 pts vs 146 pts isn't a huge difference in IPC in 3.5GHz normalised Cinebench testing. So a 13% increase (best case scenario) would be 165pts for Zen 2. That's 11 points higher than Kaby Lake and Coffee Lake. With 4.6GHz on the clock it's faster than equivalent Intel at 5.1GHz and the same core counts....If you start from a lower base it's easier to make bigger % improvements.
Which is it?
With 4.6GHz on the clock it's faster than equivalent Intel at 5.1GHz and the same core counts....
It's both.
For example, I try really hard to improve at Rocket League, however I've improved little.
Doesn't fit this context his second quote says they've done very little - which is what I was saying - without much competition over the last 10 years they've not particularly pushed themselves to wring out IPC gains so not really much of a metric for what good progress should look like. They might be trying now but that is irrelevant.