Dual Core Optimiser for X2 CPU's

maverik-sg1

New member
Inq reports here, I will test it later ;) :

It is the first improvement to Chimpzilla's dual-core offerings of a bundle, so expect a lot of software and BIOS updates coming in July.

Intended for all owners of dual-core processors, AMD Dual-Core Optimizer is a piece of software that allegedly improves gaming performance by bypassing Windows' API in some specific situations.

This utility helps that software access the RDTSC instruction directly, in order to correct the Time Stamp Counter on separate cores. It effectively keeps both cores working in parallel and reduces the differing video performance effects, or any other flaws that relate to timing situations in dual-core or upcoming multicore processors.

You can find and download the utility using our complementary L'Inq. L'Inq

Chimp maker
 
I predict no SuperPi boost, but Windows will load faster ;)

Its part of R-HT, apparently, but the other pieces are still to be released.
 
name='Animal' said:
Isnt R-HT only in AM2..?

I think it is mate - this is a sort of driver to do the same for skt 939, almost a 'for being loyal' bonus.

Tried it, think without a MOBO BIOS update there's not much going with this right now.

Mav
 
I hate to say this but told you so. I think this is all myth personally and I can't see it working very well at all even on AM2.

Boardy
 
name='boardy' said:
I hate to say this but told you so. I think this is all myth personally and I can't see it working very well at all even on AM2.

You love to say this don't you mean? ;)

I think if it works at all (in the same way SLI works) - AM2 owners will be happy, after all it was not something that was mention during the launch so anything for nothing is a bonus.

If it works and proves to be a core duo killer application - who knows?

Mav

(waiting patiently)
 
name='maverik-sg1' said:
You love to say this don't you mean? ;)

Well yea :D

I love and hate to say it really, cos atm AM2 is looking pretty pants compared to 775, I would like to see AMD get sumthin tht could beat Core 2, but atm I just dont see it at all, short of buying this STUPID idea of 2 CPU's on 1 board.

£1400 for a Core 2 beater (excluding board I might add) which costs £600 (including board) anyone? I don't think so somehow.

Boardy
 
Thing is what about Intel kentisfiled when she arrives. It's out in rev 2 es format still with a massive mod and overhal thnak god. Its listed as core duo currently though it's quad core so AMD will have to beat that and at 3.3ghz it's eating everything alive
 
name='Drazic' said:
Thing is what about Intel kentisfiled when she arrives. It's out in rev 2 es format still with a massive mod and overhal thnak god. Its listed as core duo currently though it's quad core so AMD will have to beat that and at 3.3ghz it's eating everything alive

I am not sure what this has to do with this thread?

I think the point is that for early adoptors of AM2 and for existing 939 dual core owners - if this feature works (just a bit) it is somehting that we were not aware of when we made the purchase.

Free performance boosts are most welcome.

I do share Boardy's concerns that it may not be enough (if any).

Although I maintain that any true comparisions are irrelevant until AMD go 65nm - even then, quad core is just round the corner, but (with the exception of benchmarking) how important will that be? It's obvious that current software packages do not stretch todays processors - and high end gaming is well catered for also, even the almighty FEAR can be played on my rig at 1900x1200 with full candy.

The big question is - do we need to make the leap to these new platforms now, knowing that quad core is less than 8 months away? Which means that whoever sets the best scores now - will need to rebuff a challenge from quad cores cpu's before they beat my record for longest stint at no1 :)

Mav
 
Heres tuppence!!

Who remembers when Intel moved to DDR2? That was a new mobo, CPU, RAM AND that was when the move to PCI-E was made as well.

This move by AMD isnt quite as bad, tho its far from ideal.

AMD have 65nm CPUs demoed for `939 that were given to the VERY VERY select few. they used the exisiting 939 platform as a testing ground to make sure all was well. This was 4 months ago.

AMDs 65nm is fine. They`ll have had time to make other tweaks, and start on what became K8L

AMD also have quad-core demoed on 65nm. Just because it aint publicised, doesnt mean it dont exist. Coolaler and Hicookie have been allowed to publish results coz Intel want to be the ones to get there first...big themselves up, if you will. Though..they dont need to.

AMD and Intel TALK to each other. They hate each other, but they talk. Both benefit from neither side royally owning the other for too long.

Edited for tact and relevance
 
I certainly recall the move to DDR2 was less than spectacular, also that the PCI-E frequency was not locked so overclocking was a pain and the systems were proven to be generally slower, hotter and used more power than its skt478 counterparts.

Tis true the AMD move is at least (generally) faster than it's 939 counterparts and also, less power/heat concerns - and thats on the same die size.

It all adds up to see if you are able to play the waiting game??
 
The thing that the inquirer mentions does not allow for reverse HT having sync'ed TSC eliminates problems with run away cores etc, its likely the patch is booting a switch that stops house keeping accessing the TSC and then syncing the TSC on load up. It would make windows a little more responsive in some situations and stop problems with mulit-threading applications.

Short of the AM2 reverse hyperthreading being an additional part of the driver its possible that its going to be released at a later date. You also need BIOS that supports the "multiplexing" so even if it does support no on will no yet.

The reverse hyperthreading isnt the really hardcore version AMD hopes to bring in with K10 which will do it all via hardware. But its a step in the right direction and should be brillant for AMD 4X4.
 
Back
Top