OH god can we stop with global warming and get backto i7?
Like for example, what d'you think of that "confirmed" release date Francois? =D
OH god can we stop with global warming and get backto i7?
Like for example, what d'you think of that "confirmed" release date Francois? =D
yeap, i can confirm, the release date is the day of the release of Core i7Originally Posted by Slovnaft
I forgot ... Core i7 will be able to run Dos 3.2 , even Dr Dos ...lol Legacy!
try to run Moto Racer on your graphics card ... I just tried ... does not work! no legacy there!
if you go there http://www.old-pc-games.com/ , you can pick up one of the game and try! As soon as Direct X 3, 5 or 6 are involved, you are screwed!
What I call a Royal mess!
Last edited by Drwho?; 10-19-2008 at 08:54 AM.
DrWho, The last of the time lords, setting up the Clock.
Talking about legacy, sure you would like legacy when the other players must pay for the use of intellectual property rights,
with donīt we seek legacy in a free open code? offff course there is no greenies in there right?
I would love some legacy, some free open source legacy, not the monopoly waanabe were intel is trying to go.
There are more company producing industrially x86 than there is GPU vendors ... what are you talking about?
I am not interested in talking about licenses, I don't know anything about it, and I don t want to know either ...
What I know is that the world of 3D games is full of whoop ass cans, try to install any games that you bought 5 years ago ... it does not work. In my case, some games, I can t stop playing to it, Mario 64 ... (N64) , and I like Duke Nukem 3D too, I got to have a old GPU with a old Windows to run this, this is just so frustrating! I like playing wilth all of those vantage games.
Just imaging if tomorrow, we tell you that you can run Windows XP on the next CPU ... wait wait ... they already did it for DX10!
This is just a royal mess.
[edit]
I forgot , creating legacy is more difficult that it seams, you got to plan for the future, and design it to scale in the future, Every 2 or 3 years, ATi and NV had to change the way they were design, and I understand why they had to do so, in the mean time, they did not do the effort to try to plan the next steps, they went for immediat competition, without thinking further. If you look at the Moto 68xxx world against the x86, the scalability of the architecture got them in the long run.
PCI, PCIex , AGP, and USB specs after a while become the main connecting points, all of those are open specs ... What I am brain storming about here is about legacy, Legacy always won, because it generate free consumers, no Whoop ass can at the corner of the store.
[/edit]
last remark : The x86 providers never agreed on price ... some other guys did ;-) and agree the settle down about it ... (Very personal comment!!!)
Last edited by Drwho?; 10-19-2008 at 10:13 AM.
DrWho, The last of the time lords, setting up the Clock.
the end of the NDA have been advanced of 2 weeks (2nd of november instead of 17th of november), is the release date too ?
AFAIK Intel is the only company that has open sourced its graphic drivers for example.
The NDA end is under NDA ...
DrWho, The last of the time lords, setting up the Clock.
This is a different time, when the gains to be had specifically in heavy workload arenas with GPGPU and Larrabee could very well make the CPU's moderate micro-arch improvements irrelevant. They're also up against limits of physics (thermal wall), as well as thresholds of perception in some cases like every day usage. I believe the average PC doesn't have enough workload to justify a new CPU like this (in light of other technologies which can massively speed up parallel applications).
Why spend top dollar on a new processor to process video at 10-40% improved speeds when you can do already do it at 400%-1,000%?
Granted, if they can stick more cores on there, and software developers can figure out how to thread 8 and 16 ways, you'll get comparable speed-ups. But such is the challenge before them.
Last edited by Sr7; 10-19-2008 at 01:38 PM.
wait wait ... there is a lot of marketing done about GPGPU right now, but your GPGPU will never boot your OS, or will never send email, or do the spelling correction
Your GPGPu can t encode for DivX 3.0 (or 99% of the video codecs), neither for MP3 (or 99% of the audio files) or many other legacy format, and thinking that all of those tasks will be done on the GPGPU is thinking that the world will be recompiled: This is naive.
You may get a marketing hype going, but the majority of the software will stay on the processor, why? : LEGACY!
people told me many time that Pentium III was more than what they need ... I hear this for 15 years.
DrWho, The last of the time lords, setting up the Clock.
Another question for you, DrWho??
Socket 1366 comes in November, we all know, in the 2. half of 09 socket 1156 will arrive to replace 1366 in the m1 and p1 segment, if I remember that roadmap correctly.
My concern is: what happen to 1366 after the 1156 was launched? Only XE? Do we have to buy XeonDP for our bloomfield rigs?
Tell us please, if you know and allowed to.
Thanks
beeeep beeeeeeeeeeeeeeeeeep beeeeeeeeeeeep beep beep beeeeeeeeeep beeeeeeeeeeep.Another question for you, DrWho??
Socket 1366 comes in November, we all know, in the 2. half of 09 socket 1156 will arrive to replace 1366 in the m1 and p1 segment, if I remember that roadmap correctly.
My concern is: what happen to 1366 after the 1156 was launched? Only XE? Do we have to buy XeonDP for our bloomfield rigs?
Tell us please, if you know and allowed to.
Thanks
DrWho, The last of the time lords, setting up the Clock.
LGA1156/1160 doesnt replace 1366. It replaces the rest of the LGA775.Another question for you, DrWho??
Socket 1366 comes in November, we all know, in the 2. half of 09 socket 1156 will arrive to replace 1366 in the m1 and p1 segment, if I remember that roadmap correctly.
My concern is: what happen to 1366 after the 1156 was launched? Only XE? Do we have to buy XeonDP for our bloomfield rigs?
Tell us please, if you know and allowed to.
Thanks
High: 1366
Mid/low: 1156/1160
Crunching for Comrades and the Common good of the People.
See, your perspective is so legacy-centric, you've missed my point. My point is that.. who *cares* what boots your PC? Sure it makes the CPU an essential component, but not a special/interesting component that can sell for large margins. There is a difference between "essential" and "special". For example, every system must have ram slots, but who gives a crap where the motherboard makers buy them from. Catch my drift? Essential components does not imply *important* components, in terms of what is doing the heavy lifting and saving you a lot of time.
If you're not doing the heavy lifting on a system, you become a commodity. Just because you end up in every system doesn't imply that you'll be able to sell your technology for large margins.. look at onboard sound and networking chips.
What about GPGPU is stopping someone from making an MP3 or divx encoder? Do you mean "it's impossible", or "it's not out there right now"? If it's the latter, your point is fairly weak.
There's no doubt the majority of software will stay on the processor. That's the hole in your argument though.. the majority of software doesn't need 8 cores/threads either. And who cares if the majority stays on the processor, if the *interesting* work is done elsewhere?
Last edited by Sr7; 10-19-2008 at 02:17 PM.
I don 't know if you remember, with the Pentium II, if you wanted to play DVD, you needed a Special MPEG2 card ... many people were saying that you need only a low Pentium II and this card, and you had a good PC.
Now, DVD play back takes 5% of a Core 2 Quad ... The generic processor always catch up with custome pipe line, it is just a matter of time, and you keep the Legacy
See my point now?
DrWho, The last of the time lords, setting up the Clock.
As for the subject of encoding and such, It's not impossible to make MP3 or divx encoders for each GPGPU API available out there. It just has to be done for each and every variety of API. That API also has to be rewritten and expanded to cover each generation of hardware that falls under it's scope. Right now, just about every time the API gets updated to add new features, on some level programs coded to work with older versions of the API have to be rewritten to work with the new API, as the update broke some functionality in order to make room for a new way.
Everything is incompatible without someone going in and making that compatibility. CUDA and other API's do grant huge speedups to certain types of workloads, but until there's some standard legacy support, updates and new hardware ends up breaking functionality for something. It's still progress, but it requires a lot more work to shoehorn programs onto ever changing hardware that at the base level, looks incredibly dissimilar to it's previous incarnations.
Main-- i7-980x @ 4.5GHZ | Asus P6X58D-E | HD5850 @ 950core 1250mem | 2x160GB intel x25-m G2's |
Wife-- i7-860 @ 3.5GHz | Gigabyte P55M-UD4 | HD5770 | 80GB Intel x25-m |
HTPC1-- Q9450 | Asus P5E-VM | HD3450 | 1TB storage
HTPC2-- QX9750 | Asus P5E-VM | 1TB storage |
Car-- T7400 | Kontron mini-ITX board | 80GB Intel x25-m | Azunetech X-meridian for sound |
I think what some people miss in this is that a computer is a total of the systems involved.
Yes, today all the excitement is on GPU processing but that is just a part of the whole process.
I look at it all and see that one step progresses then another catches up or even surpasses and then the others follow.
You have to look at it from that perspective to understand the cause and effect that drives the development of these systems.
Crunch with us, the XS WCG team
The XS WCG team needs your support.
A good project with good goals.
Come join us,get that warm fuzzy feeling that you've done something good for mankind.
I see your point but that was long before CPUs had hit the thermal wall, when each subsequent CPU generation really was night and day from it's predecessor. I feel like once you hit the thermal wall, you must rely on micro-arch improvements, and even integrating the IMC gets you a 1 time speed-up. Where do you go from there? It seems like all you have is more cores to throw at the problem, leaving developers to fend for themselves WRT utilizing them via threading and parallelized workloads.I don 't know if you remember, with the Pentium II, if you wanted to play DVD, you needed a Special MPEG2 card ... many people were saying that you need only a low Pentium II and this card, and you had a good PC.
Now, DVD play back takes 5% of a Core 2 Quad ... The generic processor always catch up with custome pipe line, it is just a matter of time, and you keep the Legacy
See my point now?
I don't doubt they WILL get faster, I'm just thinking maybe only trivially faster for current day workloads, and not faster by as much as P2->P3->P4->Core 2 transitions were.
If you can't do encoding/decoding/transcoding nearly as fast as another technology *today*, then what relevant workload are you left with that you're going to show your new processors performance benefits via, in the average system? Opening browsers at 2ms instead of 10ms. My point is you end up with gains where the % gain is technically huge, but where the absolute gains are below a perceptible threshold in applications where people don't really care/notice.
Updates to APIs don't break functionalities... that's the whole point of an API. To rely on that function being there and completing the task you expect it to. What you're saying is like me saying that updated CPU generations break legacy programs. It doesn't happen. Think of an instruction set as a hardware version of an API.
On a GPU, inter-generationally, you end up doing the architecture specific part of the API in a manner that is not exposed to the API user. So they don't worry about the architecture differences.
browse ytmd or /f/ and you will see what happens with a el chepo Pentium E5200 compared to a QX9650.
ANd i think thats what avarage joe also does.
AMD does it too, and their driver team still does a proper driver.
Open source =! Good. Mostly it's actually inversely proportional to software quality (I'd rather use the proprietary Radeon Linux driver than the other 2 OSS drivers)
I agree that's what average joe does. So in terms of absolutes, sure, maybe I exaggerated. But between my 2.33 Core 2 Duo and Core 2 Quad at 3.0GHz that I use, I don't see much/any gain in terms of sites like this. And keep in mind this is supposed to be the difference between $80 and $550... this thing that isn't perceptible to many users out there.browse ytmd or /f/ and you will see what happens with a el chepo Pentium E5200 compared to a QX9650.
ANd i think thats what avarage joe also does.
By the way, now I can get a Core 2 E8400 at 3.0GHz for $150. That kind of product used to cost in the $300-$450 range. Think about what that means.
Last edited by Sr7; 10-19-2008 at 02:47 PM.
take 3Dmark up to 05, and try to run it on a 4.0Ghz Core 2 Quad with fast memory (OC too), you are up to a surprise with a G45... if you choose the software rendering drivers
Raterization is very memory limited, so, when you can, try it on Core i7 again ...
DrWho, The last of the time lords, setting up the Clock.
Yes, not to mention the driver they open sourced was an incredibly stripped down, skeleton version of the linux driver with the bare-bones basics. It was more a token gesture to appease the open-source community and make them feel important than anything else.
But the reality is the current cpu's do show increases over previous generations and also run cooler.
I see that myself with the Harpertowns vs the previous clovertowns.
Clovers( on good air) max in the 3150 range while the Harpers max in the close to 4000 range and with identical cooling run 15C less.
Then they also produce close to 40% more work in a given timeframe.
Cooler and more work done in the same time.
That is the advantage of the newer cpu's then add in the lesser current draw.
My clovers at 100% load at 3150 draw 420w, the Harpers at 3758 draw 320w at 100% load.
There is your "absolute gains"" in real numbers..
Crunch with us, the XS WCG team
The XS WCG team needs your support.
A good project with good goals.
Come join us,get that warm fuzzy feeling that you've done something good for mankind.
Well one year ago there whern't cheap 45nm quads around...
why do you insist that Ci7 has no use, the people who are buying it know what the buy, everyone else buys cheap dualcore anyway.
Heck on my main forum the regulars there hardly recommend buying quads, cause most guys dont need it. Only the guys who want to have fun and strive for benching have quadcores.
Posting Permissions
Reply With Quote
Bookmarks