Intel X599 Chipset to Drive 28-core HEDT+ Platform

[dj-electric]

AMD has threadripper
Intel has FuseTripper?

Per given core speed X core amount, the balance is actually tilted to the red side in this case

 

[Raendor]

Thats true, but I dont blame them one bit, I think AMDs doing great in doing so and its a wise move. I cant say im an AMD guy anymore (not since brisbane) but im not a dick swing Intel dude either. AMDs response with zen 1 and now zen 2 must have made multiple mouths drop.

What Zen 2? It’s coming only next year. Stop using Ryzen and Zen interchangeably when these are different things. There’s not a single Zen 2 based cpu on the market, but there’s Ryzen 2 lineup based on Zen+.

 

[biffzinker]

but there’s Ryzen 2 lineup based on Zen+.

Except for the 2200G/2400G is based off the first Zen.

 

[Mussels]

Zen, Zen+ and Zen 2

Annoying that AMD have Zen 2 as the third gen release, but thats life

 

[Caring1]

Zen, Zen+ and Zen 2

Annoying that AMD have Zen 2 as the third gen release, but thats life

How many generations has Intel had on 1151?

 

[StrayKAT]

Correction it's two quad-core clusters (CCX) per die.
8 cores multiplied by 4 four individual dies on one PCB Substrate.


There's always VIA that might surprise.

Doesn't Cyrix still make chips? Or they bought out?

edit: Haha.. nvm. They're long gone.

I'm not sure "duopoly" is the right word for the market either. Isn't AMD still beholden to Intel for some x86 license?

 

[nemesis.ie]

They cross-licence, Intel is effectively beholden to AMD for the 64 bit extensions if you look at it that way.

 

[stimpy88]

Thank you AMD, keep the good work up!

 

[BluesFanUK]

Wow, when buying my 5820K I thought having an option of getting an 8 core CPU in the mainstream market was big news (and a big price!), but these are just bonkers. You can get double the cores for what the 5960X cost just a few years ago. AMD blew the CPU market wide open.

 

[Vayra86]

No. It's a result of die design.

Simple and cheap approach. It's a design aimed at lowering costs, not improving performance. And it starts to show in 16-32 core benchmarks.

Zen approach scales badly with high core count. It's very unlikely that Intel makes something similar. Generally speaking: I'll miss ring bus, but Intel Mesh is still acceptable. I guess we have no choice in the "moar cores" era...
Notice how Intel went for 6 memory channels for 28 cores, while AMD remains with 4 channels. This will make a difference.

Actually no. Its a design that is both cheap to implement AND capable of scaling up performance while still maintaining strong single core numbers. Threadripper has functionality to disable entire die's allowing the remainder to clock higher. Intel has nothing of the sort, instead it has turbo clocks for each core count.

As for the performance scaling across higher core counts, TR wins the day. SMT works a lot better and core/threads on TR scale better than the higher clocks per core on Intel. At the same time, TR can achieve a higher baseclock on all-core loads.

So... no. You're wrong. Jumping to hypotheticals about # of memory channels in the near future won't change that either. Besides, this is not even about the 28/32 core halo products but about the entire product stack and having a cheaper but better performer at every price point. AMD checks that box, Intel does not. Performance is relative. Perf/dollar is what really counts.

 

[chaosmassive]

Per given core speed X core amount, the balance is actually tilted to the red side in this case

in HEDT space, it assume everyone have deep pocket
so efficiency is out of window, and CPU will be given much more freedom in power limit

 

[notb]

Actually no. Its a design that is both cheap to implement AND capable of scaling up performance while still maintaining strong single core numbers. Threadripper has functionality to disable entire die's allowing the remainder to clock higher. Intel has nothing of the sort, instead it has turbo clocks for each core count.

It doesn't scale up performance - you haven't been paying attention. Check the reviews of 2950X and 2990WX.
We've already seen this issue in both 8C Ryzen and 32C EPYC. High latency of core-core communication results in very bad performance in many typical scenarios.

In a consumer segment Intel's 6C compete with AMD's 8C thanks to higher clocks.
In servers Intel doesn't need frequency advantage. 24C Xeon beats a 32C EPYC with ease - just thanks to a better (albeit more expensive) mesh design.

And why would anyone disable a TR die to get higher clocks? Is this why people buy so many cores? To disable them? Bonkers argument. :-D

As for the performance scaling across higher core counts, TR wins the day.

It doesn't! I'm not sure if you understand how IF works.
All IF issues you've read about - like when Ryzen gaming performance grows after disabling one CCX or switching SMT off - are multiplied in high core count systems.
And AMD can't fix this. The only thing they can do is give us more and more cores - that will give them an advantage in some tasks (like encoding) and attract some users.

Performance is relative. Perf/dollar is what really counts.

You're wrong here as well. If a CPU is faster, it is faster. End of story.
If Intel's CPU offers higher performance, people will keep buying it. It doesn't matter that AMD offers 80% for 60% of price. Yes, it wins in "perf/price", but that shrinks drastically when you look at the cost of the whole machine. And in the end you have a single CPU in your consumer PC.

People have exactly the same problem understanding how servers work. Yes, "perf/price" could work in huge datacenters. But smaller workstations/servers are limited to 1, 2 or 4 CPUs. It doesn't matter that you can buy 5 EPYCs instead of 4 XEONs, because you'll end up with a slower 4S server and a very expensive drink coaster.

 

[nemesis.ie]

You're wrong here as well. If a CPU is faster, it is faster. End of story.

I suppose it is then also true that "if a CPU had more security holes, it has more security holes" or "if it loses performance for patching security holes, it loses performance".

Or indeed "it can be faster because it cuts corners leading to security holes"?

I know this is a bit O/T, but I think you get the point.

 

[eidairaman1]

in HEDT space, it assume everyone have deep pocket
so efficiency is out of window, and CPU will be given much more freedom in power limit

Only time that really matters is in laptops

Actually no. Its a design that is both cheap to implement AND capable of scaling up performance while still maintaining strong single core numbers. Threadripper has functionality to disable entire die's allowing the remainder to clock higher. Intel has nothing of the sort, instead it has turbo clocks for each core count.

As for the performance scaling across higher core counts, TR wins the day. SMT works a lot better and core/threads on TR scale better than the higher clocks per core on Intel. At the same time, TR can achieve a higher baseclock on all-core loads.

So... no. You're wrong. Jumping to hypotheticals about # of memory channels in the near future won't change that either. Besides, this is not even about the 28/32 core halo products but about the entire product stack and having a cheaper but better performer at every price point. AMD checks that box, Intel does not. Performance is relative. Perf/dollar is what really counts.

He is a troll, just ignore his arrogance/ignorance, he was put on a 4 month ban for stupidity on his part.

 

[Vayra86]

It doesn't scale up performance - you haven't been paying attention. Check the reviews of 2950X and 2990WX.
We've already seen this issue in both 8C Ryzen and 32C EPYC. High latency of core-core communication results in very bad performance in many typical scenarios.

In a consumer segment Intel's 6C compete with AMD's 8C thanks to higher clocks.
In servers Intel doesn't need frequency advantage. 24C Xeon beats a 32C EPYC with ease - just thanks to a better (albeit more expensive) mesh design.

And why would anyone disable a TR die to get higher clocks? Is this why people buy so many cores? To disable them? Bonkers argument. :-D

It doesn't! I'm not sure if you understand how IF works.
All IF issues you've read about - like when Ryzen gaming performance grows after disabling one CCX or switching SMT off - are multiplied in high core count systems.
And AMD can't fix this. The only thing they can do is give us more and more cores - that will give them an advantage in some tasks (like encoding) and attract some users.

You're wrong here as well. If a CPU is faster, it is faster. End of story.
If Intel's CPU offers higher performance, people will keep buying it. It doesn't matter that AMD offers 80% for 60% of price. Yes, it wins in "perf/price", but that shrinks drastically when you look at the cost of the whole machine. And in the end you have a single CPU in your consumer PC.

People have exactly the same problem understanding how servers work. Yes, "perf/price" could work in huge datacenters. But smaller workstations/servers are limited to 1, 2 or 4 CPUs. It doesn't matter that you can buy 5 EPYCs instead of 4 XEONs, because you'll end up with a slower 4S server and a very expensive drink coaster.

Recent TR2 reviews show improved performance from the new modes that allow you to disable entire dies, which is useful for all the stuff that doesnt scale well into >16 cores. Or >8.

Really what you see in reviews now is similar to many things we saw with new core count procs on both msdt and hedt. Lack of optimization and support. I contest that Intel does it faster, I rhink they are mostly still riding on clockspeed advantage but all the rest isnt up to what Zen is growing into. This doesnt hapen overnight but its clear AMD is pushing in that direction. This is why they take drastic measures to increase market share.

Its really hard to state Intel Core is better for the HEDT environment these days. Its the same thing you see in msdt: Intel leads by such a tiny margin it becomes irrelevant for most except some tiny niche.

 

[Prima.Vera]

So what is this Xx99 numbering competition from both AMD/Intel.?!? Are they starting trolling eachother now over this??

 

[StrayKAT]

So what is this Xx99 numbering competition from both AMD/Intel.?!? Are they starting trolling eachother now over this??

If anything, AMD started it.. But not sure why. It doesn't help, but confuses matters. edit: As if Intel's product lineup already wasn't confusing enough.

 

[AlwaysHope]

There's always VIA that might surprise.

Oh yeah, they will come back & knock the guts out of Intel + AMD. (sarcasm)

 

[Berfs1]

I bet $10 that AMD reserves the X999 for their Zen2/3 based HEDT+
I think you're new to this, every company needs to trash their competition with (model) numbers, that's just how things work in consumer space [/QUO

I bet $10 that AMD reserves the X999 for their Zen2/3 based HEDT+
I think you're new to this, every company needs to trash their competition with (model) numbers, that's just how things work in consumer space

Or maybe because there was a leak of X499 coming... maybe you are new to that leak?