• 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 announce EPYC

And add to that Intel's supposed supply shortages hitting the big OEMs. For the mass-produced boxes OEMs churn out, Ryzen 2000 is sufficiently performant, is readily available to buy (and doesn't cost the earth accordingly) and is easier to cool. If the OEMs flock to AMD as a result then AMD's mindshare increases because the Ryzen logo will be plastered over everything.

That's not just maintaining momentum, that's kicking things into overdrive.
 
think this could be amd 's time ,last time it was they got shafted when they were better ...this time round we all know better i hope zen 2 bends them over the desk
 
think this could be amd 's time ,last time it was they got shafted when they were better ...this time round we all know better i hope zen 2 bends them over the desk

Think the world is more wise now, plus Intel got their pants pulled down by the EU and others over their previous misdemeanours, so i think although they will still try to halt the flow of AMD, they wont be so blatantly illegal about it.

Offering huge discounts on their overpriced Enterprise guff is the start, which is insulting in itself, as people realise how much Intel has been bending them over pricewise for a long time, as there was no competitor, and as a result some are spiting Intel by swapping to AMD as a big 2 fingered salute over their "Generous offer" of price drops. Lets be honest here, if AMD was not able to compete, Intel would still be ruling from on high with massively overpriced products, its actually not a great situation for them, as the only place they can go currently is to drop prices, and as i say this is upsetting some customers who were paying top dollar for a product that Intel has subsequently slashed in price by upto 50% in some instances, and even then its still not that great a deal when compared to AMD.

I think EPYC Zen2 is going to move a lot of enterprise market back towards AMD's way, and i think that over the next few years AMD will increase their hold on the Enterprise market, especially if Zen2 is a decent uplift over Zen1 and Zen1+... bodes well then for Zen2+, Zen3, Zen4 etc, if AMD can consistently improve generation on generation as they are promising. Intel have a hard task to fight that, performance alone wont do it, they'll need to start coming back with better pricing from the get go as well.

Everything trickles down from the Enterprise market, the cost of desktop chips will drop if there is a decent performance / price war between AMD and Intel... right now Intel still offer the best single threaded performance, AMD offer 90% of that for 50% of the price, so they offer the best performance per pound spent. Some people will just pay whatever the cost for the absolute best performance, if AMD achieve that with Zen2 and still well under the price of Intel, then Intel need to beat AMD on both fronts as reputation alone wont sell as many units as it used to, not now people have woken up to Intels business model.
 
AMD’s Rome is indeed a monster
https://www.semiaccurate.com/2018/11/09/amds-rome-is-indeed-a-monster/

"Intel has nothing to answer this with and won’t until 2022. By then AMD will have two more generations out"

You have to give it to them Rome is basically looking to be an engineering masterclass, mixing nodes, sticking it all on an organic package and delivering class leading performance out of one low power chip which also happens to be nice and cheap to produce. You can imagine how smug the engineers over at AMD must be feeling right now. I said it a page back but this does feel like proper next gen stuff.
 
You have to give it to them Rome is basically looking to be an engineering masterclass, mixing nodes, sticking it all on an organic package and delivering class leading performance out of one low power chip which also happens to be nice and cheap to produce. You can imagine how smug the engineers over at AMD must be feeling right now. I said it a page back but this does feel like proper next gen stuff.

Agree. Thing is though, this sort of innovative engineering solution comes from peeps that want to and love to solve engineering problems. Intel, with all the cash they have and the huge R+D dept have no response. It proves that they really don't give a monkeys about driving computing or innovation forward and all they care about is milking us, the consumer of there products.
 
Agree. Thing is though, this sort of innovative engineering solution comes from peeps that want to and love to solve engineering problems. Intel, with all the cash they have and the huge R+D dept have no response. It proves that they really don't give a monkeys about driving computing or innovation forward and all they care about is milking us, the consumer of there products.

I think we all knew that was the case though, any company in such a position always milk the consumer, if there is no competition then you are sometimes forced into a particular product or service. Another great example of this would be my train into the city each day, 5k a year for travel is ludicrous but what can I do? It goes up every year as well, while the service is getting worse but there really aren't any other realistic options. Intel were in that same dominant position just in a different market. The difference between my train line and the CPU market is that nobody has engineered me a better means of travel than the train given cost of congestion charges etc if I drive.

Back to Rome and the suggestion that in the consumer cpu space we are not that far away from tech that just a few years ago was reserved for big data centres and supercomputers. I genuinely can't wait to see what TR looks like, will we get like a mini Rome? I mean up until the 2990x it was generally accepted that memory channels were tied to specific clusters because of architecture limitations. Breaking that tie between memory channels scaling on core count is the real master stroke here I think, all the clues of what Rome could be were in there but how they have delivered it looks to be properly impressive. I know what ill buy when it comes to HEDT, now all I need to find is somebody to make a teleportation device or those tubes things you see in futurama and my travel might get cheaper as well :)
 
Agree. Thing is though, this sort of innovative engineering solution comes from peeps that want to and love to solve engineering problems. Intel, with all the cash they have and the huge R+D dept have no response. It proves that they really don't give a monkeys about driving computing or innovation forward and all they care about is milking us, the consumer of there products.

My impression of Intel is world leading Accountants, Male Bovine Manure-ra-ras, not an enthusiastic, thinking outside of the box, technology innovators, that's AMD.
 
You have to give it to them Rome is basically looking to be an engineering masterclass, mixing nodes, sticking it all on an organic package and delivering class leading performance out of one low power chip which also happens to be nice and cheap to produce. You can imagine how smug the engineers over at AMD must be feeling right now. I said it a page back but this does feel like proper next gen stuff.

One of the absurd things is how industry leading AMD actually is. With a tiny fraction of Intel and even Nvidia's budget for the past couple decades AMD have been on the forefront of most memory technologies and advances, they were the first to bring an HBM/interposer chip, they are the first to make chiplets, first to make a proper IMC, first to 'proper' dual cores. Hell, they had one of the best mobile/power efficient gpu architectures on the market but had to sell that part of the business due to Intel's illegal market manipulation being hugely responsible for AMDs debt.

Thing's like HBM, interposers, chiplets are huge huge steps in the silicon industry and yet it's AMD making those steps, not Intel with an embarrassingly larger R&D budget. I think Intel spent more on buying startups than AMD did on their own R&D by a massive margin and then there is Intel's own R&D and monumentally larger engineering staff. It really shouldn't be possible for AMD to make these huge leaps before Intel.

Adjusting to chiplets and interposers/on package memory is something AMD has been planning for a decade and yet Intel seeing this, seeing industry sentiment basically being everyone else agreeing on chiplets and splitting chips into blocks to be made on the best node for that function being the way forward somehow still Intel manages to be miles behind on it.

But that is kind of the thing, sometimes lack of budget causes you to be ultra innovative while huge money makes people arrogant and very conservative in how they plan for the future.
 
Back
Top Bottom