AMD Bulldozer Thread

[M.Beier]

Well when the first generation Pentium 4 came out, it was very expensive and clock for clock was slower against Pentium III, but after a few revisions it stood up to the top model Athlon XP 3200+. Then later Nehalem was released, inheriting Hyperthreading which was developed thanks to Pentium 4.

There is no reason why tweaking cannot get the Bulldozer on track, remember it has excellent multithreading performance, just the IPC that needs to be tweaked.

Also AMD engineers are speaking out that they are underfunded by the new management which is why they don't have the resources to create a good chip.

But highest frequency PIII was 1266mhz when when 2ghz P4 was released.

[ice_chill]

But highest frequency PIII was 1266mhz when when 2ghz P4 was released.

Pentium 3 reached 1400Mhz, first Pentium 4 came out as 1600Mhz. Even then the PIII was faster even with the slower clock.

[rozzyroz]

Willamette came out at 1.3ghz i believe. and Tualatin did hit 1.4 in the end. willies started doing better at 1.8ish, but still lagged behind.

[M.Beier]

Pentium 3 reached 1400Mhz, first Pentium 4 came out as 1600Mhz. Even then the PIII was faster even with the slower clock.

But look at the launch dates, some PIII came after P4 launch

[Hornet331]

It took them 5 monts to realease a P4 that was faster then any of the P3s (1.7ghz Willamette was the first one that was able to outperform a P3 for good) the inital release of the 1.4/1.5 ghz part where awefull... god damit there even was a 1.3ghz part that sucked so much the quitely removed it form there table...

[M.Beier]

http://en.wikipedia.org/wiki/List_of...icroprocessors
http://en.wikipedia.org/wiki/List_of...icroprocessors

[bro20000]

Canceled my order for bulldozer! Sticking with sandy till ivybridge now!

[haylui]

Canceled my order for bulldozer! Sticking with sandy till ivybridge now!

yay!! and never turn back.

[Lanek]

Put 16 cores 16 thread in the equation ...
and you get your response vs Xeon ... Sadly AMD look to have made this architecture only for server and workstation ,,,,

Piledriver is made for increase IPC and performance per core .... now we know what have AMD in mind.. look to Bulldozer as a baseline of the architecture, perf will come next.


Lol, Windows 7 is not even capable of take parts of this architecture yet... it can't recognise the modules as dispatcher for cores/ threads yet .

[AliG]

.....if i have amd systems already built and i can just drop bd in, then no it does not make more sense.

lol.... i mean that's on you. I pointed out performance, and you have twice made excuses for why it would make more sense to get BD over sandy bridge. Maybe in your situation specifically, but as a whole no it really doesn't at all. Especially considering your Thuban is quite a bit faster in most tests not named video rendering

[kromosto]

when you read more reviews it becomes more funny

[SKYMTL]

Legion Hardware and Guru3D reviewed FX-6100

AMD FX 8150 - 8120 - 6100 and 4100 performance review
http://www.guru3d.com/article/amd-fx...rmance-review/

From what I see, all Guru did was use a FX-8150 processor to "replicate" the results of the lower end chips rather than actually having additional samples on-hand.

[Piotrsama]

There happens a lot of discussin about that, many say on of the reasons is that happend because of the heavy use of automated desgin tools and next to no hand optimization happend. If you look at what we got and when you look at what Cliff Maier said back in 2010:

And when you look back, how furious AMD fans where at this, calling him a disgruntled ex-employee who only wants to put amd into a bad light...

That's quite interesting.

[Gogeta]

... god damit there even was a 1.3ghz part that sucked so much the quitely removed it form there table...

That 1.3GHz Willy was in the box that got me addicted to overclocking and hardware. I had nowhere to go but up! Just thank someone we're no longer using RDRAM.

[RussC]

Oh man, the Hitler video was really funny and disturbing at he same time.

Its clear to me that AMD just decided scerw it, we just have to get this processor out, regardless of performace, wattage or otherwise. Release it and let the cards fall were they may. Its painfully obvious that AMD(and bigger companies also, like IBM, Samsung, TI ect) will not catch up to INTC anytime soon soon when it comes to manufacturing. Thats what a monopoly can do for your business, just like MSFT(these two tech giant monopolys are being allowed to stand, why, I dont know). Or at least till we get to the point were the node shrinks stop somewhere ~9nM. Whent the node shrinks slow to a crawl or stop, that will allow the other players to catch up.

AMD clearly is the loser in all of this. Those wattage #s are appalling. Thats trully is the worst thing about this whole mess. Its more than clear their engineering expertise is way behind INTCs, in all facets across the board. Its really suprising to me the stock hasn't acted to the downside more, but I guess its already so low, there just not more downside in it.

As Ive said before, my criteria for success or failure are the adds in PC World from IBUYPOWER and CYBERPOWER. If the FX line could make those advertisements, its a success, if not, then fail. Im really looking forward to the next PC World in the mail. My guess is it wont be there.

I was going to upgrade to one just to run on my chiller, but sadly I will be with my 955 for a while longer.

RussC

[ice_chill]

From what I see, all Guru did was use a FX-8150 processor to "replicate" the results of the lower end chips rather than actually having additional samples on-hand.

So the Athlon II X4 640 remains the best budget quad core.

[FlawleZ]

That 1.3GHz Willy was in the box that got me addicted to overclocking and hardware. I had nowhere to go but up! Just thank someone we're no longer using RDRAM.

I had to use one as a workstation several years back. I hated it.

[DedEmbryonicCe1]

Pentium 3 reached 1400Mhz, first Pentium 4 came out as 1600Mhz. Even then the PIII was faster even with the slower clock.

Willamette came out at 1.3ghz i believe. and Tualatin did hit 1.4 in the end. willies started doing better at 1.8ish, but still lagged behind.

Both of you are wrong. Look at M.Beier's links and you'll see P4 Williamette debuted with only two speeds; 1.4 & 1.5 GHz. I know because I had a P4 1.4 GHz Williamette with RDRAM PC-800 that was purchased when those were the only choices available (SDRAM equipped P4 systems hadn't arrived yet). Yes, I learned my lesson from that mistake.

[FlanK3r]

lol.... i mean that's on you. I pointed out performance, and you have twice made excuses for why it would make more sense to get BD over sandy bridge. Maybe in your situation specifically, but as a whole no it really doesn't at all. Especially considering your Thuban is quite a bit faster in most tests not named video rendering

I dont agree, maybe Im alone, who really read most of reviews from all world (to time about 25 reviews). If u use superpi and wprime, yes, u can stay with Thuban. But in real aplications is FX better than Thuban-Photoshop, rendering, compression/decompression, videoencoding, microsoft office, in gaming simillary (in BF3 is FX better!). So, what do u need to proof? No, FX is really better than Thuban. U must more write or tetsing with your hands, no listen some people here. Without experinece is not knowledge ,-).

[Sparky]

So you think the power consumption numbers are higher because people used the stock coolers and more voltage than they could have? How much power usage would you expect to shave off with the ideal scenario?

You can be surprised how much power consumption can be reduced by what seems a relatively minor reduction in voltage. I'm not saying that will fix Bulldozer's power consumption issues, I just mean it in a general statement.

[SUPERKAMES]

@FlanK3r:
Agree with you .
//I'm using an AMD X6 1090T & still looking for a FX-8120 . Look at this article --> Some real world apps such as 3dsmax renderring show me that a FX8150 is 8% slower than an I7 2600K with a decent headroom for overclocking , not too bad .

[slaveondope]

You can be surprised how much power consumption can be reduced by what seems a relatively minor reduction in voltage. I'm not saying that will fix Bulldozer's power consumption issues, I just mean it in a general statement.

Much like Llano at stock

[kaktus1907]

any boinc numbers ?

[villa1n]

Put 16 cores 16 thread in the equation ...
and you get your response vs Xeon ... Sadly AMD look to have made this architecture only for server and workstation ,,,,

Piledriver is made for increase IPC and performance per core .... now we know what have AMD in mind.. look to Bulldozer as a baseline of the architecture, perf will come next.


Lol, Windows 7 is not even capable of take parts of this architecture yet... it can't recognise the modules as dispatcher for cores/ threads yet .

are there any interlagos 6282 se reviews up yet? specint vmmark etc? was movieman going to drop his numbers soon, as it would be interesting to see how it compares to an e7 10c

[rozzyroz]

Both of you are wrong. Look at M.Beier's links and you'll see P4 Williamette debuted with only two speeds; 1.4 & 1.5 GHz. I know because I had a P4 1.4 GHz Williamette with RDRAM PC-800 that was purchased when those were the only choices available (SDRAM equipped P4 systems hadn't arrived yet). Yes, I learned my lesson from that mistake.

i remember when there were many people on madonion.com that were bragging about getting these 1.3ghz parts, only to be shocked at how the lower end p3 would outperform them.... it was much like the lead up to this launch. it really made you shake your head wondering what happened. however when they reached 1.8ghz, i bought a p4/rdram system as well... it replaced my coppermine system with 550e running at 770mhz. the 1.3's were dropped very fast.

here's alink for ya. http://ark.intel.com/products/27421/...he-400-MHz-FSB