• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** AMD "Zen" thread (inc AM4/APU discussion) ***

Because relying on developers to optimize for architectural 'quirks' has always turned out well...
A few titles did get patches for the unusually bad performance Ryzen had in them, but what about the thousands of older titles, will developers ever bother to update those? Relying on developers to optimize is always a losing strategy, and it's going to happen with Intel's Skylake-X too (mesh & cache hierarchy), it's never going to be properly utilized by consumer applications.
But relying on developers to optimise has ALWAYS been the case, it just so happens that for the last 10 or so years it hasn't been as necessary as it used to be and developers got lazy (in a manner of speaking). You wouldn't notice that a certain application isn't well optimised for Intel CPUs because there was nothing to compare it with. Any application that Ryzen "did well in" at launch was probably better optimised for Ryzen than it is for Intel, just by accident. Going forward, optimising for Intel and AMD platforms will be more intentional.

And I think you missed the point of the dual socket latency comparison, it's noteworthy that a fabric between two dies on the same substrate has a higher latency impact. Here's an article on it: https://www.servethehome.com/amd-epyc-infinity-fabric-latency-ddr4-2400-v-2666-a-snapshot/ latency is similar to a 4P Xeon configuration, which isn't really ideal and AMD will have to tackle the issue for Zen+ or Zen 2.
Yeah it's a problem but you get far more cores per dollar with AMD right now in the server space, so it's a good trade off for a lot of customers surely. Get an AMD 1-socket system with similar latency issues to circumvent as an Intel 2-socket system, but get the same number of cores for less money.

Like you've said, decoupling it from DRAM frequency would be a good start, in my opinion they should remove it from desktop parts.
If possible, sure, but if it was at a noticeable cost increase to the product line-up I'm not sure it'd be worth it to be honest.
 
People buy into hype. The problems the platform have are ignored, until they experience said problems and then change their opinion.

Ryzen has never and will never be perfect, but as someone who was very critical of Piledriver, I'm very happy with my Ryzen set up.

This man gets it. Seems to be an attack on anyone that mentions a bad thing about ryzen in here lately.
 
This man gets it. Seems to be an attack on anyone that mentions a bad thing about ryzen in here lately.

You seem.... quite insistent to pan Ryzen for running games optimised purely for Intel, not quite as well as Intel.
The "inconsistencies" you insist on focusing on are clearly ONLY to do with Ryzen being crap. Nothing to do with the different code being run by the various programs (then put through an Intel intended compiler), that's not even a remotely variable possibility, it must just be those damned AMD chips that are sloppy made and all over the place. :rolleyes::rolleyes::rolleyes:

When you realise your narrative is so one sided, a conversation can be had.

If, despite all Ryzens apparent terribleness it's often within spitting distance of a chip that wasn't remotely close to market when it was developed, how poor is it? Really now..? You're going to use a history of software developed purely with Intel and single thread performance in mind to put it down. "It's crap cos it doesn't run Mass Effect 3 very well" (or similarly pointless critique).
"Jesus, these boats don't fly very well, do they?!"

In general it's about 10-15% off Kaby performance levels (correctly setup. Yes, requires correct accompanying parts, unlike Intel who always launch architectures perfectly... wait) . Unless your budget is unlimited, the ~50% cash saved on the same cores and threads can be spent elsewhere to equal a better machine overall. How is this such a massive fail? It's done completely well enough. If you can accept that, I'd think we can agree we're all talking with unbiased opinion.

IPC is... within 5%, depends what it's doing. Clocks are the issue due to the lithography process it was built on. 14nm finfet is power efficient, it's not intended to be a juggernaut. A move to 12nm and a change in process (not seen what, exactly is being used, I assume 12nm finfet and that there's some sort of "special sauce" when finfet is die-shrunk) IF (as speculated though) it achieves even 10% performance improvement, with development now taking note of it, the gap (to Kaby at least) is effectively closed. It will win in some, it will loose in some. It will probably still be half the price. Once we're at "equal stature, half the price" - what exactly is the point in sticking with Intel?

I suggest you start getting a "lol it runs hot" type defence ready. If you're going to insist on carrying a flag for Intel going forward, you'll likely be finding increasingly less solid ground to stand on.
 
Last edited:
I doubt we'll get to a situation where it's "equal stature, half the price". Things are usually priced lower or higher for a reason. If AMD's CPU's truly are as good as Intels soon, I expect pricing to be very similar, definitely not half the price. Either AMD will raise prices or Intel's will come down IMO.
Currently it's clear Intel are still ahead. There's not a lot of reviewers I'd take too much notice of, but the common theme still seems to be "if money no object, buy Intel".
AMD has done well though, definitely shaking things up a bit! Even Intel fans have to agree on that at least :D.
Looking at the TR1900, it already seems to be around the same price as the 7820X, both of the 8 core enthusiast range.
 
Last edited:
I doubt we'll get to a situation where it's "equal stature, half the price". Things are usually priced lower or higher for a reason. If AMD's CPU's truly are as good as Intels soon, I expect pricing to be very similar, definitely not half the price. Either AMD will raise prices or Intel's will come down IMO.
Currently it's clear Intel are still ahead. There's not a lot of reviewers I'd take too much notice of, but the common theme still seems to be "if money no object, buy Intel".
AMD has done well though, definitely shaking things up a bit! Even Intel fans have to agree on that at least :D.
Looking at the TR1900, it already seems to be around the same price as the 7820X, both of the 8 core enthusiast range.

Pre Ryzen Intel was around £100 per core. AMD are offering very similar performance for half that and they don't really need to. They seem to be turning a nice profit and have a healthy turn over because they're charging the right price.
 
Pre Ryzen Intel was around £100 per core. AMD are offering very similar performance for half that and they don't really need to. They seem to be turning a nice profit and have a healthy turn over because they're charging the right price.
I think it's a case of "if they could, they would". Think last quarter results didn't blow investors away so we'll have to see what the next(current) quarter brings. It could have course be that they're keeping prices low to get market share which is no bad thing. When a company is owner by investors, the bottom line matters greatly.Since last quarter it's actually Intel shares that are holding up better, much better. Maybe the market is wrong, we'll see soon I suppose.

Note that I actually think AMD will do well and hopefully next(current) quarter gives investors a nice surprise.
 
Last edited:
Like... wow.

image.png


image.png
 
^^^ I don't think Amd's market share has nosed dived since Ryzen has been on the market. I would have though it would be the opposite so I'm not sure what to think of that to be honest.

That graph reads like Amd has lost way more market share during the launch of Ryzen than it did the whole time they had Bulldozer.
 
Its a benchmarking app ^^^^ it looks like it was DDosed with Intel uploads by some Intel Shill....

Don't see this as an argument but who are we meant to believe here?

zjco4i.png


We need someone like OCuK to give us numbers!

Thats who downloaded a benchmarking app and used it and uploaded the results, which is obviously not sales, the one i posted is actual sales statistics.

That BTW does not look normal, that's an obvious scripted mass fake Intel submission.
Like... wow.

image.png


image.png
 
Its a benchmarking app ^^^^ it looks like it was DDosed with Intel uploads by some Intel Shill....
Thats who downloaded a benchmarking app and used it, which is obviously not sales, the one i posted is actual sales statistics.

People downloading the app to get a feel for how much of an upgrade Ryzen is?
 
Its a benchmarking app ^^^^ it looks like it was DDosed with Intel uploads by some Intel Shill....



Thats who downloaded a benchmarking app and used it and uploaded the results, which is obviously not sales, the one i posted is actual sales statistics.

That BTW does not look normal, that's an obvious mass fake Intel submission.

Can I ask the origin of that graph? So far it looks like it come from reddit.
 
^^^^^ https://www.passmark.com/products/pt.htm

https://www.cpubenchmark.net/

People downloading the app to get a feel for how much of an upgrade Ryzen is?

No one is downloading it like that latest chart suggests, look at the spike, its a mass submission script, very similar to a DDos attack where instead of hitting the site with mass requests to deny service its hitting the site with mass Intel Submitions.

Some Intel shill didn't like where the chart was heading and hit it with a mass Intel submission script.
 
^^^^^ https://www.passmark.com/products/pt.htm

https://www.cpubenchmark.net/



No one is downloading it like that latest chart suggests, look at the spike, its a mass submission script, very similar to a DDos attack where instead of hitting the site with mass requests to deny service its hitting the site with mass Intel Submitions.

Some Intel shill didn't like where the chart was heading and hit it with a mass Intel submission script.

I'm aware of what passmark is. Do you have any evidence to suggest these results come from a script?
Do the mindfactory.de images origin from reddit?

*EDIT* Your image come from imgur/reddit. Do mindfactory themselves actually share this info?
 
^^^^^ https://www.passmark.com/products/pt.htm

https://www.cpubenchmark.net/



No one is downloading it like that latest chart suggests, look at the spike, its a mass submission script, very similar to a DDos attack where instead of hitting the site with mass requests to deny service its hitting the site with mass Intel Submitions.

Some Intel shill didn't like where the chart was heading and hit it with a mass Intel submission script.

Yeah seems odd to see such a big spike. Very strange.
 
Back
Top Bottom