AMD Further Unveils Zen Processor Details

haste. said:
You knew exactly what I meant. Most everyone assumes when someone refers to "Intel core architecture" they are referring to 2006+. Yes it was born from Pentium Pro, but your post comes off like the neck beard but that says "but actually" and everyone rolls their eyes.
Click to expand...

But actually 'neck beard' is formed by concatenating the words "neck" and "beard" to form neckbeard.
 
Dan_D said:
If it only matches Broadwell in IPC but doesn't clock worth a shit, people will complain about the clock speed. If final silicon shows clocks that significantly favor Intel, then it won't matter. Zen will be a loser in the desktop market. Low stock clocks also likely indicate poor overclocking performance. If the fastest Zen CPU comes at at 3.0GHz, I wouldn't expect it to clock as high as Broadwell or Skylake. As I've said before, IPC is only part of the equation. Actual clock speeds are what we need to really get an idea of how Zen will perform against Intel's offerings.
Click to expand...
I agree totally from an enthusiast point of view. But 3.0 - 3.2 is pretty much standard for every other market even for 2S servers. You would be hard pressed to get too much higher than 3.2 - 3.4 with more than 4 cores without paying some serious cash in the server market and most people hit that or under for HTPCs and stuff. Of course the higher Zen clocks (if the IPC is the same) then the more competitive it will likely be but IPC and especially efficiency matters A LOT. At this point for me IPC and efficiency almost matter more because it affects so much especially when we start talking about density.

This was the main reason I dropped my 2S Opteron build. While it was cheap to build and performance wasn't an issue density and especially efficiency were major issues. The only way that build made sense is if I was dealing with streaming or well threaded workloads. In all other scenarios the power cost just made the build more of a fun project since the only time it made sense is when I was doing VM load testing and video encoding.

If Zen can hit I'd say 3.2 - 3.4 in any core configuration then it will be fine for 95% of the market especially if it comes in under current pricing. Could Intel release higher clock parts? Sure but right now the IPC difference is so massive between BD arcs and Broadwell that even meeting IPC is more than enough to at least make AMD viable in anything above ULV/P builds.
 
kac77 said:
I agree totally from an enthusiast point of view. But 3.0 - 3.2 is pretty much standard for every other market even for 2S servers. You would be hard pressed to get too much higher than 3.2 - 3.4 with more than 4 cores without paying some serious cash in the server market and most people hit that or under for HTPCs and stuff. Of course the higher Zen clocks (if the IPC is the same) then the more competitive it will likely be but IPC and especially efficiency matters A LOT. At this point for me IPC and efficiency almost matter more because it affects so much especially when we start talking about density.

This was the main reason I dropped my 2S Opteron build. While it was cheap to build and performance wasn't an issue density and especially efficiency were major issues. The only way that build made sense is if I was dealing with streaming or well threaded workloads. In all other scenarios the power cost just made the build more of a fun project since the only time it made sense is when I was doing VM load testing and video encoding.

If Zen can hit I'd say 3.2 - 3.4 in any core configuration then it will be fine for 95% of the market especially if it comes in under current pricing. Could Intel release higher clock parts? Sure but right now the IPC difference is so massive between BD arcs and Broadwell that even meeting IPC is more than enough to at least make AMD viable in anything above ULV/P builds.
Click to expand...

I agree with you. I wasn't really thinking about it as a server part at the time I made that comment. Without knowing the TDP, clock range or platform details it is hard to speculate on the viability as a server part. My guess is that AMD will probably do ok there so long as the TDP isn't horrible, assuming the IPC performance is where this information says it is.
 
Blender is one of the major, popular 3d modelling programs.... I don't know anyone that makes money with cinebench, nor anyone who plays fire3dstrikemarkbench..
 
N4CR said:
Blender is one of the major, popular 3d modelling programs.... I don't know anyone that makes money with cinebench, nor anyone who plays fire3dstrikemarkbench..
Click to expand...

You are preaching to the choir. I've been telling people that Cinebench isn't a test of stability, and that 3D Mark is about as useless a "tool" as there is. We've seen systems that can do Cinebench all day and not crash. Fire up Handbrake on the same box and it will lock up in seconds. 3D Mark's performance doesn't equate to anything in the real world as the variables that effect its performance aren't the same ones that impact performance in a game engine. Its also not a great stability test.
 
haste. said:
You knew exactly what I meant. Most everyone assumes when someone refers to "Intel core architecture" they are referring to 2006+. Yes it was born from Pentium Pro, but your post comes off like the neckbeard but that says "but actually" and everyone rolls their eyes.
Click to expand...
Only someone who thinks Intel's marketing labels are meaningful and has no knowledge of processor microarchitecture thinks "Core" begin in 2006+.
I was working as a processor architect for Intel during that period. I knew the people doing the work. You want to spread nonsense based on marketing terms, go ahead, but don't expect it to pass muster among people with an engineering clue.
 
Dan_D said:
You are preaching to the choir. I've been telling people that Cinebench isn't a test of stability, and that 3D Mark is about as useless a "tool" as there is. We've seen systems that can do Cinebench all day and not crash. Fire up Handbrake on the same box and it will lock up in seconds. 3D Mark's performance doesn't equate to anything in the real world as the variables that effect its performance aren't the same ones that impact performance in a game engine. Its also not a great stability test.
Click to expand...

People put to much weight to certain benchmarks that is all. What is relevant for "gamers" might not be for everyone , some of the benchmarks mentioned have some weight but in the end if all you do is play DOTA 2 it really does not matter how much firestrike or cinebench does.

But the idea is that there is a measurable difference in Zen , how you care to show it is clear if AMD can compile the source it can compile with the right flags and so on. But for games this is different. Some benchmarks suffer from the same problem where source is not compiled when AMD cpu would benefit from it.

If Zen cleaned up in all the benchmarks then there would be more then just blender.
 
Pieter3dnow said:
People put to much weight to certain benchmarks that is all. What is relevant for "gamers" might not be for everyone , some of the benchmarks mentioned have some weight but in the end if all you do is play DOTA 2 it really does not matter how much firestrike or cinebench does.

But the idea is that there is a measurable difference in Zen , how you care to show it is clear if AMD can compile the source it can compile with the right flags and so on. But for games this is different. Some benchmarks suffer from the same problem where source is not compiled when AMD cpu would benefit from it.

If Zen cleaned up in all the benchmarks then there would be more then just blender.
Click to expand...

if that would be the case then I think, they would have proudly show any possible benchmark in their hands where Zen Outperformed or performed on pair with Intel similar CPUs.. that's why im skeptical until real reviews go out to the light.
 
N4CR said:
Blender is one of the major, popular 3d modelling programs.... I don't know anyone that makes money with cinebench, nor anyone who plays fire3dstrikemarkbench..
Click to expand...

How much can one make running Blender?
 
Araxie said:
if that would be the case then I think, they would have proudly show any possible benchmark in their hands where Zen Outperformed or performed on pair with Intel similar CPUs.. that's why im skeptical until real reviews go out to the light.
Click to expand...

Rightly so. AMD has often made outlandish claims about its performance numbers, showcasing its CPUs in one or two benchmarks hand picked to show them in the best light. Once they land in the hands of reviewers the truth comes to light and we see that the actual performance was only worth a damn in those benchmarks AMD showed us ahead of time and nothing else.

There has also been a lot of commentary in this thread about AMD showing off more benchmarks if the IPC was truly competitive with Broadwell, or showing off higher clocks than they have if Zen could achieve them. That's not necessarily true. Even when AMD has a really competitive product, it has always played its cards close to the vest. AMD usually goes pretty much silent before a processor launch. In most cases this is thought to be because of its total lack of competitive products, but I believe its more symptomatic of their business strategy. Many people don't understand just how much more R&D budget Intel has, or how much more manufacturing capacity it has. Intel is very conservative with its clocks as well all know from the massive overclocking headroom its worst chips generally have. If AMD shows an edge too early, Intel will follow up quickly with new products shortening AMD's window to sell chips before Intel closes the performance gap again.

I've not seen Zen in person yet, so I'm speculating on what we can expect naturally. I don't know if Zen can match Broadwell's IPC in anything but what they've shown, nor what clocks we can expect out of its enthusiast parts. It's easy to say they'd show higher clocks if it could, but its actually better if they don't. Showing some IPC equality with Broadwell is enough to show that they've likely exceeded our expectations with their earlier claims about Zen having 40% more IPC over Excavator. We've often speculated that this puts it in range of Haswell more than Broadwell or Skylake. So AMD's done just enough to show that Zen will be more competitive than many of us might have thought. At the same time, they are playing it safe. Its not about disappointing the consumer, but about not showing its hand to its rival.

The truth is, Intel could wipe AMD off the face of the CPU market at any time it decides it want's to sit in court for its anti-competitive practices. Financially, this has pretty much always been the case for the last 25 years.
 
Last edited:
New regime seems to keep the super stupid claims to a minimum. The community seems to perpetuate the outlandish things like RX 480 at 1700+ on air. Or was that WCCFTECH? Hmm. :)
 
cageymaru said:
New regime seems to keep the super stupid claims to a minimum. The community seems to perpetuate the outlandish things like RX 480 at 1700+ on air. Or was that WCCFTECH? Hmm. :)
Click to expand...

Indeed, very little information, and not a lot of hype from AMD. Now the RX 480 launch remains interesting to me cause of the rumor that it could reach +1400 using AIBs, that Wccftech reported on, and I think did so in a matter of being positively correct, where Videocardz placed it as rumors. Now the +1400 rumors had stuff supposedly coming from China, and I personally remember this one guy on Overclock that also appeared in Wccftech's comments supposedly saying things like how AIBs could reach it, etc, etc.

Now, part of it is always hype, hype is what leads people to believe that these things are possible in the first place, but I wonder about the rumor mill and "leaks" that came out prior to the launch, there were a few that claimed so, I wonder where they came from. One person? Multiple people that wanted to crash the party?

The fun thing was the Pascal releases were restrained in comparison to hype, leaks, and when leaks did come out, at least according to Videocardz, were on point to the product. Of course, at this point, Nvidia products don't really get the pre-launch hype hope that AMD gets, so there is that.
 
Simple fact is you dont show too much, the competition is always watching and Intel has the cash to make a adjustment if they really think they need to. This whole Blender bench to me is just showing us they did in fact make a pretty good new chip but the rest was kept in the dark on purpose. Amd looks to be on the way back to being competitive with Intel and if they are really lucky might take the crown away depending on clock speeds for a bit. But I think they would be just happy as can be to be within 90% of Intel performance.
 
I agree. It would be foolish for AMD to tip their hand too soon. What they have revealed is that Zen can be competitive, without showing their whole hand.

Like others, I would guess that Zen won't outright beat Intel head-to-head (especially if we are talking about a 2017 release) but should come close enough that they are a viable option on the high-end.
 
  • Like
Reactions: N4CR
like this
cybereality said:
I agree. It would be foolish for AMD to tip their hand too soon. What they have revealed is that Zen can be competitive, without showing their whole hand.

Like others, I would guess that Zen won't outright beat Intel head-to-head (especially if we are talking about a 2017 release) but should come close enough that they are a viable option on the high-end.
Click to expand...
Gideon said:
Simple fact is you dont show too much, the competition is always watching and Intel has the cash to make a adjustment if they really think they need to. This whole Blender bench to me is just showing us they did in fact make a pretty good new chip but the rest was kept in the dark on purpose. Amd looks to be on the way back to being competitive with Intel and if they are really lucky might take the crown away depending on clock speeds for a bit. But I think they would be just happy as can be to be within 90% of Intel performance.
Click to expand...
Guys get real Intel hasn't lost sleep on anything AMD done last few years why should they start now? Intel has a strangle hold on certain key areas so they will make a little bit less money, for Intel not a big deal the market as it is is "shrinking" so it is expected to make less money. In the long term AMD has to do a lot of work to get "back" and that Intel will try to counter but not the "second" they launch Zen.
 
Pieter3dnow said:
Guys get real Intel hasn't lost sleep on anything AMD done last few years why should they start now? Intel has a strangle hold on certain key areas so they will make a little bit less money, for Intel not a big deal the market as it is is "shrinking" so it is expected to make less money. In the long term AMD has to do a lot of work to get "back" and that Intel will try to counter but not the "second" they launch Zen.
Click to expand...

Neither of them you quoted mentioned anything about Intel being worried about AMD, just that AMD needs to manage information about Zen well and close to their chest.
 
N4CR said:
Ask this guy Interview: Sean Kennedy on Blender in Hollywood

Dan, that wasn't directed at you, I know [H] doesn't like canned crap. I should've quoted Ieldra, sorry. Figured someone into cpus would generally know blender.
Click to expand...

Well, canned benchmarks for CPUs are one thing. I don't personally have issues with that. The problem is that the results with video / GPU testing has shown that the canned results and real world test data do not always correlate. With CPU and motherboard testing you don't get the same problems.
 
Pieter3dnow said:
Guys get real Intel hasn't lost sleep on anything AMD done last few years why should they start now? Intel has a strangle hold on certain key areas so they will make a little bit less money, for Intel not a big deal the market as it is is "shrinking" so it is expected to make less money. In the long term AMD has to do a lot of work to get "back" and that Intel will try to counter but not the "second" they launch Zen.
Click to expand...

Intel won't counter the second Zen is released, but the gap that AMD has to sell processors at a higher profit margin can be shrunk signifantly if AMD tips its entire hand MONTHS before release day. Intel can easily counter if given enough time. They probably wouldn't do so unless AMD outclassed a significant portion of Intel's product line which seems highly unlikely. Either way, companies keeping secrets from each other is a very real concern and I've seen the back and fourth, IP stealing and all of that behind the curtain. They mitigate it as much as they can knowing full well that it's like weeds in a garden. At best all they can do is control the problem. They can't cure it forever.
 
Dan_D said:
Intel won't counter the second Zen is released, but the gap that AMD has to sell processors at a higher profit margin can be shrunk signifantly if AMD tips its entire hand MONTHS before release day. Intel can easily counter if given enough time. They probably wouldn't do so unless AMD outclassed a significant portion of Intel's product line which seems highly unlikely. Either way, companies keeping secrets from each other is a very real concern and I've seen the back and fourth, IP stealing and all of that behind the curtain. They mitigate it as much as they can knowing full well that it's like weeds in a garden. At best all they can do is control the problem. They can't cure it forever.
Click to expand...

Both AMD and Intel knows one another's products way in advance. I bet you Intel got Zen running in its labs to see exactly how it performs. While AMD got Kaby Lake and others in theirs. The chains and logistics are to big, with so many partners involved that you cant keep any secrets. And so does any big customers. Google, Baidu, Amazon, MS, Alibaba, Facebook and so on.

The only thing you can keep a secret is exact consumer release date and price.
 
Shintai said:
Both AMD and Intel knows one another's products way in advance. I bet you Intel got Zen running in its labs to see exactly how it performs. While AMD got Kaby Lake and others in theirs. The chains and logistics are to big, with so many partners involved that you cant keep any secrets. And so does any big customers. Google, Baidu, Amazon, MS, Alibaba, Facebook and so on.

The only thing you can keep a secret is exact consumer release date and price.
Click to expand...

Man I'm pretty sure I read about AMD/Intel paying 3rd parties to 'reverse engineer' their CPUs; accurate die shots and scanning tunneling microscopy
 
I'd buy Zen if it was somewhat competitive with Intel and using a modern chipset, not something from 8 years ago.

Seriously AMD, update your stinking chipsets.
 
Trimlock said:
I'd buy Zen if it was somewhat competitive with Intel and using a modern chipset, not something from 8 years ago.

Seriously AMD, update your stinking chipsets.
Click to expand...

I'm sure they will. To what extent is largely unknown. In the past AMD has been pretty good about some aspects of it, but they've always suffered when it comes to things like USB I/O and the like. They could half ass it and let companies like ASUS integrate third party controllers to make up for a lot of it. We'll have to wait and see.
 
Dan_D said:
I'm sure they will. To what extent is largely unknown. In the past AMD has been pretty good about some aspects of it, but they've always suffered when it comes to things like USB I/O and the like. They could half ass it and let companies like ASUS integrate third party controllers to make up for a lot of it. We'll have to wait and see.
Click to expand...
I'm mostly on with that except the fact that OEMs are terrible with driver support.

Asus has been pretty good with their motherboard support though.
 
Shintai said:
Both AMD and Intel knows one another's products way in advance. I bet you Intel got Zen running in its labs to see exactly how it performs. While AMD got Kaby Lake and others in theirs. The chains and logistics are to big, with so many partners involved that you cant keep any secrets. And so does any big customers. Google, Baidu, Amazon, MS, Alibaba, Facebook and so on.

The only thing you can keep a secret is exact consumer release date and price.
Click to expand...

Once again, an opinion marketing as truth. Do you have verifiable proof of this? Sorry but, I cannot imagine, short of violating a NDA or giving a product to a competitor from a reviewer that was not allowed to do so, I cannot see it happening. Both are definitely giving up corporate secrets and are illegal.
 
Trimlock said:
I'd buy Zen if it was somewhat competitive with Intel and using a modern chipset, not something from 8 years ago.
Seriously AMD, update your stinking chipsets.
Click to expand...
What a weird comment to make since this year plenty of AM3+ motherboards were released with updated features as USB 3.1 , if you followed Kyle's Test Setup and Reasoning - DX11 vs DX12 AMD CPU Scaling and Gaming Framerate

Then you might have spotted GIGABYTE - Motherboard - Socket AM3+ - GA-990FX-Gaming (rev. 1.0)

which should be pretty decent feature wise ...
 
I hope the release date doesn't continue to keep slipping. So it's "sometime" in 2017 now?

That would imply to me that either:

> Genuinely more can be gotten out of the Zen design - I suppose that's a positive

> The Zen design doesn't deliver as expected (i.e. Bulldozer) i.e. they're now in salvage mode.

Comparisons with down clocked Intel doesn't exactly inspire. If they were trying to demonstrate IPC improvements - why didn't they do they inverse and overclock the Zen? (You could say that's a rhetorical question)


Won't more people just pick up a Intel chip in the mean time?

Given the slow rate of progress over the past few years - if people purchase anything recent then I can't see why they'd have a genuine to reason to switch to AMD unless they have money to burn.

If the STEAM survey is any indication then there's a good 30% of people there currently who have mid-range <=2GHz~ chips
Steam Hardware & Software Survey

I'd like to see all of those people pick up 3GHz+ Zen Quad/Octa cores.


If Zen does deliver couldn't Intel just afford to slash their pricing? Won't that destroy the usual incentive to buy AMD?
Be nice if AMD could make more of a dent into the business market.
Scrapping the bottom of the barrel in the bargain basement market as usual won't get them anywhere.
 
N4CR said:
Ask this guy Interview: Sean Kennedy on Blender in Hollywood

Dan, that wasn't directed at you, I know [H] doesn't like canned crap. I should've quoted Ieldra, sorry. Figured someone into cpus would generally know blender.
Click to expand...

Yeah, my bad. I thought it was just another benchmark.

That said, I almost get a "preemptive excuse making" vibe. I think it's great that Zen is neck and neck with BW-E in a Blender render, but I think it's a bit silly to do a wholesale dismissal of all other known CPU benchmarks since they are not what AMD chose to use in this demonstration. What if Zen kills it in Cinebench? Still worthless? I downloaded the Blender BMW benchmark, but of course it wouldn't run without Blender itself installed so I'll get around to that now that the weekend's over. In the video, AMD did a render of the Zen logo. Looks like that file would need to be provided if everybody is going to use it as a benchmark. And let's face it - most guys are going to fire off a known benchmark to compare rigs, not download a rendering suite and one of their provided benchmark files.

Dan mentioned Handbrake, that's a program I am more familiar with, but useless as a benchmark without a predetermined file and settings - a "canned" scenario, if you will. Anybody running Handbrake will darn sure appreciate 8c/16t. I run it on 4c i5 w/o HT, 4c/8t i7, and 8c/16t i7 and the difference across platforms is noticeable to say the least. ASUS ROG Realbench has a benchmark that runs 2 GPU intensive tests (one is a system-wide test, but the score relies partially on GPU performance) and 2 CPU intensive tests, a GIMP render and a Handbrake encode. It provides scores for the individual tests. I guess we'll have to see how Zen performs in Realbench, it might need to be dismissed as well.
 
  • Like
Reactions: N4CR
like this
N4CR said:
Blender is one of the major, popular 3d modelling programs.... I don't know anyone that makes money with cinebench, nor anyone who plays fire3dstrikemarkbench..
Click to expand...

Interesting. I had never heard of it before, but I haven't really played with 3d modeling since the days of PovRay, 3D Studio Max and Lightwave.

Dan_D said:
You are preaching to the choir. I've been telling people that Cinebench isn't a test of stability, and that 3D Mark is about as useless a "tool" as there is. We've seen systems that can do Cinebench all day and not crash. Fire up Handbrake on the same box and it will lock up in seconds. 3D Mark's performance doesn't equate to anything in the real world as the variables that effect its performance aren't the same ones that impact performance in a game engine. Its also not a great stability test.
Click to expand...

I think the reason Cinebench has been pretty popular as a benchmark is because it is easy to isolate a single core and run a single threaded test, so that core count doesn't obfuscate the result.

There aren't enough single threaded benchmarks in wide use. When I go to anandtech.com/bench to get some data, single threaded Cinebench R11.5 is really it. (not sure why they don't use the newer R15)
 
Zarathustra[H] said:
Interesting. I had never heard of it before, but I haven't really played with 3d modeling since the days of PovRay, 3D Studio Max and Lightwave.



I think the reason Cinebench has been pretty popular as a benchmark is because it is easy to isolate a single core and run a single threaded test, so that core count doesn't obfuscate the result.

There aren't enough single threaded benchmarks in wide use. When I go to anandtech.com/bench to get some data, single threaded Cinebench R11.5 is really it. (not sure why they don't use the newer R15)
Click to expand...
Blender got a lot of use in Skyrim modding. Making new armors and baddies is fun but a lot of work.

As far as cinebench, they likely haven't moved to the newer because it makes the database irrelevant as far as comparison material. Wish they would though to get away from the ICC issue in 11.5.
 
JustReason said:
Blender got a lot of use in Skyrim modding. Making new armors and baddies is fun but a lot of work.

As far as cinebench, they likely haven't moved to the newer because it makes the database irrelevant as far as comparison material. Wish they would though to get away from the ICC issue in 11.5.
Click to expand...


Well, the funny part is that they have Cinebench R15 results under the "legacy workloads" tab, along with R10.

So at some point they used R15, and discontinued it, but kept 11.5 which is a little odd..
 
ManofGod said:
Once again, an opinion marketing as truth. Do you have verifiable proof of this? Sorry but, I cannot imagine, short of violating a NDA or giving a product to a competitor from a reviewer that was not allowed to do so, I cannot see it happening. Both are definitely giving up corporate secrets and are illegal.
Click to expand...

What, you think just because it's unethical, and illegal, they don't do it?
 
Ruoh said:
What, you think just because it's unethical, and illegal, they don't do it?
Click to expand...

To a certain extent, yes. (Based upon if they think they can get away with it and if the fines are something they are willing to eat, that is.) Otherwise, no, they do not do it as a rule, in my opinion. That is not something you can simply hide and I would have to see evidence of that to completely buy it.
 
Few thoughts on this benchmark now dust has settled.

For one it's telling they are showing multi-core performance, they're using a workstation orientated benchmark. This leads me to think they may not have the clockspeeds we're after, maybe close but just not enough, so they target the multi thread workload market, lower cost multi core option than Intel. This naturally extends to servers if they have TDP in check.

Canned benchmarks sure have their place. But for someone like me who wants to know how well it does encoding video, in a real scenario when I'm working with it, real world tests are the one.
 
This just seems like a complete and utter bullshit benchmark, here's a listing of results: CPU benchmarks | BlenchMark
According to this particular benchmark the 6 core 3.47GHz Intel X5690 from five and a half years ago is 20% faster than a 10 core 3GHz Haswell-E, and the fastest AMD is a 6 core K10 from 2009 (Opteron 8425), which, if you go solely by this benchmark that so many here are claiming vindicates AMD, is over 60% faster than an 8 core AMD FX-8350. Does any AMD fan really want to stand up and claim the Opteron 8425 is 60% faster than the FX-8350 or can we all just agree this is a very strange benchmark that obviously cares more about some odd features of certain CPUs and isn't exactly indicative of any sort of real world performance, because even if Zen can stand neck and neck with Haswell-E on this particular benchmark the 5.5 year old X5690 is still beating both of them soundly, and they're only about 30% faster than a 7 year old AMD part.

What we really need is a wide range of benchmarks testing lots of different features and usage scenarios next to this blender benchmark, then we need to throw blender out because it's stupid and go by everything but blender.
 
We need zero benchmarks is the point. All we need run is some real world applications doing real tasks. (Watch someone name another benchmark and explain how it translates to real world performance.)
 
-Dragon- said:
This just seems like a complete and utter bullshit benchmark, here's a listing of results: CPU benchmarks | BlenchMark
According to this particular benchmark the 6 core 3.47GHz Intel X5690 from five and a half years ago is 20% faster than a 10 core 3GHz Haswell-E, and the fastest AMD is a 6 core K10 from 2009 (Opteron 8425), which, if you go solely by this benchmark that so many here are claiming vindicates AMD, is over 60% faster than an 8 core AMD FX-8350. Does any AMD fan really want to stand up and claim the Opteron 8425 is 60% faster than the FX-8350 or can we all just agree this is a very strange benchmark that obviously cares more about some odd features of certain CPUs and isn't exactly indicative of any sort of real world performance, because even if Zen can stand neck and neck with Haswell-E on this particular benchmark the 5.5 year old X5690 is still beating both of them soundly, and they're only about 30% faster than a 7 year old AMD part.

What we really need is a wide range of benchmarks testing lots of different features and usage scenarios next to this blender benchmark, then we need to throw blender out because it's stupid and go by everything but blender.
Click to expand...

Chances are those chips at the top are running multi socket configurations if that's the case then yes an Opteron 8425 could very well beat an 8350 if it's running multi-socketed. That model I believe runs in sockets 4 and up. Same holds for those Xeons at the top. Those are 2 socket processors.
 
Last edited:
You must log in or register to reply here.
Back
Top