![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
AMD Zen Performance Preview
- Thread starter HardOCP News
- Start date
AMD for a very long time did have a faster clock for clock architecture. Longer then most admit... because for a few years Intel was only the technical winner, as they had some top of the line cherry picked chips running at high enough clocks to claim a paper victory. Intel smartened up and improved their core design and abandoned some of the stupid ideas that made up the P4. (yes that was a long time ago now). AMD likewise made some bad choices with core connects ect with their line.
It would seem AMD learned from that and are about to come back with a core that has solved some of those mistakes... just like Intel did when they came back with the Core line. The main issue for AMD is going to be Fabs. Its no knock on Intel to say they have a solid advantage owning their own fabs and having spent loads of $ on their foundry tech. For AMD to compete with that on the high end... I don't see it being all that realistic. However most chip sales are of the middle of the road variety, in that regard as long as their not turning out wafers with high fail rates at fabs they don't own they should be able to get in the game again. Giving up their own fabs years ago may have been the real moment AMD died, and the following years have just been the final rattle. We'll see... yes it may be wishful thinking to hope things workout for them. If they fail though it won't be the chip design.
AMD knows how to design a high performance core. I believe the question is can they turn out wafers of them that have a high enough % of cores that can be clocked high enough to be in the race. If only a handful of cores on each wafer can hit speeds high enough, or worse a high % of chips on those wafers just fail and require cache slashes to function and be sold as lower end (Zen-eron couldn't resist sorry) parts it will be very bad for their future. If Intel needed a performance part to get back in a race owning their own fab makes that issue hurt a little less... and in a market with such small margins it could make all the difference.
AMDs cores are fairly mature. Its not like Zen is coming from a company that has never designed a X86 chip or something. AMD has(or had) pushed Intel with every release for a long time. Clock for clock they have often (almost always) had the better design. Granted they haven't pushed Intel in a long time and I don't know if they can now or not. I hope so... if not x86 will likely die in the next 5-7 years imo. ARM will very likely overtake x86 as the go to for desktops/laptops and the like by then. That is unless a strong AMD come back in the next couple years push Intel on performance... other wise as I see it Intel will instead go after lower power usage in an attempt to head off ARM and they will almost for sure get slaughtered on that score. It may always be true that Intel will have the fastest processor, but at some point it won't matter for the market. ARM at some point will be fast enough to cover even power users everyday usage.
For the record;
AMD K6 was the first real AMD chip that pushed Intel, it was based on the Nx686 designed by NexGen before AMD bought them. K6-2 was the release of the 3dnow extensions, AMD helped usher in all the cool Jumpsuit dancing Intel commericals .
K6 wasn't always faster then Intels PIIs but it was always close and did win some important stuff like old school 3dmark. It helped push up the date for PIII. (also why we got those stupid brick slot designs, so AMD couldn't release a faster K6-3 built to slot into the same mother boards... those stupid PIIIs Intel just stuck em on a board to keep AMD from slotting their chips in the same motherboards... Which backfired cause well it lead AMD to Athlon.)AMD K7(Athlon) I don't think I need to go over... they put the hurting on Intel. The only thing that kept Intel from really dropping into their rightful spot as the second place guys was good marketing. Damn those dancing techs back in the day my grandmother loved them and made me build her machine with Intel inside. lol In regards to Intel we can thank AMDs Athlon and A64 for putting a quick death to the abomination that was P4.
AMD K8 (Athlon64) was a very very good chip it was the first 64bit processor for the rest of us... and forced Intel to release their own 32 / 64 bit chip so fast that they just accepted AMDs extension of x86. So we can thank them for X86_64... and for Intel killing off the Itanium chips a few years later.
AMD K9 (Phenom) was also a solid chip at its release. You can thank AMD for still pushing Intel enough to push I7. Intel made a lot of changes to the Core chips to boost single thread performance... and with AMD selling inexpensive 8 core parts you can thank them for pushing Intel to introduce hyper threading with the Nehalem (Core i7), they had used the tech in P4s before... still Hyper threading wasn't ever part of the CoreDuo design, Intel wouldn't likely have bothered using it for i7 if AMD wasn't selling 8 core chips for a couple bills.
The Bulldozer design is where they made mistakes... they fell behind. They made some bad design choices. Its not a terrible chip its just wasn't enough to really push Intel... and its shown. Since Bulldozer was released 5 years ago x86 chips from Intel haven't really gotten any better. Sure they release some new stuff here and there and yes they have gotten faster but not at the rate they did when they had some pressure in the market. Had Bulldozer been been a real challenger in the mid range parts... Intel would have responded.
Zen is another new architecture from AMD... and their record as I see it right now is 4 of 5 winners. 4 of their last 5 architectures produced products that competed with Intel for market in a real way. 2 of those releases (K7 / K8) where hands down better chips then what Intel was selling at the time of their release. Two others (K6 and K9) could be argued where the winners in at least part of the market if not all. The win for all of us is Intel always came back with an answer.
Last edited:
It's all good. His reply may have been slightly curt. I can handle it though I have heard worse. He did get his point across in a linguistically economical way, there is value in that.
- Joined
- Aug 16, 2004
- Messages
- 12,429
Thanks for the history.. brings back a lot of memories. Still have my K6-2 333mhz in the drawer somewhere. I am due for an upgrade next year since I am still on a first-gen i7 platform.
I went from AMD K6-2 333mhz -> Athlon 64 3400+ -> Opteron 144 -> Too many Core 2 Duos -> Core 2 Quad Q6600 -> Q9650 -> i7 920 (2009) -> X5650 (2015).
Shintai
Supreme [H]ardness
- Joined
- Jul 1, 2016
- Messages
- 5,677
Even if we imagine that. Remember IPC is easy if the clocks are low. High clocks and high IPC, now that's the challenge. And obviously without having a nuclear reactor attached for power.
JustReason
razor1 is my Lover
- Joined
- Oct 31, 2015
- Messages
- 2,483
no prob wasn't complaining and yes it was relevant for HT for Zen just that it was all about current Intel so had to look after a number of post to be sure. I am interested to see how well that front end works on Zen and it is the whole reason I see this as a forward step even if it doesn't beat Intel (which I don't expect at all).
I had a K6 keychain at one time.
I haven't ran a full on AMD machine since Athlon 64 either. I skipped Phenom and ended up going Intel. Its made me pretty sad... I resolved I am heading back to AMD no matter what. At this point might as well hold out and see what Zen is all about.. worse case just get a smoking deal on an FX already, I mean heck there still great chips and already dirt cheap. FX chips are still some of the fastest for a few things like 265 compression... and not that it matters now with everything so darn fast but I do plenty of audio stuff and the FXs do the job very well. I better stop or I'm going to sell myself before zen hits. haha
Shintai
Supreme [H]ardness
- Joined
- Jul 1, 2016
- Messages
- 5,677
JustReason
razor1 is my Lover
- Joined
- Oct 31, 2015
- Messages
- 2,483
Or maybe just spend every waking moment blindly bashing a product real or imaginary without reason. Sound good to you? Seems like it. Do you actually game any at all? I do a lot hence why I only really post here in the morning (est) and sparingly through the day. I game at night be it WoW or currently FO4. Also why spend so much of your time bashing a product from a company you obviously have great disdain for. I dare say you spend more time and effort in AMD threads than any rabid AMD fan has ever. That doesn't make any sense at all.Yes, why buy on performance metrics when blind loyalty is better.
Yes, why buy on performance metrics when blind loyalty is better.
It's about realizing I was part of the problem. I bought into the bench mark is god crap last cycle and I won't again this time. Bottom line is for the stuff I actually do compressing video and working with large amounts of audio data even AMDs current chips are likely the best bang for my buck. As I can run a FX chip and spend the clams I save on a higher quality board and SSDs. For my exact use I know right now if I spend to the limit of my budget the current AMD option will allow me to build a faster machine. So ya Zen is right around the corner might as well wait and see what AMD has cooked up... worse case I pull the trigger on the FX system I was considering and perhaps score a good blow out price and use that extra $ for a faster or larger SSD.
- Joined
- Oct 4, 2007
- Messages
- 13,080
Yeah, I rather gave up on him a long time ago.
He does give me a good laugh from time to time but I realized nothing I say will change his point of view on AMD hardware. Me, I am looking forward to this and if I can time it with my tax return and they release in January or February, I will just upgrade both my computers and pass down my FX parts to someone else. 
Spirit_Retro
Gawd
- Joined
- Apr 1, 2010
- Messages
- 525
Yeah, I rather gave up on him a long time ago.
Hey- let me know on those FX parts.
These puppies still have legs... I'm having a hard time justifying an upgrade at this point. Yea the new stuff is faster- but I'm pretty sure I do not need faster.
Shintai
Supreme [H]ardness
- Joined
- Jul 1, 2016
- Messages
- 5,677
I think its already established that you buy AMD no matter how they perform. And that you already gave your money to AMD for Zen. Fanaticism at its best.
JustReason
razor1 is my Lover
- Joined
- Oct 31, 2015
- Messages
- 2,483
Pot meet Kettle. Are you seriously gonna condemn him on his personal rationale for his own purchase? He isn't trying to convince anyone to do the same just what he is gonna do. You on the other hand are applying your own criteria and belief to anyone in contradiction to it and then labeling them in any negative manner you can. Don't damn lest you be damned the same.
- Joined
- Oct 4, 2007
- Messages
- 13,080
Wow, he ha now labeled me.
Sure must be getting under his skin now. Zen full steam ahead for me! Yet you seem to make shit up about things you don't know, to support your twisted views? I suggest you should take a step back and start understanding some things before your post about any technical stuff? Shintai doesn't make things up so I suggest you take a step back and see what you do.
ManofGod we know he is an avid AMD supporter, nothing wrong with that, just don't get the crazy hype train going cause Zen will not the "best" it will be good but doesn't seem to be better then what Intel has right now or coming out shortly. Just hell of a lot better then what AMD have right now which is more important then anything else.
Last edited:
Gideon
Supreme [H]ardness
- Joined
- Apr 13, 2006
- Messages
- 4,186
Razor, Shintai does nothing but troll anything AMD. His sad argument is that AMD cant compete due to R&D costs, yet AMD has in the past produced better chips despite that fact. History already proved him wrong but facts rarely get in his way of trolling. Fact is some people love certain companies and that is fine if they want to do that. What is not ok is bashing someones choice due to the fact you just hate that company. This is a speculation thread so some will have high hopes others wont and thats fine. But to say someone is a idiot for having high hopes is nothing but trolling. Shintai is a very angry bitter person and I hope he gets the therapy he needs in life. Zen may be a great cpu or it may not, this is supposed to be a discussion on Zen, not bash the other guy cause he has different thoughts on how it will perform. Razor you may not always be the nicest guy but you always were willing to discuss your points, not just bash people. Shintai may not be the best person to defend.
JustReason
razor1 is my Lover
- Joined
- Oct 31, 2015
- Messages
- 2,483
first prove I made anything up. You keep skewing any argument I make into some tangent that really has nothing to do with the point at hand. I know it frustrates you so when you can't sway an argument into your favor. Even here where you make statements about a persons choice and still manage to throw in a little dirt as well.
maybe it is hard for you to understand but some people have other criteria for purchases that do not include performance crowns. Even in this case assuming~$400 16thread CPU, the number of threads is the metric with cost included as well as possible platform updates over previous CPUs. Pretty easy to understand if you aren't blinded by bias or arrogance.
- Joined
- May 18, 1997
- Messages
- 58,059
The next person to address the poster and not the topic will be banned. If you see this happen again in the thread, use the report post button.
Well then use facts to stop his trolling if that is the way you guys feel, that is what I do, but no you have people like JustReason, who has no reasons in his posts you have People calling people names because they don't know how to use facts, cause they have no facts or can't understand them!
What can someone say when they use facts in their statements to make a point?
Nothing, outside of countering with other facts......
Everything else is BS.
I'm not defending him, I'm defending the way he posts, he doesn't talk shit, he just presents his case with facts.
A troll by definition doesn't use facts to support arguments, they use half truths and lies or make shit up and go down the path of name calling, the only people here is who? Two guys that are posting against Shintai.
Just because you don't like what is posted, doesn't make a person a troll! Separate your affinity for what ever brand and just use facts to support your case and that is it.
Last edited:
aztekk
Limp Gawd
- Joined
- Nov 4, 2014
- Messages
- 166
First of all, it wasn't "K9", that never existed. The architechture was referred to as K10 upon release and K8N for the most part before it was launched (since it was based on K8). And it was not a huge success upon release, actually the original Phenom chips had a lot of issues, like the TLB bug but the biggest problem was their poor performance compared to Intel at the time, Conroe was in fact kicking its ass, and it only got worse after motherboards disabled parts of the cache. They had to actually stop shipments to the server market since no one wanted the B2 revision. B3 fixed it but the damage was already done.
K10 started to shine after the release of Phenom II, that's when it got competitive to a good extent. And that's probably the launch you're referring to.
P.S. razor1, you still here?
AMD hasn't been this far behind before and for this long,
For the record;
AMD K6 was the first real AMD chip that pushed Intel, it was based on the Nx686 designed by NexGen before AMD bought them. K6-2 was the release of the 3dnow extensions, AMD helped usher in all the cool Jumpsuit dancing Intel commericals .
AMD K7(Athlon) I don't think I need to go over... they put the hurting on Intel. The only thing that kept Intel from really dropping into their rightful spot as the second place guys was good marketing. Damn those dancing techs back in the day my grandmother loved them and made me build her machine with Intel inside. lol In regards to Intel we can thank AMDs Athlon and A64 for putting a quick death to the abomination that was P4.
AMD K8 (Athlon64) was a very very good chip it was the first 64bit processor for the rest of us... and forced Intel to release their own 32 / 64 bit chip so fast that they just accepted AMDs extension of x86. So we can thank them for X86_64... and for Intel killing off the Itanium chips a few years later.
Nope not correct, Intel's marketing had little do with what was going on during that time, AMD was able get 25% server marketshare and 50% pc marketshare outside of OEM's with Athlon 64 even with Intel's underhanded tactics of stopping in roads with OEM's. And now actually there are flaws and limitations with the x64 currently being used, Intel was trying to get around those flaws with Itanium, but this had nothing to do with either company, it had to do with Microsoft, Microsoft backed x64 by AMD because it was a natural fit from their OS and kept their software stack intact. If MS backed Itanium their software stack would have been brought back down to the level of Apple's OS and Unix, Linux, and lesser used OS's, which was not an option for them.
Nehalam crushed K9, this was the down fall for AMD, phenom wasn't that good of CPU and it wasn't pushing Intel much.
You are talking about something that never happened, Bulldozer was a fail, a big fail that AMD is still going through since phenom that is 1 decade, 2 designs both failed.
now with Zen we have no clue what its going to end up at. All we have is what AMD has shown us thus far against a cut down Broadwell, and we have no idea of the things they did to Blender either. So just going by the way AMD markets their stuff in the past, we can say that is the best they are going to get..... Which is not enough to take market share away from Intel, cause Intel will match price with AMD, cause they will not want to concede market share.
So what is the motivation for companies, individuals, to move away from Intel? A platform they have been comfortable with for the past 10 years? Nothing at all
First of all, it wasn't "K9", that never existed. The architechture was referred to as K10 upon release and K8N for the most part before it was launched (since it was based on K8). And it was not a huge success upon release, actually the original Phenom chips had a lot of issues, like the TLB bug but the biggest problem was their poor performance compared to Intel at the time, Conroe was in fact kicking its ass, and it only got worse after motherboards disabled parts of the cache. They had to actually stop shipments to the server market since no one wanted the B2 revision. B3 fixed it but the damage was already done.
K10 started to shine after the release of Phenom II, that's when it got competitive to a good extent. And that's probably the launch you're referring to.
P.S. razor1, you still here?
Think your right... thanks for the correction. I was thinking the second version... I sort of forgot about that crap first gen hicup. Point was at least Intel was concerned at that point... it helped drive them a bit, even if it turns out it wasn't really required.
They aren't that concerned, they have been planning for this yeah, but I don't think AMD is going to be able to push Intel much at least on to the degree we see a jump from Pentium to Nehalem.
Short term margin loss for Intel that they will sustain for maintaining status quo with AMD, will be back to the same differentiation if AMD doesn't go into a faster iteration of Zen. And we know Intel is prepping for transitions to 10nm mid to late next year at least for some of their chips, that means they will have a year advantage for 10nm, maybe more...
Can't argue that. Your right they are not in a great place right now... everyone knows that. Its the sort of thing that doesn't matter much imo. AMD only really needs a handful of OEM wins these days.... lets face it the majority of the desktops that do sell these days come from fewer and fewer oems. If zen is good enough and AMD can scrap together a bit of marketing money and sway a couple big wins that is all it would take to be back in the game. We'll see though. Everyone admits they have a steep hill to climb.
Itanium was a terrible solution that was never ever aimed at you and me. (assuming your using desktop chips... and not posting from some Cray supercomputer cluster). Could you imagine where we would be if we still all had a 4gb ram limit. Cause that is what Intel had planned for the industry based on X86. Itanium would never have shipped in a desktop ever. Intel accepted the standard as they clearly had zero choice. If MS selected amd over intels solution I think that speaks to me point. AMD was laying some hurt on Intel, not just with customers... but with their largest long time allies like MS. They where forced X86_64 or face the reality of being the second fiddle chip company and not being able to hide with any amount of marketing.
Seriously though MS did in fact ship Windows for Itanium. If you remember the performance of Itanium was so bad that it earned the nickname Itanic. MS did end windows for itanium before intel killed it... simply because no one was buying them. In fact Red hat dropped support before MS did. Opteron servers where destroying Itanic machines. For the record MS still offers support for Windows server Itanium version until July 2018.
I was corrected by another poster... I think I was thinking of Phenom II a bit the first gen I remember now did have a few launch issues. Having said that I don't disagree with you... I wouldn't list the phenom as AMDs best processor. The tech was solid enough though to again keep Intel worried enough to push. On paper the tech is solid even if the first few fab runs killed them... which is why I say Zen is going to come down to how well AMD executes their fab runs.
Anyway not really disagreeing with you... no it wasn't a better chip. The tech though was solid and enough to keep intel on their toes.
I don't believe it was the fail you think it was. They are still around they are still selling chips. They don't have the high end... no doubt. AMD is not selling anything right now that can really compete with the I7s in 95% of the tasks people are going to throw at a processor. We all know that. They are however doing a much better job with the low end stuff. the A10 line may not be sexy and get a lot of attention from tech sites... cause no power user [H] type is going to care much about chips they may buy for grandmas machine. Still the A10s compete very well. Don't get me wrong... Zen clearly needs to be more then that. Honestly we should all be wishing and hoping they have something more. If not... type it again X86 is on life support. Waiting around for the eventual invasion of the high performance ARM chips that will be good enough to push Intel into a very small market of high performance servers, gaming computer companies and the power user custom build market. Intel needs AMD more then anything. lol
Last edited:
You right... its the fabs. I did say it earlier as well, its possible AMD really died the day they sold the fabs. I'll still hold some hope for AMD... they surprised Intel way back with Athlon, and it was good. Intel was sure they beat that little upstart AMD and their purchased K6 chip over the head with there stupid socket change on PIII... and AMD went and made a chipset. Lets hope they can pull out one more surprise, perhaps lightning can strike twice.
Well even though AMD is in a bad place, they have made quite a few strides to solve their problems.
Itanium was a terrible solution that was never ever aimed at you and me. (assuming your using desktop chips... and not posting from some Cray supercomputer cluster). Could you imagine where we would be if we still all had a 4gb ram limit. Cause that is what Intel had planned for the industry based on X86. Itanium would never have shipped in a desktop ever. Intel accepted the standard as they clearly had zero choice. If MS selected amd over intels solution I think that speaks to me point. AMD was laying some hurt on Intel, not just with customers... but with their largest long time allies like MS. They where forced X86_64 or face the reality of being the second fiddle chip company and not being able to hide with any amount of marketing.
Seriously though MS did in fact ship Windows for Itanium. If you remember the performance of Itanium was so bad that it earned the nickname Itanic. MS did end windows for itanium before intel killed it... simply because no one was buying them. In fact Red hat dropped support before MS did. Opteron servers where destroying Itanic machines.
Oh yeah I know MS did, but did that much after they endorsed AMD's x64. And with the delays to Itanium, and Itanium 2 with negligible performance benefits, Itanium wasn't really an option.
I don't believe it was the fail you think it was. They are still around they are still selling chips. They don't have the high end... no doubt. AMD is not selling anything right now that can really compete with the I7s in 95% of the tasks people are going to throw at a processor. We all know that. They are however doing a much better job with the low end stuff. the A10 line may not be sexy and get a lot of attention from tech sites... cause no power user [H] type is going to care much about chips they may buy for grandmas machine. Still the A10s compete very well. Don't get me wrong... Zen clearly needs to be more then that. Honestly we should all be wishing and hoping they have something more. If not... type it again X86 is on life support. Waiting around for the eventual invasion of the high performance ARM chips that will be good enough to push Intel into a very small market of high performance servers, gaming computer companies and the power user custom build market. Intel needs AMD more then anything. lol
Bulldozer and Phenom were not good chips lol, that is why AMD lost so much ground as quickly. It was like within 2 quarters they lost everything Athon 64 and x2 gave them.
I have very little hope with GF's, they haven't showed anything about being capable of being better then another other fab let alone Intel, where Intel is just years in front.
Its going to be what makes it or breaks it for sure. They have lost a lot of ground to Intel the last 5-6 years. If the leaked benchmarks aren't complete BS though.. Zen still may turn out to be a great mid range chip. I am not crazy I don't believe AMD is going to release a Zen part that destroys Intels top end. What they really need is a solid $200-300 dollar chip that makes the I5s look bad. If they can sneak in a few OEM wins with that sort of part things may get interesting again. As you have said though Intel has built a pretty sizable war chest. When the first Athlons hit Intel did almost give chips away to oems and spent a fortune paying Casey Kasem to narrate dancing tech and blue man group tv spots. If AMD manages something like that who knows we could be seeing a I7 rebrand and dancing techs again... ok honestly now I sort of wanna see some new tech commercials. (I swear though if I see a Dancing Intel tech commercial set to dubstep music I'll murder someone.) haha
Well I expect from what we have seen so far and all the white papers and patents about Zen seem to show it will do well, but again, I'm not putting too much stock into anything till we get independent reviews lol. Too many times has AMD showed or talked about upcoming products and when they come out they aren't what was shown or talked about.
OEM's are already selling AMD (after the lawsuit of monopolistic tactics of Intel was settled) I don't see much change here, because they aren't offering anything Intel can't. (HP, Dell, Lenovo, all have AMD products, which are priced cheaper then Intel counterparts right now, but sales for those systems are just lower, because people want Intel since they know Intel is better even though price parity wise they aren't)
These kinds of markets don't shift because something new comes out that are the same as whats already out there. The only way these types of markets shift is when something truly disruptive is released, like the Athlon and x64 and x2's. That is not what Zen is about at least from what we have seen thus far. It will disrupt the mobile market (because AMD is not there right now) but not the desktop or server market.
The GPU market is the same way, did AMD really gain much since the launch of Polaris? Yeah mobile they did, but desktop discrete, not really, not only that they have priced themselves to a point where margins look to be razor thin, and they still aren't gaining ground? That is why its hard to shift markets like this.
Just think of this, why would vendors, data centers, server houses, want more then one type of system for their infrastructure, sales, etc, when they don't need to? Cost effectiveness drops. You need to have different contracts, different warranty programs for different systems. Also technical expertise will have to be different too when things go wrong. Now if AMD was 25% faster at the same cost, well then yeah they will start shifting over as they replace existing systems and for vendors people will want that extra performance. But this isn't the case. Not only that, Intel is already pricing their xeon chips much less then what we see in Ark, 2/5ths the price, and with volume discounts. So AMD to compete really with Intel doesn't have the resources to do a battle in price. They have to figure something else out. This is why if the rumors are true of 600 or 500 8 core part, better believe its going to performance like a 500 or 600 Intel part, or just a bit higher, but nothing ground breaking.
OEM's are already selling AMD (after the lawsuit of monopolistic tactics of Intel was settled) I don't see much change here, because they aren't offering anything Intel can't. (HP, Dell, Lenovo, all have AMD products, which are priced cheaper then Intel counterparts right now, but sales for those systems are just lower, because people want Intel since they know Intel is better even though price parity wise they aren't)
These kinds of markets don't shift because something new comes out that are the same as whats already out there. The only way these types of markets shift is when something truly disruptive is released, like the Athlon and x64 and x2's. That is not what Zen is about at least from what we have seen thus far. It will disrupt the mobile market (because AMD is not there right now) but not the desktop or server market.
The GPU market is the same way, did AMD really gain much since the launch of Polaris? Yeah mobile they did, but desktop discrete, not really, not only that they have priced themselves to a point where margins look to be razor thin, and they still aren't gaining ground? That is why its hard to shift markets like this.
Just think of this, why would vendors, data centers, server houses, want more then one type of system for their infrastructure, sales, etc, when they don't need to? Cost effectiveness drops. You need to have different contracts, different warranty programs for different systems. Also technical expertise will have to be different too when things go wrong. Now if AMD was 25% faster at the same cost, well then yeah they will start shifting over as they replace existing systems and for vendors people will want that extra performance. But this isn't the case. Not only that, Intel is already pricing their xeon chips much less then what we see in Ark, 2/5ths the price, and with volume discounts. So AMD to compete really with Intel doesn't have the resources to do a battle in price. They have to figure something else out. This is why if the rumors are true of 600 or 500 8 core part, better believe its going to performance like a 500 or 600 Intel part, or just a bit higher, but nothing ground breaking.
Last edited:
EuphoricRage470
[H]ard|Gawd
- Joined
- Apr 15, 2011
- Messages
- 1,105
There are basically no laptops with Polaris in them, aside from the Apple MacBook Pro and huge, bulky, DTR laptop like this one (http://radeon.com/en-us/alienware-unveils-new-generation-of-radeon-rx-470-laptops/). The big unit share gains in mobile came from sales of old, cheap, low-performance AMD discrete GPUs in laptops. Nothing that would really be considered gaming class.
Go to Newegg and look at the gaming laptop selections. All GeForce.
Well which way Zen goes will decide the future of desktop computing as I see it. If its good enough to keep Intel pushing and people talking about X86 it might help stave off the ARM uprising that is on the horizon. I mean heck AMD is already shipping at least in small numbers one ARM server chip. From what I have seen of numbers it has pretty decent performance as well. No not high end desktop type performance... still impressive numbers for a low power sipping part. It can more then hold its own agains the likes of FX8300s which isn't bad for a 23w part. Intel has also announced they will be designing and fabbing some ARM chips for phones at least.
If Zen flopps... I could see AMD talking a swing for the fences respin of their Opteron A chips. At least one advantage for AMD would be the fab situation. Would be a lot easier to hire out quality fabs for some ARM core chips. If they got a company like Samsung (who could use some good press) to do the fabbing on a beefed up A1100 built for a mid range cost Samsung desktop with a nice screen, something like the surface studio... but with some interesting Samsung take on the idea and a price tag that would move them. haha ok now I'm dreaming.
Seriously though go Zen... x86 is the only hope if we still want to be able to build our own machines in the future.
If Zen flopps... I could see AMD talking a swing for the fences respin of their Opteron A chips. At least one advantage for AMD would be the fab situation. Would be a lot easier to hire out quality fabs for some ARM core chips. If they got a company like Samsung (who could use some good press) to do the fabbing on a beefed up A1100 built for a mid range cost Samsung desktop with a nice screen, something like the surface studio... but with some interesting Samsung take on the idea and a price tag that would move them. haha ok now I'm dreaming.
Seriously though go Zen... x86 is the only hope if we still want to be able to build our own machines in the future.
Zen is AMD's last hope, but how long it will last as a savoir is what is worrisome, Intel is not going to give much room for AMD to make strides, they have the capabilities to really hurt AMD even with Zen,
Personally I think it would be better if AMD is bought out once their debt is taken care of, cause they really don't have the resources to go against both nV and Intel. Just going up against either of those two companies is bad enough, but both.
Personally I think it would be better if AMD is bought out once their debt is taken care of, cause they really don't have the resources to go against both nV and Intel. Just going up against either of those two companies is bad enough, but both.
drescherjm
[H]F Junkie
- Joined
- Nov 19, 2008
- Messages
- 14,933
I fully expect Zen will be a success if the 40% IPC (putting the IPC somewhere between Ivy Bridge and Haswell) that AMD has stated is accurate. If not and the 8C / 16T is has to be priced lower than an i5 to compete this is the end of the game for AMD.
http://wccftech.com/amd-zen-8-core-35ghz-cpu-spotted/
Well if this is to be believed... SR7s their top end will have a MSRP of $500 and $350 for 3.0ghz and 3.2ghz. They claim it will be punching in the i7 5960X class. They have an engineering number f1D3201A2M88F3_35/32_N could be good info. We'll see where they shake out when they're out and about getting reviewed by people like Kyle for sure. From leaks we have seen from older samples running 400-500mhz slower though... if we believe them. A zen clocked at 3.2 may well surprise some folks.
EuphoricRage470
[H]ard|Gawd
- Joined
- Apr 15, 2011
- Messages
- 1,105
AMD took care of a bunch of its debt by issuing stock. The recent rise in AMD's stock price allowed them to clean up the balance sheet a bunch.
That kind of success is only good enough to get into the black and get rid of the rest of their debt. But its not enough to gain market share and take the fight to Intel, they need to do more to do that, which if Intel has no response to Zen right now or not planning for anything from Zen, then AMD has 3 years, which if you go by their balance sheets, they will not get too much money maybe a billion a year? Three years, 3 billion. Well guess what that will not be enough to sustain them if Intel decides to up the ante with future products. Its easier for Intel to do because they have the resources and they are still per core faster and have node advantages.
This is a multi year, multi generation battle, not something that shifts within 6 months to year or even two or 3 years.
AMD shouldn't have to many issues with cash... they aren't going away anytime soon. When it comes down to it plenty of large investors are still willing to sink cash into AMD. Cause lets face it if they ever did go tits up their IP alone would be worth a fortune. I'm not saying the have the cash on hand to go to war like Intel does though. Should be a fun Jan-Feb 2017. If zen turns out to be what the leaks are claiming and performs = to Intel chips selling for 2x the price. Even if it does't sell like mad it will keep investors happy. Also I know I mentioned AMDs ARM ambitions, I think that still has plenty of investors happy on that front as well. Lots of people still betting on ARM servers to take off at some point and AMDs optron name still has some weight in that market. Even Polaris... as an investor I may not really care that Nvidia is selling more gaming cards that have razor thin margins. I'm looking at that slimmer gpu design and thinking AMD is considering it for SOC chips. At this point AMD could enter the high end phone market by simply retooling the Optron A1100 core and slapping on a Polaris GPU. The Polaris based cards may be low ball for PC VR... but a AMD ARM chip with a Polaris GPU in an android phone? Why not they have the tech and Samsung is clearly a friend in regards to the fabs. Both companies are going to be under some pressure this year... who knows I wouldn't be shocked to see them announce a High end Samsung phone powered by and AMD-ARMtron / Polaris Samsung Fabbed SOC.