AMD '#Rekt son' reply to Intel's "four glued together" statement

[JohnnyCorporalTech]

AMD has been relatively calm about the whole "4 dies glued together" fiasco. After Intel took a swing at AMD regarding their EPYC CPU lineup, AMD has proven to be the more mature of the two silicon giants, releasing a cool and collected response. AMD's Scott Aylor responds by stating:

Quote

There’s a theory out there that EPYC is just 4 desktop processors glued together. When you look at throughout the presentations today, from Mike Clark, about the memory sizes and features that he has built into the core and when you hear from Kevin the security enhancements that are in there, when you hear from Jerry about the over provisioning of memory bandwidth between the devices, you hear from Sam determinism features that aren’t available from the competitor that were built for enterprise. Hopefully you will realize this is not a glued together desktop processor.

AMD's cool and collected response speaks volumes about their confidence. AMD also provided some slides explaining how their EPYC products are

Quote

Harnessing the power of the X86 Ecosystem.

It is clear from Intel's response that EPYC is disruptive, and that it has already had an effect on its competitor on the blue team. 

 

here are some of the slides AMD has provided:

 

Source:http://wccftech.com/amds-reply-to-intel-calling-naples-epyc-4-glued-togather-desktop-dies/

 

[RadiatingLight]

TBH that ecosystem slide looks very unimpressive.

[sazrocks]

Just now, RadiatingLight said:

TBH that ecosystem slide looks very unimpressive.

At least they don’t have companies on there twice...

[RadiatingLight]

Just now, sazrocks said:

At least they don’t have companies on there twice...

I'm willing to believe that Hitachi was just an honest mistake, if it was really the only duplicated brand.

[f22luke]

2 minutes ago, RadiatingLight said:

TBH that ecosystem slide looks very unimpressive.

Why because they didn't cram in as many meaningless names like Intel?  The Intel slide had duplicates and was full of names that had nothing to do with "ecosystem" unless you define it in the broadest possible sense.  

[2FA]

2 minutes ago, RadiatingLight said:

I'm willing to believe that Hitachi was just an honest mistake, if it was really the only duplicated brand.

There were three brands duplicated...

[ARikozuM]

2 minutes ago, RadiatingLight said:

I'm willing to believe that Hitachi was just an honest mistake, if it was really the only duplicated brand.

Unless one of the "Hitachi" were supposed to be HGST as WD and Hitachi are related as is HGST.

[RadiatingLight]

Just now, DeadEyePsycho said:

There were three brands duplicated...

Which ones?

this is interesting

[jeffery7466]

1 minute ago, f22luke said:

Why because they didn't cram in as many meaningless names like Intel?  The Intel slide had duplicates and was full of names that had nothing to do with "ecosystem" unless you define it in the broadest possible sense.  

They were totally dishonest by claiming that AMD had no ecosystem. So to me it doesn't matter whether they were cramming or not.

[Marinatall_Ironside]

#DamnSon

 

Looks like Intel shit their breeches on this once    

??????

[The Sloth]

4 minutes ago, jeffery7466 said:

They were totally dishonest by claiming that AMD had no ecosystem. So to me it doesn't matter whether they were cramming or not.

or having misleading once like Gigabyte and Microsoft. 

 

anyone wondering 

[NumLock21]

7 minutes ago, RadiatingLight said:

Which ones?

this is interesting

Costco Wholesale

[kiska3]

Triggered yet? 

/s

[sazrocks]

8 minutes ago, RadiatingLight said:

Which ones?

this is interesting

Here is a picture with the duplicates circled:

[2FA]

3 minutes ago, RadiatingLight said:

Which ones?

this is interesting

3 minutes ago, NumLock21 said:

Costco Wholesale

Hitachi

DDN

 

[Hunter259]

How is saying it's "Glued together" some childish statement. They are, in a way, glued together. Intel did it in the past when they couldn't figure out how to make a 4 core die and have it not be a money sink. They left that as soon as they could. It's an inherently technologically inferior way of making a CPU.

[Bcat00]

It really boggles mind how some dumb slide can blow up into something news worthy. If i ask my cousin who works as a script writer for cbc, he would tell me to get lost and quit wasting his time.

[Marinatall_Ironside]

3 minutes ago, Hunter259 said:

How is saying it's "Glued together" some childish statement. They are, in a way, glued together. Intel did it in the past when they couldn't figure out how to make a 4 core die and have it not be a money sink. They left that as soon as they could. It's an inherently technologically inferior way of making a CPU.

Or was in this case with AMD and their Infinity Fabric technology.

 

Now, what is more of a money sink is trying to squeeze as many cores and as much cache (plus FPUs and other shenanigans) as you can onto a single chunk of silicon.

Edited July 19, 2017 by JurunceNK

[Hunter259]

Just now, JurunceNK said:

Or was in this case with AMD and their Infinity Fabric technology.

 

Now, what is more of a money sink is trying to squeeze as many cores and as much cache (plus FPUs and other shenanigans) as you can onto a single die.

It's better to have it be a single die with all cores together. Less latency that way. Obviously you can make a fast chip with dual dies/CCX's but it is a worse way.

[Marinatall_Ironside]

Just now, Hunter259 said:

It's better to have it be a single die with all cores together. Less latency that way. Obviously you can make a fast chip with dual dies/CCX's but it is a worse way.

We'll see when the actual benchmarks come out. What was true in the past may not be true today. But again, we will see when the third parties gets their hands on these EPYC CPUs and benches them.

[Hunter259]

Just now, JurunceNK said:

We'll see when the actual benchmarks come out. What was true in the past may not be true today. But again, we will see when the third parties gets their hands on these EPYC CPUs and benches them.

Uh. This is something that is not changing and is basic Computer Science/Engineering. Multiple dies is a less elegant and technologically advanced, although as we've pushed silicon this far it may be the only good way to get past 40 cores, and anytime you leave the die you will get a latency/bandwidth hit. It'll perform like Ryzen does just with more cores due to Ryzen already being a somewhat multi-die chip.

[The Sloth]

8 minutes ago, Bcat00 said:

It really boggles mind how some dumb slide can blow up into something news worthy. If i ask my cousin who works as a script writer for cbc, he would tell me to get lost and quit wasting his time.

CBC? i thought they were busy covering random places shirtless Trudue goes. 

 

[mr moose]

Dear AMD,  cows aren't very good at climbing trees. 

 

 

[The Sloth]

Just now, mr moose said:

Dear AMD,  cows aren't very good at climbing trees. 

 

 

they are good at eating things that are green( its a joke) 

[Jito463]

31 minutes ago, Hunter259 said:

Multiple dies is a less elegant and technologically advanced

I can't say I agree with that.  As with anything in life, there are pros and cons to both methods.

 

31 minutes ago, Hunter259 said:

anytime you leave the die you will get a latency/bandwidth hit.

True, but recall that the Intel method was flawed, due to their reliance on the NB bus for interconnects between the dies.  While IF does scale based on your RAM speed, the actual communication between CCX's still occurs on-die (and the communication between dies on TR/Epyc occurs on the silicon).  Yes, there will be a latency hit, but the extent of the latency determines whether it's a "worse way".  The primary thing holding back Ryzen isn't IF, it's the silicon process from GloFo.  If they could breach that 4GHz barrier, then Ryzen would be a knockout.

 

Being a different method, does not automatically make it an inferior method.