http://www.tweaktown.com/news/55341/amd-radeon-gpu-tech-power-intels-next-gen-igpus/index.html
Here's tweaktown saying the same thing
Here's tweaktown saying the same thing
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
AMD Discussion
- Thread starter koyoot
- Start date
- Sort by reaction score
In march 2017 the deal of licensing patents between Intel and Nvidia is going down. Intel had to get new deal, without Nvidia because of their politics. There will be no AMD GPUs in Intel CPUs.
The deal is allowing AMD to get the Thunderbolt technology on their hands, and other things.
The deal is allowing AMD to get the Thunderbolt technology on their hands, and other things.
http://wccftech.com/amd-zen-16-core-cpu-pictured/
I'm not really sure how they managed to see a 16 core there, but that's no surprise.
I guess this could be an option for the nMP, but hope not yet.
Naples is also a bit far off.
I'm not really sure how they managed to see a 16 core there, but that's no surprise.
I guess this could be an option for the nMP, but hope not yet.
Naples is also a bit far off.
Last edited:
First demos for AMD Zen:
Blender benchmark, Zen @ 3.4 Base clock, and no boost matches i7 6900K at stock clocks 3.2 GHz/3.7 Turbo.
In Handbrake it transcodes 5 secs faster than the 6900K. 54 secs vs 59 secs.
And best part: Zen is 95W CPU, 6900K - 140W.
Blender benchmark, Zen @ 3.4 Base clock, and no boost matches i7 6900K at stock clocks 3.2 GHz/3.7 Turbo.
In Handbrake it transcodes 5 secs faster than the 6900K. 54 secs vs 59 secs.
And best part: Zen is 95W CPU, 6900K - 140W.
AMD really seems to be coming back this time around.
Hope all goes well.
I',m starting to think that Ryzen (Naples) will indeed get into the nMP.
Maybe it's still too early but eventually.
Hope all goes well.
I',m starting to think that Ryzen (Naples) will indeed get into the nMP.
Maybe it's still too early but eventually.
Remember, take TDPs with a giant grain of salt. We won't have any idea of efficiency until reviews are out. Both AMD and Intel often exceed the processors TDP.
Also, one more thing: 8 core config appears to be dual channel only.
Last edited:
I don't understand what that restriction means. Dual CPU configs only? Something to do with memory controllers ("dual channel" RAM)?
Memory controllers.
Mainstream Intel Core i 2 memory controllers ( for max 4 x86 cores )
Workstation/Server Intel Xeon E5-E7 (and HEDT i7 x9xxx) 4 memory controllers from 4 up to 12-18+ x86 cores.
Ryzen 2 memory controllers for 4-8 cores
The Zen L3 cache is a pure victim cache ( stuff bounced from the L2 caches ). Coupled with the "AI" prefetch , SMT , and prediction they have probably counterbalanced much of the memory pressure that 8 cores would incur. As long as it is code and data have relatively highly predictable data locality this can work out.
For workloads like high number of multiple users and/or DBMS thread loads.... probably not as well. However, for single user desktops (w/ one focus app at a time) it should fly for many workloads and a even larger set of simple benchmarks. Even more so for workload coupled to HDDs and/or SATA SSDs.
it is a bit dubious how this scales past 8 core clusters and doesn't incur non linear NUMA effects. A hard core Xeon E5-E7 threat this doesn't look like on high loads.
Last edited:
AMD played VERY good game of smoke and mirrors.
I cannot say right now more. Watch for more information in short future. They have shown a lot, but there is still a lot more to be shown.
If anyone wants to test AMD claims:
https://www.reddit.com/r/Amd/commen...ks/db8hpfe/?context=3&st=iwur5a96&sh=ad479dc3
Here is Handbrake source file, to test it. Blender file is easy to find in the internet.
Edit: Handbrake demo suggest that Zen core have high INTeger code throughput. And that is what is very important in Server market...
Interesting what Apple will do. They can stick with Intel. Zen however appears to be AT LEAST as good as Intel's offerings. But more power efficient, an cheaper.
I cannot say right now more. Watch for more information in short future. They have shown a lot, but there is still a lot more to be shown.
If anyone wants to test AMD claims:
https://www.reddit.com/r/Amd/commen...ks/db8hpfe/?context=3&st=iwur5a96&sh=ad479dc3
Here is Handbrake source file, to test it. Blender file is easy to find in the internet.
Edit: Handbrake demo suggest that Zen core have high INTeger code throughput. And that is what is very important in Server market...
Interesting what Apple will do. They can stick with Intel. Zen however appears to be AT LEAST as good as Intel's offerings. But more power efficient, an cheaper.
Last edited:
I can think of a big reason why Apple could be holding out for Zen. If Apple wants to push the number of TB3 ports in the next mac pro its going to need a lot of bandwidth. More than Intel with Broadwell-EP/Skylake-EP can provide. AMD's Zen/Naples is rumored to have 64 PCIe lanes. That would give them plenty of bandwidth for dual GPUs, dual SSDs, 4 TB3 controllers, etc.
Imagine a 32 core machine with dual Vega GPUs, dual SSDs, 8 USB-C/TB3 ports on the back and can drive 4 5k displays. Thats fun to think about and would make the 4 year wait almost worth it.
Imagine a 32 core machine with dual Vega GPUs, dual SSDs, 8 USB-C/TB3 ports on the back and can drive 4 5k displays. Thats fun to think about and would make the 4 year wait almost worth it.
Naples is 32 core CPU, 4096 GCN core GPU, 16 GB HBM2 APU monster with 180-225W TDP. It will not end up in MBP
.35W quad core APU with 16 CU's and 2-4 GB of HBM2 - that is what we will see if Apple will decide to ditch Intel from their computers. And its called Raven Ridge
.HEDT, 8C/16T CPUs are going to have higher amount of PCIe lanes than 16. There is only one design for Summit Ridge, that is shared between the designs. Only 4C/8T has 16 PCIe lanes.
Also Professional CPUs most likely will start with 16 Core layout, and from that it can be cut down to 12 core design. So most likely highest end MCM will look like this: 16C/32T, 64 PCIe lanes(!), 4 channel memory controller. TDP is most likely 125, 150 and 180W(for APU, most likely).
Actually it is quite interesting alternative. However, in lower thermal envelopes, for example in 35W, the CPU will most likely not be faster than 7700T.
Yes, it may have 3.0 GHz base clock at that setting(we are talking about 4C/8T variant), cost much less, than Intel competitor, but the IPC is 10% below Skylake/Kaby Lake.
But the price, and low power version makes it extremely interesting from my perspective. I have just started a company that designs low-power/high-performance desktop computers
.
Actually it is quite interesting alternative. However, in lower thermal envelopes, for example in 35W, the CPU will most likely not be faster than 7700T.
Yes, it may have 3.0 GHz base clock at that setting(we are talking about 4C/8T variant), cost much less, than Intel competitor, but the IPC is 10% below Skylake/Kaby Lake.
But the price, and low power version makes it extremely interesting from my perspective. I have just started a company that designs low-power/high-performance desktop computers
It's my intent to go with the 8/16 version the only question is do I add a second card which I can use to pass through to VM's or do I sell the processor, MoBo, and 480 and go with Vega. What'll make it even better is I'll be in the States by the time it hit's and I'll be able to pay less and not wait a couple weeks
RX 480 is too weak to pair it with 8C/16T
.But thats my opinion
. Until AMD releases something more powerful, my vote goes to dual GTX 1070 setup.
RX 480 is too weak to pair it with 8C/16T
But thats my opinion
I'm not going to mix green and red..if you do that you get brown