Jump to content
Sign in to follow this  
Dasa

AMD Zen

Recommended Posts

Those are high TDPs on the high end chips and the real consumption will probably double it, based on the i9 9900K using over 200W at peak performance (TDP 95W)

Share this post


Link to post
Share on other sites

A reply from a user on whirlpool forums I found interesting.

 

Dasa writes...

Quote

I am interested in the 12c for gaming but not so much for the extra cores I am wondering if the extra 32MB L3 cache will be of benefit

Jag writes...

They haven't disclosed the cache layout as yet, but I don't think it's a unified L3 cache. It'll probably be 32MB L3 used by 6 cores in the same chiplet, and another 32MB used by the other 6 cores in the other chiplet. I doubt it'll be any faster than the 32MB in the single chiplet designs except in limited circumstances (running very different tasks on each chiplet).

Quote

or if the dual chiplet design will hurt more than help due to increased latency between cores.

There shouldn't be increased latency any more because there's no longer any local access to memory. The memory controller has now moved from the CCX to the I/O controller chip, each chiplet connects to the I/O controller. This means that all cores have equal latency to memory. The previous design was NUMA, with each CCX having direct access to half the RAM, and accessing the other half via the Infinity Fabric and other CCX.

Now all chiplets access RAM over Infinity Fabric to the I/O Controller. ie: Instead of having half memory fast and half slow, now all uses the slower mode. Reducing the impact of this design choice is the primary reason the L3 cache has doubled.

BoganTimmy writes...

Quote

So far AMD have been pretty decent on their performance presentations.

Jag writes...

I thought the PCIe 4.0 bandwidth test to be presented in such a way as to be easily misleading. If you miss any of the qualifiers it appears to be showing something that it wasn't.

At this point in time there was the comparison in gaming that suggests it's at least on par with Intel.

WIth no testing parameters it's impossible to know at this stage. Perhaps both CPUs were hitting a GPU bottleneck? They were suspiciously close. Independent testing will show this up pretty quick.

the_other_guy writes...

Quote

70m vs 16m at this point intel is in trouble

Jag writes...

If you were going to count the L2 and L3 like this than it'd be 70MB vs 18M.

However this isn't really an apple to oranges comparison for several reasons:
1. As per above, it's likely that the 64MB L3 actually operates as two independent 32MB caches, rather than one combined cache.
2. Zen's L3 cache operates as a non-inclusive victim cache and not a general purpose write-back cache like on Intel. That means this cache contains data that was evicted from the L2 cache. When you need to access data in the L3, it must swap data between the L2 and L3, which slows a read operation into a move (L3 <-> L2) and read (from L2). This increases cache latency. Hopefully this is something that Zen 2 improves on, but there's no architecture information indicating that they've changed this as yet.
3. Due to the above, AMD's L2 acts more like a window into the L3, so adding them together isn't really appropriate. Technically it can hold <L2 size> + <L3 size>, but it acts more like a cache of <L3 size>.
4. Zen generally has a higher latency path to memory than Intel does, so cache size is more important.

Due to the above it'd be closer to say in operation it's more like 32MB vs 18MB, but even then that is a fairly irrelevant comparison.

sysKin writes...

Quote

hindrance (because now you have to look up another chiplet's cache before knowing whether DRAM access is needed)

Jag writes...

It would add a fair bit of complexity to the I/O Controller. If they weren't victim caches it may be worth always checking the other L3 for the information before checking RAM, but I don't think this will happen for this generation. They haven't released much info about the I/O Controller, and the I/O controller does some unexpectedly strange things like directly support an audio codec and USB ports (Meaning these don't compete for chipset bandwidth), so there's still some hope yet.

Sov writes...

Quote

Why in the hell would anyone want to use NVMe in RAID for a desktop system?

Jag writes...

RAID 1 maybe? RAID 0 seems a bit silly except for fairly useless bragging rights. If you really need RAID 0 NVMe, then you probably want a HEDT system with non-M2 PCIe SSD. The way SSD access their NAND memory is effectively already RAID 0 built-in to the drive. There may be some edge cases where you can get multiple PCIe 3.0 x4 SSDs for cheaper than the newer PCIe 4.0 SSDs of similar speed, but I imagine there will be a fairly narrow window.

Nukkels writes...

Quote

I'm quite surprised that Intel can get much higher boost clocks than AMD, on a larger process node too.

Jag writes...

Mature node vs new node. This behaviour is fairly typical, it's just at an unusual extreme because Intel's 14nm process has had extended tuning in comparison to previous nodes. Intel's original node roadmap acknowledged this by expecting to run 10nm alongside 14nm+*, they'd use the 10nm node for power efficiency applications and the 2nd iteration 14nm+ for performance applications.

I think the biggest surprise in the new CPUs is that the 3700X has a lower base clock than the 2700X. Too early to tell if they've had binning issues, or if they just needed to make a bit more room for the 3800X so that they could up-sell the typical price point.

Note that we're also yet to see how AMD will be calculating TDP for this generation, which is fairly important info. Their previous method used a load temperature that was tied to the GoFlo nodes, this may change with the move to TSMC. Boosting may also work differently.

  • Like 2

Share this post


Link to post
Share on other sites
15 hours ago, gamble said:

Seems that since amd can match or better Intel in certain price points they are holding back higher clocks and more cores.

They are definitely holding back. Now that they can match Intel and have them by the short and curly's, they can just refine and release. AMD's product roadmap seems to be very strong. There's talk of 2.5D for Zen 3 in the form of HBM on the I/O die. and 3D stacking for Zen 4

 

Intel however are stuck on the confinements of 14nm desktop and server for another two and a half years, that means whatever refinements their new cores will have, they won't get more logic in there so IPC is going to be super hard to eek out.

 

I just hope that AMD don't stuff it up now they have the upper hand.

  • Like 2

Share this post


Link to post
Share on other sites
Posted (edited)
8 hours ago, SceptreCore said:

I just hope that AMD don't stuff it up now they have the upper hand.

 

Yea, like get to greedy... If they can gain some market share first while being profitable.   Too me it looks like they probably were planning to release at better prices but saw the market and what people were willing to play considering the performance they gained in IPC too. 

 

@Dasa  That is some good info.  12c cache will be interesting if it is only using 32m at a time depending on the tasks.  If there is benefits from running two separate caches in games, I bet it would have to be programmed for such.  Latency could be worse if it tries to use both possibly or if there is a way for it to know what to use when.  If the base 12c 'all core' is running lower than the 8 core part, the 8 core might just be better than the 12c for gaming if running the same speeds.   I know that when I run my current cpu with all cores locked at 4.15ghz, I get less fps and benches are lower than just running it stock... AoS might be a great game and a few others to compare the two top cpus coming out (8c vs 12c) as it seems to actually use the cores it has available.  It would be interesting if the 8 core is just as fast or if the 12c loses some ground once more than 8c are used but that extra 100mhz might make up for that on the 12c. 

 

For me, I was wanting a 12c 5ghz for under 500 from the rumors so now Im looking more at if they will have a better 8 core coming with a bit higher clocks.  I guess it would be fun to play with but the last two gens there was very little in the way of OC headroom.    

 

Edited by gamble

Share this post


Link to post
Share on other sites

It's definitely better then Ryzen 1 and 2 which had I guess huge issue getting to 4.5GHz at all, so they can push the speeds to 5 but it seems the stable speed are around high 4s. The rumoured prices are USD so I would double that then add Australia tax.

Share this post


Link to post
Share on other sites

Anyone here get AMD stock expecting it to suddenly increase?

Share this post


Link to post
Share on other sites
3 hours ago, Jeruselem said:

Anyone here get AMD stock expecting it to suddenly increase?

I wish I did... but I don't know how. It's jumped ~10%. That's big.

 

12 hours ago, gamble said:

Yea, like get to greedy... If they can gain some market share first while being profitable.   Too me it looks like they probably were planning to release at better prices but saw the market and what people were willing to play considering the performance they gained in IPC too. 

Well it's all for nothing if this better technology doesn't get them any money. But on par single thread performance with a 9700k and 30% greater multithreaded performance for $50 cheaper. No reviewer out there will recommend Intel. It's not a bad move.

 

AMD is still undercutting Intel. But it could be a little sweeter. 😄 

Share this post


Link to post
Share on other sites
34 minutes ago, SceptreCore said:

I wish I did... but I don't know how. It's jumped ~10%. That's big.

 

 

But if AMD stuff things up (again), then the stock falls 20%!

Share this post


Link to post
Share on other sites
47 minutes ago, Jeruselem said:

 

But if AMD stuff things up (again), then the stock falls 20%!

You need a slap

Share this post


Link to post
Share on other sites
Posted (edited)
4 hours ago, Jeruselem said:

Anyone here get AMD stock expecting it to suddenly increase?

Wanted to buy some before Ryzen when there stock was about the lowest ever as I new it would be easy money but didn't bother working out how to buy stock at the time.

Buying now would seem a little risky as stock wont increase by anything near that much again and like Jeruselem I have run over black cats while I was towing a trailer that are luckier than me.

Edited by Dasa

Share this post


Link to post
Share on other sites

Provided it is at least as fast as my 6700K in games that use 4 cores or less yes.

Share this post


Link to post
Share on other sites
Posted (edited)

Ryzen 3 and chipsets! If you have the A320 chipset, get a new motherboard.

 

 

Edited by Jeruselem
  • Like 1

Share this post


Link to post
Share on other sites
50 minutes ago, SceptreCore said:

I'd be inclined to get a new one anyway. A320 are not great boards.

 

I guess it was the super-budget one ... to be expected

Share this post


Link to post
Share on other sites

Something kinda interesting I posted in Patreon:

 

Since say the new AMD multi chiplet design will have two physically different core stack locations connecting to the IO module, will that produce more latency when gaming? If the L3 cache is not shared between chiplets (or is it?), could this actually make it slower than a single 8 core chiplet for things that are sensitive to latency? Even if they had the same amount of actual cores, a 4+4 (unlikely design, just for conversation) vs an single 8, the single chiplet you would think would be faster... I guess Im curious on how it will actually use resources between each other with the IO die. Oh... Love the new site!
 
 
Yeah it'll be interesting to see how the I/O die works...today I saw some pretty interesting latency numbers though, check this out. https://twitter.com/hms1193/status/1134168659004338176
26min
 
 
Oh wow nice find, interesting... comparing my 2600x and 3200 Dimms.. Im at 67ns... 1, 3 and 9.4 respectively.

Share this post


Link to post
Share on other sites

If true that is extremely good latency to the RAM and would mean a massive improvement in gaming performance.

L1 and L2 have good bandwidth and acceptable latency.

L3 is a little slow but the increased size could more than compensate for that.

Share this post


Link to post
Share on other sites

Apparently that Impact MB wont be ready at launch and may come about a month later which sucks for me.

Share this post


Link to post
Share on other sites

I kinda expected that to happen, better launching a working product than one which isn't right.

Share this post


Link to post
Share on other sites
Posted (edited)
On 5/30/2019 at 11:40 PM, SceptreCore said:

A ryzen 5 3600 looks pretty tasty to me.

 

I mean my 6 core does a great job already... Im not sure if I'll go 8 or 12 core.  The cache is what has me questioning and if the 8 core can OC well then might go that route.  The last two ryzens were crap OCers.. I'd be curious if the 3700x with its lower tdp OC better or as good as the 3800x.  3700x at 2/3rds of the price of 3900x hmmm... 

 

 

Edited by gamble
  • Like 1

Share this post


Link to post
Share on other sites
Posted (edited)
9 hours ago, gamble said:

 

I mean my 6 core does a great job already... Im not sure if I'll go 8 or 12 core.  The cache is what has me questioning and if the 8 core can OC well then might go that route.  The last two ryzens were crap OCers.. I'd be curious if the 3700x with its lower tdp OC better or as good as the 3800x.  3700x at 2/3rds of the price of 3900x hmmm... 

 

 

Look how much better the 3600 is to the 2600X

 

❤️ AMD

Edited by SceptreCore

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×