AMD 8-core ZEN Packs a Whallop with Multithreaded Performance

[64K]

I think that situation is changing, Lisa Su is (AFAIK) an engineer. Also, I think part of AMD's issue when they were very competitive was actually a lack of marketing. So if they can restore the balance; get out good products but also make sure people know about them then things will improve.

I started a course recently and 90% of the folks in it who have been looking for "back to school" laptops are purely asking the question "Should I buy an I5 or I7" they have no clue about the rest of the system nor balance nor the importance of a good screen etc.

So the marketing dept definitely has a place; if it is used to (honestly) educate the consumers it's all good IMO.

Yes, Lisa Su has a Doctorate in Electrical Engineering from MIT. You don't see many of those. She's a very smart person.

 

[bug]

I think that situation is changing, Lisa Su is (AFAIK) an engineer. Also, I think part of AMD's issue when they were very competitive was actually a lack of marketing. So if they can restore the balance; get out good products but also make sure people know about them then things will improve.

I started a course recently and 90% of the folks in it who have been looking for "back to school" laptops are purely asking the question "Should I buy an I5 or I7" they have no clue about the rest of the system nor balance nor the importance of a good screen etc.

So the marketing dept definitely has a place; if it is used to (honestly) educate the consumers it's all good IMO.

Well, there's no AMD in the Ultrabook market which is all people are interested these days, so of course everybody's looking for i5 or i7. That's why I say Zen has to do a whole lot more than win or come close in a number of benchmarks.

 

[Audiophizile]

I would also add that we are now moving into a post single core era. Take BF1's scaling across multiple cores for example.

My projection is that AMD will have about a 10% IPC deficit against Intel, but will counter that with aggressive core count pricing. Whilst postulating about server pricing might be a bit rich, the easiest example I can think of is pushing an 8 core part against 1151 4 core parts.

Still, hoping for some competition!

A 35w dual core i3 6100T@3.2Ghz beats a 125w 8core fx-8370@4.3GHz in battlefield 1...

 

[cdawall]

A 35w dual core i3 6100T@3.2Ghz beats a 125w 8core fx-8370@4.3GHz in battlefield 1...

And? Does one poorly coded game make or break a CPU?

 

[Audiophizile]

And? Does one poorly coded game make or break a CPU?

Well the game runs exceptionaly well on not top of the line hardware. One of the most optimized games I've played, more so than bf4 or any recent AAA title I've played. I was quoting his point that bf1 scales well across multiple cores and showing him a dual core beats an 8 core.

 

[cdawall]

Well the game runs exceptionaly well on not top of the line hardware. One of the most optimized games I've played, more so than bf4 or any recent AAA title I've played. I was quoting his point that bf1 scales well across multiple cores and showing him a dual core beats an 8 core.

It is optimized for high single core IPC. That makes it poorly coded to anyone who doesn't own an i whatever. There is literally NO excuse for a game released now to for one not be coded correctly in DX12 or vulkan, and on top of that for it not to be multithreaded well is an absolute joke.

and I say correctly coded because DX12 removes the majority of CPU overhead or at least it should. So good job on them.

 

[Audiophizile]

It is optimized for high single core IPC. That makes it poorly coded to anyone who doesn't own an i whatever. There is literally NO excuse for a game released now to for one not be coded correctly in DX12 or vulkan, and on top of that for it not to be multithreaded well is an absolute joke.

and I say correctly coded because DX12 removes the majority of CPU overhead or at least it should. So good job on them.

So you hate almost single every game in 2016?

 

[cdawall]

So you hate almost single every game in 2016?

Only the ones that can't multithread and there are several that function fine...

 

[Camm]

A 35w dual core i3 6100T@3.2Ghz beats a 125w 8core fx-8370@4.3GHz in battlefield 1...

Go bench on DX12 and not 11...

 

[Audiophizile]

Only the ones that can't multithread and there are several that function fine...

List the ones that perform better on multi thread vs faster single core please.

 

[Audiophizile]

Go bench on DX12 and not 11...

You mean the garbage that is dx12 in bf1? Bf1 does terrible in dx12. No getting around it.

 

[cdawall]

List the ones that perform better on multi thread vs faster single core please.

List ones what? Games? Most games now are coded for a console that uses 6/7 of 8 amd jaguar low power cores. If having a strong single core IPC was that important don't you think those money whore's would have used something else? Multithreading is the way of now. Single threading is lazy coding and comparing one poorly coded game is stupid.

But since you asked

That's the basis for vulkan.

 

[Melvis]

Wasnt Crysis 3 one of those games that really made use of the 8cores well and in a few cases topped the FPS chart over the current i7 at the time (3770K) ?

I rarely see games these days still (not that I have many modern titles) that use all 8 cores well.

 

[bug]

It is optimized for high single core IPC. That makes it poorly coded to anyone who doesn't own an i whatever. There is literally NO excuse for a game released now to for one not be coded correctly in DX12 or vulkan, and on top of that for it not to be multithreaded well is an absolute joke.

and I say correctly coded because DX12 removes the majority of CPU overhead or at least it should. So good job on them.

You seem to think every piece of code is inherently parallelizable and the more threads you throw at it, the faster it runs. That's not true.
Not only is code hard to parallelize, even if you manage to do it, you still have to synchronize some actions, thus reducing some (but not all) of the benefits of threading.

Add to that that multithreaded code is by default harder to develop and test and you may start to understand why more cores is not an automatic win.

 

[Audiophizile]

List ones what? Games? Most games now are coded for a console that uses 6/7 of 8 amd jaguar low power cores. If having a strong single core IPC was that important don't you think those money whore's would have used something else? Multithreading is the way of now. Single threading is lazy coding and comparing one poorly coded game is stupid.

But since you asked

That's the basis for vulkan.

You realize this shows the opposite of your point? It shows 4 core i5s and 4c/8t i7s beating $1000 6c/12t i7s and amds 8c. Also, its not a game from 2016... So again, please show me these games.

 

[cdawall]

You realize this shows the opposite of your point? It shows 4 core i5s and 4c/8t i7s beating $1000 6c/12t i7s and amds 8c. Also, its not a game from 2016... So again, please show me these games.

Look again there is a 3 yes a 3 fps difference between all of those chips from top to bottom and the 4590 is far from the highest clocked. The point was to show a well coded game doesn't care. That's the point of dx12/Vulcan (mantle).

You seem to think every piece of code is inherently parallelizable and the more threads you throw at it, the faster it runs. That's not true.
Not only is code hard to parallelize, even if you manage to do it, you still have to synchronize some actions, thus reducing some (but not all) of the benefits of threading.

Add to that that multithreaded code is by default harder to develop and test and you may start to understand why more cores is not an automatic win.

How about this, I already said coders are lazy. Which fits exactly what you said. I don't care if it is harder, the hardware exists and they get paid plenty to utilize it. Hence my hinting towards consoles. They managed to get some damn good looking games to run on the PS3's weird ass architecture, but they can't convert an x86-64 platform to an x86-64 platform? The Xbox runs on a modified Windows 10 now?! Somehow games can use 6 or 7 cores fine on it. Lazy coders are why pc gaming is on its last legs. Don't stand up for those lazy fucks.

 

[bug]

How about this, I already said coders are lazy. Which fits exactly what you said. I don't care if it is harder, the hardware exists and they get paid plenty to utilize it. Hence my hinting towards consoles. They managed to get some damn good looking games to run on the PS3's weird ass architecture, but they can't convert an x86-64 platform to an x86-64 platform? The Xbox runs on a modified Windows 10 now?! Somehow games can use 6 or 7 cores fine on it. Lazy coders are why pc gaming is on its last legs. Don't stand up for those lazy fucks.

I don't know whether to laugh or cry.

 

[cdawall]

I don't know whether to laugh or cry.

Got another way to explain constant bad console ports?

 

[Audiophizile]

Look again there is a 3 yes a 3 fps difference between all of those chips from top to bottom and the 4590 is far from the highest clocked. The point was to show a well coded game doesn't care. That's the point of dx12/Vulcan (mantle).

I agree that it should help lower end hardware perform better when implemented properly. It is extremely rare that it happens and is sad. Usually games I play don't run worse on PC, if youre looking for way better than console graphics or constant 100+fps is where you start to see more problems. If every PC game was benchmarked at console level graphics and resolution you'd see REALLY impressive numbers all across the board. Hardware does and should make at least some difference though. Speed of a CPU, memory or GPU should all at least slightly impact a games performance. Obviously GPU being the most difference.

 

[msimax]

http://www.guru3d.com/articles_pages/battlefield_1_pc_graphics_benchmark_review,9.html

http://www.guru3d.com/articles_page...e_warfare_pc_graphics_benchmark_review,8.html

http://www.guru3d.com/articles_pages/dishonored_2_pc_graphics_performance_benchmark_review,8.html

 

[cdawall]

I agree that it should help lower end hardware perform better when implemented properly. It is extremely rare that it happens and is sad. Usually games I play don't run worse on PC, if youre looking for way better than console graphics or constant 100+fps is where you start to see more problems. If every PC game was benchmarked at console level graphics and resolution you'd see REALLY impressive numbers all across the board. Hardware does and should make at least some difference though. Speed of a CPU, memory or GPU should all at least slightly impact a games performance. Obviously GPU being the most difference.

If you benched at 720P/med&high settings you would see AMD processors perform awful. They don't have the IPC to push as high FPS.

http://www.guru3d.com/articles_pages/battlefield_1_pc_graphics_benchmark_review,9.html

http://www.guru3d.com/articles_page...e_warfare_pc_graphics_benchmark_review,8.html

http://www.guru3d.com/articles_pages/dishonored_2_pc_graphics_performance_benchmark_review,8.html

This is the part I am getting at. With a good API you don't need anything, the GPU does the work like it should. This is what all proper coding should look like.

 

[Assimilator]

Got another way to explain constant bad console ports?

A console isn't a PC, no matter how much AMD hardware is in it.

 

[cdawall]

A console isn't a PC, no matter how much AMD hardware is in it.

The Xbox literally run Windows 10 now

 

[BiggieShady]

Zen rumors are going wild these days at semiaccurate. https://m.reddit.com/r/Amd/comments/5cffyt/newish_details_about_zen/

• There are some errata issues present in the current testing samples, similar in a way to the TLB bug of the Phenom. The workaround right now is done via the BIOS. The workaround however, strips around 30 ~ 40% of the CPU performance.
  
  
• The CPUs are well behind schedule and every day there's real progress and bug fixing being done. Unlike with INTEL's E0 CPUs which make it to the wild that are almost completely final silicon. AMD's samples will continue to get bug fixes right up until retail spec sampling to partners.
  
  
• In August Clock speeds were 3.8GHz, right now 4.2GHz overclocking is possible, with LN2 5GHz is doable. Again this will change of course, but it is just the current silicon that is behaving like this.
  
  
• AM4/ZEN uses an SOC design, that means even CMOS/BIOS configuration is on package (not necessarily on silicon, I can't confirm this) so it is possible to clear the "BIOS" and still have old value applied 30 minutes later. How this will be addressed remains to be seen. Perhaps it won't be the same scenario for final silicon
  
  
• Operating voltages (nominal) are 1.3v and all the way up to 1.5v should be fine it seems for AIO cooling. Frequency scaling isn't a strong point but again that may have everything to do with the process at this point rather than an inherent design limitation.
  
  
• Performance is particularly strong at this point vs. INTEL's latest offerings. Single thread performance is matching Haswell-E and of course multi-threading performance as well. Tests that are memory bandwidth dependent may go to the INTEL platform simply as a result of having more memory channels, but I can't confirm that right now and have no info on that. The important thing here is that the 16Thread/8-Core CPU is minimum 5960X performance if not better actually. (Based on Cinebench R15) with the error fix disabled.
  
  
• Can't speak to how well the IMC is working as current samples are locked to low DRAM frequencies (2133MHz and lower) and of course this has an impact on performance.
  
  
• As stated in the beginning, every week is progress and AMD is working at an unprecedented rate to get these ready by March.
  
  
• You're unlikely to see any high end boards for the CPUs prior to launch or at launch, simply because no vendors can commit to too much right now as plenty is changing at a rapid rate.
  
  
• All overclocking is done via Overdrive, you can't change any performance features at all in the BIOS (on to that next) at all.
  
  
• BIOS or UEFI is actually built into the CPU, so only AMD can update the "BIOS" or microcode. All overclocking must take place within the Operating system
  
  
• Right now it takes up to 30 minutes to clear the BIOS. If you remove the CPU and place it on another motherboard, it'll have the same settings applied as on the previous board. So debugging is a nightmare
  
  
• 6850K SKU (May not be final designation) is wait for it.... $300 roughly. That's 8 Cores and 16 Threads
  
  
• AMD's Hyper Threading is called SMU and it is ************ good. The same efficiency as Intel's HT.
  
  
• Performance is really good, be it SuperPi, Cinebench, 3DMark etc, it's FPU performance is incredibly good and easily matching that of what Intel offers.
  
  
• Current performance is staggering even though it is limited to 2133MHz (as mentioned before) and NorthBridge Frequency is limited to 2400MHz
  
  
• There will be a nigher SKU than the 6850K, but it is a higher bin so it will certainly overclock better than 6850K and that may carry a premium price, but unlikely to be double.
  
  
• There's plenty of excitement from all board vendors about the platform, so we will see how it all pans out. (Especially with the hot mess that INTEL has in store for us H2 2017, that we can leave to another thread)
  
  
• For Gaming, the CPU is neck and neck with INTEL, even at low res where CPU bound.

Did I mention those are only rumors?

 

[cdawall]

Zen rumors are going wild these days at semiaccurate. https://m.reddit.com/r/Amd/comments/5cffyt/newish_details_about_zen/


Did I mention those are only rumors?

If the errata bug is true nothing that was said after matters. 30-40% performance difference makes them only as fast as steamroller.