AMD FX-8150 shows up in Passmark CPU benchmark

[grimpr]

Still sounds decent to me. I might get one. Looking at this motherboard. Looks pretty sick to me. Imagine that with an 8-core BullDozer, and four GTX460s, all crunching Distributed Computing.

Edit: I think that AMD is still ahead of Intel, not in pure computing horsepower, but as a platform company.

AMD's newest chipsets sport ALL SATA 6G ports, not this pathetic mish-mash of SATA 6G and SATA 3G like Intel. I think (could be wrong), that AMD's newest chipsets also support USB3 natively. Also, AMD's higher-end chipsets like the 890FX and 990FX, have mucho PCI-E lanes. That's one thing that you cannot get with Sandy Bridge, and that's a motherboard with four PCI-E x16 slots (that are at least PCI-E 2.0 x8 electrically). SB's on-chip PCI-E lanes simply aren't numerous enough, and even if they were, they cannot be split four ways, AFAIK.

If you're gonna spend all that cash on a AM3+ mobo, just get an ASUS. Gigabyte, for me, are going downhill for some time now.

 

[jones377]

Intel should just release a 2500K with HT enabled then. Game over for AMD. Call it 2550K

Either that or the rumored 2700K could waterfall down the prices for the 2500K and 2600K, perhaps discontinuing the 2500K altogether. Looks like neither will happen though.

 

[Abwx]

Intel should just release a 2500K with HT enabled then. Game over for AMD. Call it 2550K

Is there people who think that a 2500K with HT would be different
than a 2600K???...

 

[3DVagabond]

Intel should just release a 2500K with HT enabled then. Game over for AMD. Call it 2550K

Either that or the rumored 2700K could waterfall down the prices for the 2500K and 2600K, perhaps discontinuing the 2500K altogether. Looks like neither will happen though.

As Abwx said, they already have a 2500K w/HT.

No good reason to discontinue it because it's such a good gaming chip. Besides it helps keep the 2600K price up a bit. If they discontinued it then the 2600K would have to fill both slots. Probably at a price somewhere in between the price they can sell it for now and the price of the 2500K. That or just give up that market price point for an O/C'ing gamer's chip.

Intel has to be a bit careful about how aggressive they get. They don't want to raise the ire of the FTC. If Intel thought they could get away with it, they would have simply undercut AMD out of business by now.

 

[Dresdenboy]

Yet , Intel CPUs do better in Superpi , wich is an evidence
that they still have a robust X87 support....

SPi uses some long latency (12+) x87 ops for doing tests/comparisons at the end of loops. Might be a real bottleneck when having a separate FPU (results have to be finally evaluated in the int core). Further such loops might be difficult to predict.

 

[BlueBlazer]

SPi uses some long latency (12+) x87 ops for doing tests/comparisons at the end of loops. Might be a real bottleneck when having a separate FPU (results have to be finally evaluated in the int core). Further such loops might be difficult to predict.

Furthermore x87 FP instructions are slower on Pentium4 and all generations after that. That can be clearly seen during the K7 versus P4 days (in fact from Athlon up to Phenom had better x87 performance than Pentium4 and up to Core 2 Duo), that hasn't changed since Intel endorsed SSE and SSE2 instructions as replacement. Abwx assumptions (on SuperPi doing better on Intel due to still having "a robust X87 support") are in fact highly incorrect, as he is contradicting himself on earlier statements regarding x87 performance on Phenom versus Core 2 Duo/Quad processors in that SSE2 debate.

 

[nemesismk2]

Passmark is about as reliable as WEI. All we can really do is bear down until Anand/etc get the real tests out.

my thoughts exactly, anand will give a reliable review

 

[nonameo]

I love it when people quote passmark on slickdeals. It's pretty much the holy grail of benchmarks over there.

 

[Despoiler]

Either that or the rumored 2700K could waterfall down the prices for the 2500K and 2600K, perhaps discontinuing the 2500K altogether. Looks like neither will happen though.

The 2700k is replacing the 2600k and will be more expensive than the 2600k.

 

[mosox]

Can someone explain me why the benches in the passmark are not good while the 4-5 year old benches in Anand are better?

For instance there are no less than 4 (four) newer versions of the 3dsmax after the 3ds Max 9 featured in the Anand benches.

How about Cinebench? Take a look, see the difference between the R10 used by Anandtech and the latest version, R11.5 or it's not the same benchmark? If it's not the same what's more important?

http://www.hardwarecanucks.com/foru...2500k-core-i7-2600k-processors-review-10.html

Also, Sony Vegas 8?

Also Anand: POV-Ray 3.7 beta 23 ----2600K is 25% faster than the 1100T

http://www.anandtech.com/bench/Product/287?vs=203

HWCanucks POV-Ray 3.73 beta 40

http://www.hardwarecanucks.com/foru...2500k-core-i7-2600k-processors-review-11.html

2600K is 9% faster than the 1100T

 

[Arkaign]

Can someone explain me why the benches in the passmark are not good while the 4-5 year old benches in Anand are better?

For instance there are no less than 4 (four) newer versions of the 3dsmax after the 3ds Max 9 featured in the Anand benches.

How about Cinebench? Take a look, see the difference between the R10 used by Anandtech and the latest version, R11.5 or it's not the same benchmark? If it's not the same what's more important?

http://www.hardwarecanucks.com/foru...2500k-core-i7-2600k-processors-review-10.html

Also, Sony Vegas 8?

It's not so much that the Passmark benches themselves are all that bad (they do leave out things like gaming, however), it's that the control set is basically out the window due to letting everyone post benches. So the contents are tainted enough to be basically worthless. Nobody in the enthusiast/professional PC community takes PassMark seriously. They even have disclaimers on their site related to this.

Anand's testing choices are a seperate issue, but overall I think his tests are extraordinarily good. With new major releases there are dozens and dozens of aspects looked at, and nothing major is left out. You can always fill in the data as you should by checking other reputable sites and comparing and contrasting results.

 

[mosox]

Edited my comment. I might spend some time one of these days to compare the benches from different sites, so far I'm seeing wild differences in many areas. <br>

 

[Arkaign]

Edited my comment. I might spend some time one of these days to compare the benches from different sites, so far I'm seeing wild differences in many areas. <br>

Yeah, I think it's definitely important to really do the research if you're spending hard-earned $$ on hardware.

The Cinebench 10 vs 11.5 is a good example of getting that info, as you posted. If one is focused on such activities, and has access to the latest software, then the 6 core chips really do pay off there.

Overclocked scores also relevant or irrelevant depending on how you want to set up your system.

 

[sequoia464]

I think (could be wrong), that AMD's newest chipsets also support USB3 natively.

Not yet, at least on my 990FX board - USB 3.0 is through an AS Media chip. Supposedly an improvement from the NEC USB3.0 - according to ASUS anyway.

 

[formulav8]

I thought they added two USB 3.0 ports in the latest chipsets? Or is that the Llano chipset only? :\

 

[Kingbee13]

I thought they added two USB 3.0 ports in the latest chipsets? Or is that the Llano chipset only? :\

Llano only, 990FX is just a rebadged 890FX (from what I understand)

 

[sequoia464]

I thought they added two USB 3.0 ports in the latest chipsets? Or is that the Llano chipset only? :\

All I can speak to are my own boards - my Gigabyte 890XA had an 850 southbridge - the two USB3 ports were provided by a NEC chip, my current board is an Asus 990FX Sabertooth with four USB3 ports through the AS Media chip.

I don't know anything about Llano boards and chipsets. I'm sure someone here knows.

EDIT; I see someone already answered.

 

[Abwx]

SPi uses some long latency (12+) x87 ops for doing tests/comparisons at the end of loops. Might be a real bottleneck when having a separate FPU (results have to be finally evaluated in the int core). Further such loops might be difficult to predict.

Latencies have nothing to do...

All arise from the algorithm used to calculate pi using
an asymptotical approximation , thus , contrary to some beliefs ,
superpi will compute more integer ops than floating points ones ,
hence the superiority of core duo over K8 , due to better integer
throughput..

This of course completely destroy Blueblazer s " analysis "
about the holdings and borderings of this "bench" s results
on differents CPUs and their respective strenthes in X87....

 

[Dresdenboy]

Latencies have nothing to do...

All arise from the algorithm used to calculate pi using
an asymptotical approximation , thus , contrary to some beliefs ,
superpi will compute more integer ops than floating points ones ,
hence the superiority of core duo over K8 , due to better integer
throughput..

Did I hear my T alert

Which integer throughput is higher? ALU ops should have a t.put of 3 per cycle in each case.

But what kind of integer ops did you see being executed during main loops in SuperPi? I'm not at home this WE but IIRC I saw most hits at fmul, fadd, fxch, fld, fild, fist (your int ops?), fcomp, fstcw, ftst and sahf.

 

[Abwx]

Did I hear my T alert

Which integer throughput is higher? ALU ops should have a t.put of 3 per cycle in each case.

But what kind of integer ops did you see being executed during main loops in SuperPi? I'm not at home this WE but IIRC I saw most hits at fmul, fadd, fxch, fld, fild, fist (your int ops?), fcomp, fstcw, ftst and sahf.

Super PI is a computer program that calculates pi to a specified number of digits after the decimal point—up to a maximum of 32 million. It uses Gauss–Legendre algorithm

http://en.wikipedia.org/wiki/Gauss–Legendre_algorithm

 

[sm625]

If I had to pick one benchmark it would be passmark. Definitely better than sysmark. It speaks volumes that anand uses half a dozen sysmarks but doesnt bother to list the passmark score. But that doesnt mean this passmark score is any more legit than any of the other BD benchmarks floating around. If I cant buy it or test it myself, it doesnt mean much.

 

[zlejedi]

AMD's newest chipsets sport ALL SATA 6G ports, not this pathetic mish-mash of SATA 6G and SATA 3G like Intel. I think (could be wrong), that AMD's newest chipsets also support USB3 natively. Also, AMD's higher-end chipsets like the 890FX and 990FX, have mucho PCI-E lanes. That's one thing that you cannot get with Sandy Bridge, and that's a motherboard with four PCI-E x16 slots (that are at least PCI-E 2.0 x8 electrically). SB's on-chip PCI-E lanes simply aren't numerous enough, and even if they were, they cannot be split four ways, AFAIK.

And all of those are utterly pointless to huge majority of users.

SATA2 will never bottleneck traditional HDDs
SATA3 is needed only for fast SSDs
PCI-ex lanes x8 vs x16 have been shown many times to have no noticeable impact on multi-GPU performance

 

[StrangerGuy]

And all of those are utterly pointless to huge majority of users.

SATA2 will never bottleneck traditional HDDs
SATA3 is needed only for fast SSDs
PCI-ex lanes x8 vs x16 have been shown many times to have no noticeable impact on multi-GPU performance

AMD knows in this age better chipsets = more CPUs sold than better CPUs themselves.

 

[podspi]

AMD knows in this age better chipsets = more CPUs sold than better CPUs themselves.

There are limits to that but yes, platform does matter. For many people, the CPU is no longer a significant (or as significant) proportion of the total cost of the machine.

 

[exar333]

AMD knows in this age better chipsets = more CPUs sold than better CPUs themselves.

You bet. Lot's of shiny stickers on your machine is >>>> than CPU performance. :thumbsdown: