- Wednesday, 25 November 2009
- INQ Mobile
- RSS
A hunch is creativity trying to tell you something - Frank Capra
How to launch Nehalem when Penryn is doing so well
WE'RE A MONTH AWAY now from the upcoming Fall Intel Developer Forum in the cool windy Frisco, a place where you might see quite a few of us running around gathering the news. A very important event for us and for you.
Especially this year, where Intel is pretty much expected to unveil its final Nehalem rollout plans for the initial 4Q '08 rounds: read - DP workstation, HPC and UP gaming parts.
As we've mentioned here before, the Bloomfield UP and Gainestown DP Nehalems, both using varieties of the Tylersburg chipset platform, are doing well, and seem to have gained a bit of speed over the past few months too - at least enough to have their launch speeds matching the best GHz launch clocks of the Penryns at the same time, while hopefully providing quite a bit more performance.
At the same time, Penryns have managed to keep AMD at bay when it comes to speed, except in big MP boxen and HPC supercomputing apps, where Opteron's memory bandwidth and Hypertransport still give it an edge till... yeah, till the Nehalems are in.
So, Intel seems to be in a kind of dilemma: start pre-marketing the Nehalems early to extend the mindshare domination over AMD, but at a risk of cannibalising the current healthy Penryn sales; or 'balance out' the rollout, focusing strictly on the high-priced extreme segments first to gain platform recognition without affecting Penryn revenues?
Well, it's not all so clear. If, say, limited "extreme" Bloomfield and Gainestown benchmarks were to surface, with results drastically above the best Penryns, even many mainstream buyers could be impressed enough to consider delaying their upgrades till the mainstream Nehalems are out and about in a year from now. Those sales right now have very little risk of going to AMD, so delaying them is a pure revenue loss to Intel.
Keep in mind also that the QuickPath interconnect and on-chip memory objectively give the least advantage on uni socket desktop machines, where the existing FSB1600 is far from being saturated. Also, in a typical X48 or 790i chipset scenario, the Crossfire or SLI GPUs have fast direct access to very high bandwidth DDR3 memory banks without going through chipset->CPU hops like they have to do on the current AMDs - or future Nehalems.
Then, we have the E0 Penryn stepping coming in, sometime this month probably. Expect performance and power gains over C0 similar to what the famous G0 stepping did for the 65nm generation. Either way, E0 should enable Intel to launch at least 3.6 GHz parts at about the same declared TDP and voltages as the current 3.2 GHz ones. Whether they decide to go ahead with that, or stick with the 3.2 GHz but possibly lower TDP, depends to a great deal on the decided Nehalem launch speed.
I personally feel Intel would have no problems having 3.2 GHz UP and DP Nehalems at launch in 4Q, and, if they really push it, 3.6 GHz limited Extreme parts aren't impossible either at the same time - as far as I've seen in early tests. So, Nehalems should have the launch speeds identical to the top Penryn speeds at the time.
My feeling is: go full steam ahead with the new processors and make things clear at the IDF - the promise was made to have them out this year, and it better be kept, at least for the workstation, HPC and gaming markets which are always the best early adopters.
The sales overlap problem with Penryn, and any resulting Osbourne effect possibility, can easily be removed by placing the initial Nehalems in a price class above the best Penryns, at least until early next year. By that time, Nehalem should be ready for the mainstream.
Does it mean cheaper high-speed Penryns? I hope so. More pressure on AMD CPU lines? Yeah, definitely. More opportunity for AMD GPU lads to bundle pairs of 4870 X2 cards with each high-end Penryn and Nehalem sold? Obviously. And, in absence of Nvidian Nehalem SLI chipset, more "value sales" of Nforce cum GeForce SLI bundles on Penryns? Sure - look at the GTX280 pricing drop for an early taste of it. µ
Share this:
ShareBusiness Intelligence Software
Customer Relationship Management (CRM)
Enterprise Accounting Software
Enterprise Resource Planning (ERP)
Enterprise Systems Management
I’m tryin to make me go to Pennryn
I said "No! No! Forget it! No!"
This looks very dangerous to me. 
I'm not picking up dog sh!t. I'm a rock star.
We're the Osbournes. And I love it.

What do you want me to do with my [gflops] Sharon? Put it under my bed?
Turn that thing off, its driving me mad!!
I need 'em. I need every one.

Listen to me, I'm dead serious!

I'm not from the Mickey Mouse Club-I'm from the Chipzilla Club
I'm the Prince of fncking Darkness! 
I'm the pre-announcement bogeyman that scares new marketing people!
Being sober on a bus is,
like, totally different than being drunk on a bus.

Sandy Bridge wasn't in vogue with what they wanted to do at this time... We have people out there on the internet and in the clubs, throwing good money after bad, that people are excited about. I want to know what the kids are into, because I don't trust the industry. Northstar was built like a tank!

Occupation mission creep. It could be 2011, 100 years, as long as it takes, to be victorious in an unwinnable dollar.
Razor Blades? Penny apiece. Some cutting up of Santa Claus & Nahalem will be Final. its 700 million transistors. Basicly finalization of XP Core/NT5. (NOT Ultimate, NOT Extreme)

Big Steps are 6 memory slots with TOP(not all) being 32 Gb/s bandwidth. in 3d '6 thats new TOP of 32,000. Ddr3 is clunky ~1.3Gb/s

Yet it is merely end of long line of CPU improvements from K7, while real Game has Shifted to Dunnington, 
if You Got Cash, at 2 billion transistors. Bit high for Poor to report upon, why else is travel, often by internet fantasy, so important. Getting FREE Dunnington costs up to $2,700. or its Wholesle equivalent. So Go Figure & Remeber:Fall Isn't Today & Next year is mightly LONG Flat Spot. ALL TIME VISTA ULTIMATE SALES LIE DORMANT.
Signed:Ultie_Gezzer.
Great article, shame it seems to have attracted comments from the gibberish/markov chain brigade. Is there no way to get rid of drashek? He isn't clever or funny, hell, he's barely coherent.

I, for one, welcome our new Intel overlords, and hope they get the parts on shelves this year, forcing the price of Penryns down. It's a shame that AMD aren't snapping at their heels right now, or it wouldn't be an issue..
Well, for a start, I'm not going to rant rubbish like the posts above.

Intels main cncern is selling up curent Penryn stock, so in my opinion, the should start selling the Nehalem a an Opteronj competitior, and try to infiltrate the high end server market. I mean, finally they have a product with the memory bandwidth to compete with Opteron - and the computational horse-power also.

It wint effect Penryn slaes, as Penryn is almost mid-range pricing, even for high end parts.

I dont believe they have anything to worry about -especially on the desktop front, where Penryn is still the best performer.

And as the high end parts will be priced out of most [even enthusiast] people's pocket, they need to get their strategy right, and try make an impact in a market where change does not happen unless necessary [high end server etc].

The CPU looks great, looks like a good performer - but better than an overclocked Penryn?? We have to wait and see
Firefox's anti-Drashek addon would filter the headache-inducing post above.
I would dispute the claim that uniprocessing Penryn's are not yet saturating the bus. Unless that comment was meant to be restricted to native dual cores, my benchmarking of our memory-write-intensive neural network software has indicated very poor scaling from 2 cores to quad cores.
What the fuck are these two last comments about?!
This Intel_Ultie guy is clearly channelling Drashek. God help us. Oh right, processors. Monopoly? Stupid code-names? Can't wait until next year for the latest Intel masterpiece, codenamed "J. K. Willington Fortescue-Smythe" with an even more ridiculously large cache. Nothing against Intel, it's just boring when the big guy is winning all the time. It's exactly what you'd expect and isn't really that impressive. On the other hand it does show just how much better foreign engineers are than home-grown ones these days.
Is there any ACTUAL data on the Osbourne effect? Every generation you hear "they have to be careful, you know the whole Osbourne effect, blah, blah, blah"

Intel will follow their recent gameplan on product rollouts. As Nehalem is bigger (and probably lower yielding) than an MCM Penryn, expect the focus to be almost exclusively on server land (where the gains will also be most noticeable) with a handful of parts, high prices and shortages on desktop. 

Unlike AMD, I think Intel understands the early adopters usually will pay through the nose and sell a few pints of blood for performance, so launching energy efficient, non high end chips 'because customers are demanding it' (stifles laughter thinking that AMD actually believed folks would not see right through that crap), will not be happening.

At some point in 09, the extreme Penryn parts will disappear (if Nehalem hasn't surpassed it significantly on clockspeeds) and Penryn will become the blunt 'low end' quad instrument for Intel to see if they can push Phenoms into the 2 digit price category.

You will also not see 3500 SKU's of Nehalem quads! Would you like to try our 3.115GHz or the 3.120GHz product? Would you like that to be the black, blue or black and blue edition? How about a 1.8GHz chip?
Easy to make up lots of nifty techy product/concept names: QuickPath, QuickRing, QuickBus, QuickBand, QuickWire, QuickSwitch, QuickBox, QuickCard, QuickTech, QuickSpew ...
WTF are you saying? I've seen your comments on other sites as well. The similarities are uncanny. Ever heard of that saying...how did it go again...? Oh yeah.

"English, motha------. You speak it?"
Umm, "Hector", <anything>_Ultie was that "drashek" bellend with his tedious Nonsense Rambling With Odd CAPITALISATION, last I checked :-)

Anyway, interesting article, surprisingly thoughtful and not at all spittle-flecked, unlike some of the stuff around here (ahem, Charlie, please drink less coffee, dude). La Intella is going to have a lot of Penryn units to shift before they can carpet bomb with the new stuff, but I am sure they'll find a way to squeeze a ton of profit out of the situation, even if it means worse chips for us poor proles. That said, the new parts sound mighty fine, can't wait to take some for a spin.
Intel_Ultie is a convicted pedophile aka Drashek, with maggots instead of brains inside the cranium.

I don't believe it! I just arrived for my daily feeding frenzy of Inq. and get to these comments... WTF?

Are people trying to emulate that looney 'Drashek' OR is the looney 'Drashek' still fouling up my entertainment?

Being a programmer I thought about trying to create a text/word generator with some kind of coherance similar to that of what we see with Senor Ultie. I came to the conclusion that producing the functional requirement specification for something so dis-jointed, so 'all-over-the-place' was going to be impossible. Imagine trying to create a 'Drashek' algorithm using the laws of chaos? As for people trying to copy this freak I would very much like to understand their method of analysis. Please help me.....

Failing the above; I have looked into ways of tracking down this &*%£ (words fail me - wish it was the same for Ultie) and breaking down the door to his hovel/asylum entering the room where his PC is running, pulling out the plug and ramming it into a very dark place!!!!!

OK, I've calmed down now....I think.

Inq. please, please, very pretty please with nice berries on the top, stop this thing posting anymore comments? It's putting me off my coffee and leading to starvation at lunchtime due to lack of appetite.

Many thanks and hopes in advance,

Dave The (Ultie) Rave xxx