Read more.Quote:
At SC18 it showed off performance in several real world HPC/AI applications.
Printable View
Read more.Quote:
At SC18 it showed off performance in several real world HPC/AI applications.
I really would have kept quiet about the Stream Triad test. 50% more memory channels than Epyc, 30% more performance is a dubious win on something that scales really really really well against memory bandwidth.
Edit: Do Intel say whether they had threading turned off on the AMD platform like last time?
I wonder if Intel are aware what the acronym of Cascade Lake Advanced Performance is, do they have it. ;)
Is Hyper Threading enabled on these new glued-together CPUs?
Comparing with last year AMD product when the new one will probably come out around the same time if not sooner.
It is also fun that they compare the "deep learning inference" with their own old product instead of Epyc 7601 like the other comparison...
Should Nvidia be scared too?
I wonder if they disabled SMT for AMD lol, its not too low for intel honestly.
I'm also wondering what is the total cost of ownership difference between the two systems.
I wouldn't be surprised if AMD's setup is 4x cheaper.
Lenovo has leaked 39 different Cascade Lake processor models from the Xeon Bronze, Silver, Gold, and Platinum product lines.
The leak doesn't include the headlining 48C processor, so that will likely be released somewhat later.
As above, there is no mention of Hyper Threading.
It's a bit naff that this pile of knee jerk trash gets Hexus service and the noted IPC of AMD doesn't. Which is far more interesting than a UPI glued dual processor that's overly hot and no business will want because it's literally a stop gap platform.
It sounds like it isn't, presumably because 350W is already more than enough TDP so turning the wick up wouldn't be wise. I presume that is why on previous benchmarks the AMD platform also had threading turned off.Quote:
Originally Posted by afiretruck
Slide 6 in the Intel PDF is where the juicy details lie. SMT is off for Linpack, for the MILC, WRF and OpenFOAM tests they've left it enabled but are only running 1 thread/core, so you're still not getting full performance from the AMD chips. Also "MILC version: 7.8.1 “intel-recipe” branch" and everything possible is compiled with the Intel compiler.
Basically take the numbers with a bucketful of salt.
Quote:
Originally Posted by Corky34
"Experience the power of the CLAP, coming soon from Intel!"
Maybe they're expecting a round of applause. ;-)Quote:
Originally Posted by Corky34
Quote:
Originally Posted by Corky34
Good job they used lake instead of river.