AMD accuses BAPCo and Intel of cheating with Sysmark benchmarks

[Tomsen]

Because the FX 8350 has only 2 ALUs for every 4 per Haswell core, and because it lacks newer instructions. Software should be keeping up with new instructions. AMD can get bent or prove it. Also, games are GPU-bound. That's not a valid way to test CPU power.

But haswell is a SMT design. The 4 ALUs are shared between 2 threads.

A single thread most likely peak at 3 ALUs, anyway.

 

as for who's screwing who ..

do you recall Intel's Larrabee? nVidia fckd over Intel by retracting their licences, killin Larrabee in 2010

what did Intel do? complain? made YT videos about it? no, they started working on co-processors; they recently bought Altera and with that purchase they will put the screws to nVidia - not by "cheating" in benchmarks

You might want to look into history, on why things acted out as they did.

Nvidia didn't suddenly f*ck intel for no reason. It was basically retaliation, because Intel tried to kill off Nvidias chipset business (again, with dirty tricks).

They went into a lawsuit. Intel didn't come out as the winning party..

EDIT: Larrabee was also DOA. They went ahead an turned it into a dedicated co-processors, where it actually had a chance.

[patrickjp93]

People questioning AMD's damning, demonstrated and data-backed arguments... what are you doing? Is your objectivity switched not flipped in the 'On' position?

You have no clue. AMD is cherry picking specific tasks that use instructions which for both companies are at their theoretical lowest possible latencies, leaving only cache, memory, and clock speed differences for the most part. AMD is going to have to do much better to make me start believing this.

[Delicieuxz]

You have no clue.

I think that line rather applies to people who react to a statement along with backing information with assumptions that one party is making a mistake.

 

You aren't supposed to believe one side or the other based on them making claims. If anything, consider the information and see what it adds up to.

[Demonking]

When did I say Zen doesn't match Haswell? We're expecting it to be somewhere around 40% faster than Excavator, putting Zen on par with or even higher than Haswell.

 

Look at Skylake. That uses 14nm. The improvement from Haswell is about 10-15% faster, that's an minimal improvement at best. So you telling me that just because AMD's Zen is 14nm and on par with Haswell, that it's a joke? Laughable.

 

I don't know how Cannonlake will perform but I'm predicting it's another 10-15% improvement over Kaby Lake, judging from Intel's trends.

 

The fact you people expect that AMD will suddenly catch up to Intel is a joke in itself, actually be realistic for once. The only thing Zen is hoping to achieve is for AMD to be relevant again and making some competition.

point is for people on haswell there's no reason to move over to zen & even for new builders they're better off getting a haswell CPU which may end up cheaper especially with their motherboard being so cheap along with DDR3 being cheaper than 4 and will only get cheaper.

[HKZeroFive]

point is for people on haswell there's no reason to move over to zen & even for new builders they're better off getting a haswell CPU which may end up cheaper especially with their motherboard being so cheap.

 

Competitive pricing is going be a large factor. If it costs somewhere around Haswell but not as much as Skylake, I think many people will be appealed by the idea of having DDR4.

 

Also remember that Zen will go up to 8 cores. If they can competitively price their 6 cores, we might see more people flock to AMD.

[Demonking]

Competitive pricing is going be a large factor. If it costs somewhere around Haswell but not as much as Skylake, I think many people will be appealed by the idea of having DDR4.

 

Also remember that Zen will go up to 8 cores. If they can competitively price their 6 cores, we might see more people flock to AMD.

See that what has me worried. How are they going to get an 8 core haswell level IPC for less than an i7,  those prices are going to be higher same with the suppose 6 core. 

[HKZeroFive]

See that what has me worried. How are they going to get an 8 core haswell level IPC for less than an i7,  those prices are going to be higher same with the suppose 6 core. 

 

True. This is all speculation so I don't know what specific price AMD is aiming for, but an alternative to the i7s would appeal to quite a lot of people (again, if priced competitively). If AMD is willing to adjust their prices to rival some of the i7s, we just don't know. For now, Intel still has the dominance in the high end range.

[Sakkura]

point is for people on haswell there's no reason to move over to zen & even for new builders they're better off getting a haswell CPU which may end up cheaper especially with their motherboard being so cheap along with DDR3 being cheaper than 4 and will only get cheaper.

 

A 2016 CPU was never going to get people to ditch a 2013 Intel CPU (yes, I know many were bought in 2014, 15 or even 16). It's about replacing Sandy/Ivy Bridge, anything older from Intel, and anything pre-Zen from AMD. Haswell will most likely be poor value for money by the time Zen arrives, even if Zen itself sucks - Skylake supply issues should clear up eventually.

[Marinatall_Ironside]

A 2016 CPU was never going to get people to ditch a 2013 Intel CPU (yes, I know many were bought in 2014, 15 or even 16). It's about replacing Sandy/Ivy Bridge, anything older from Intel, and anything pre-Zen from AMD. Haswell will most likely be poor value for money by the time Zen arrives, even if Zen itself sucks - Skylake supply issues should clear up eventually.

which hopefully when summer hits, I can have myself a 6700K at the $350 USD mark

[patrickjp93]

I think that line rather applies to people who react to a statement along with backing information with assumptions that one party is making a mistake.

You aren't supposed to believe one side or the other based on them making claims. If anything, consider the information and see what it adds up to.

It doesn't add up to what AMD is claiming it does. If you're falling for this, you're woefully ignorant.

[Demonking]

A 2016 CPU was never going to get people to ditch a 2013 Intel CPU (yes, I know many were bought in 2014, 15 or even 16). It's about replacing Sandy/Ivy Bridge, anything older from Intel, and anything pre-Zen from AMD. Haswell will most likely be poor value for money by the time Zen arrives, even if Zen itself sucks - Skylake supply issues should clear up eventually.

I find that hard to believe since people say skylake isn't worth the move from a 2500K yet they are going to move to zen which is worst than skylake which they say isn't worth it. 

[Benji_w]

I find that hard to believe since people say skylake isn't worth the move from a 2500K yet they are going to move to zen which is worst than skylake which they say isn't worth it. 

 

Who said that? lol Zen isn't out yet so how can it be worse?

[Benji_w]

It doesn't add up to what AMD is claiming it does. If you're falling for this, you're woefully ignorant.

 

Intel have been done before for falsifying benchmark results, People in the FX club on OCN have recently received payouts over the class action or what ever it's called.

[Benji_w]

http://www.extremetech.com/computing/193480-intel-finally-agrees-to-pay-15-to-pentium-4-owners-over-amd-athlon-benchmarking-shenanigans

[Demonking]

Who said that? lol Zen isn't out yet so how can it be worse?

it stated to be on haswell level in IPC how is that not worst than skylake which is faster than haswell, so yes it will not be better than skylake.

[Benji_w]

it stated to be on haswell level in IPC how is that not worst than skylake which is faster than haswell, so yes it will not be better than skylake.

 

By who? AMD or other? Based on what? Speculation isn't fact.

[SurvivorNVL]

By who? AMD or other? Based on what? Speculation isn't fact.

@patrickjp93 did the math in regards to Zen's theoretical IPC performance based on the information they gave us.  It's 40% over Excavator which puts us in the ballpark, clock for clock around 5% below or above Haswell in IPC performance - that's speculation, but it's the most likely result.  So, with their own numbers and the knowledge of where it'd put them IPC wise - they're going to be at Haswell / Ivy-bridge levels of IPC when Kaby-lake and Skylake-E are released.  Zen is going to need to be in some -really- good pricing brackets as performance will definitely not be the winner in this case.

[Morgan MLGman]

Zen is already a disappointment. Its stated to be on Haswell level not even skylake so it's coming in with an L. 

 

If Zen has Haswell performance it will only be relevant if it's significantly cheaper than Skylake. 

You are mistaken, Zen was stated to have IPC of around Haswell level or a little above that with 6 and 8 cored mainstream CPUs, not to be around Haswell level of performance. If they're priced aggressively, they would not be a disappointment but a valid alternative to what Intel has/will have to offer, I'm not sure if those statements are true, but that's what has been revealed so far.

[Benji_w]

@patrickjp93 did the math in regards to Zen's theoretical IPC performance based on the information they gave us.  It's 40% over Excavator which puts us in the ballpark, clock for clock around 5% below or above Haswell in IPC performance - that's speculation, but it's the most likely result.  So, with their own numbers and the knowledge of where it'd put them IPC wise - they're going to be at Haswell / Ivy-bridge levels of IPC when Kaby-lake and Skylake-E are released.  Zen is going to need to be in some -really- good pricing brackets as performance will definitely not be the winner in this case.

 

Still could be speculation on their end too, fuck knows, there's no point talking about it til they're released.

[SurvivorNVL]

Still could be speculation on their end too, fuck knows, there's no point talking about it til they're released.

I really hope it's not speculation on AMD's end.  If they're just speculating about their own architecture and CPU's performance and don't even know if it is going to get close to Haswell performance and IPC - we should be worried.

[Benji_w]

I really hope it's not speculation on AMD's end.  If they're just speculating about their own architecture and CPU's performance and don't even know if it is going to get close to Haswell performance and IPC - we should be worried.

 

Worrying over the performance of a CPU is ridiculous, there are far more important things to worry about.

[PerfectTemplar]

You are mistaken, Zen was stated to have IPC of around Haswell level or a little above that with 6 and 8 cored mainstream CPUs, not to be around Haswell level of performance. If they're priced aggressively, they would not be a disappointment but a valid alternative to what Intel has/will have to offer, I'm not sure if those statements are true, but that's what has been revealed so far.

Well, I am not mistaken as I was just responding to the scenario where Zen was only just as good as Haswell in performance. I do doubt that would ever be the case, otherwise I don't know why AMD would bother. If they release IPC levels of Haswell with 6-8 cores and they are priced competitively then it will obviously be a great addition to the market and will make a lot of people consider AMD over Intel. 

[Sakkura]

I find that hard to believe since people say skylake isn't worth the move from a 2500K yet they are going to move to zen which is worst than skylake which they say isn't worth it. 

 

 

@patrickjp93 did the math in regards to Zen's theoretical IPC performance based on the information they gave us.  It's 40% over Excavator which puts us in the ballpark, clock for clock around 5% below or above Haswell in IPC performance - that's speculation, but it's the most likely result.  So, with their own numbers and the knowledge of where it'd put them IPC wise - they're going to be at Haswell / Ivy-bridge levels of IPC when Kaby-lake and Skylake-E are released.  Zen is going to need to be in some -really- good pricing brackets as performance will definitely not be the winner in this case.

 

AMD has only given some approximate numbers about IPC. We don't know what clocks Zen will run at yet. The problem with Intel architectures after Sandy Bridge is that they've struggled to reach the same (over)clocks, which means the improvement in IPC has been partially counteracted by lower clocks.

 

In addition, people saying they'll move to Zen but not Skylake also has to do with pricing, as well as perceived performance and maybe even brand preference. If a Zen quad core is significantly cheaper than the currently $250+ Core i5-6600K, well then it could be a more attractive upgrade even if performance is slightly lower.

 

Bottom line, we don't know yet. We don't know the price, and we only have vague ideas about the performance.

[patrickjp93]

Intel have been done before for falsifying benchmark results, People in the FX club on OCN have recently received payouts over the class action or what ever it's called.

No they haven't. Cinebench was a matter of Maxon using a compiler that didn't optimize for AMD's architecture. That said, Intel's compiler still doesn't optimize for anyone but Intel, and you don't see a fresh lawsuit over that. Maxon screwed that up, not Intel. The judge was incompetent. As for Pentium 4, that was under Otellini more than a decade ago. He and Richard Huddy are both gone. Intel is innocent until proven guilty and AMD is full of crap until it puts its money where it's mouth is or provides proof I can't debunk in my sleep.

[patrickjp93]

AMD has only given some approximate numbers about IPC. We don't know what clocks Zen will run at yet. The problem with Intel architectures after Sandy Bridge is that they've struggled to reach the same (over)clocks, which means the improvement in IPC has been partially counteracted by lower clocks.

In addition, people saying they'll move to Zen but not Skylake also has to do with pricing, as well as perceived performance and maybe even brand preference. If a Zen quad core is significantly cheaper than the currently $250+ Core i5-6600K, well then it could be a more attractive upgrade even if performance is slightly lower.

Bottom line, we don't know yet. We don't know the price, and we only have vague ideas about the performance.

We do know, and 14nm FF is not going to maintain the same clocks that 32nm Bulk did. At best they'll be in the same range. I predict low-mid 3GHz for the 8-cores.

We have a very solid idea about performance. It will be 3% over Haswell at the absolute best per-clock.