Poster | Thread |
minator
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 15-Jun-2011 20:06:29
| | [ #81 ] |
|
|
|
Cult Member |
Joined: 23-Mar-2004 Posts: 989
From: Cambridge | | |
|
| @umisef
Quote:
(a) These benchmarks were not accessing Natami's DDR2 main memory at all. They were accessing a smallish amount of fast SRAM which is located on the CPU card, and which even Gunnar tends to refer to as "L2 cache". |
Hmm, I suspected it was doing something like that.
I don't see the point of tricks like this though. The project is interesting enough without it.
All this is going to do is convince a load of folk the Natami is going to be incredibly fast. When it turns up and it's nowhere near it's claimed performance they're going to have a lot of ####ed off customers.
_________________ Whyzzat? |
|
Status: Offline |
|
|
Lou
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-Jun-2011 11:26:38
| | [ #82 ] |
|
|
|
Elite Member |
Joined: 2-Nov-2004 Posts: 4169
From: Rhode Island | | |
|
| @umisef
Quote:
umisef wrote: @AmigaBlitter
Seeing how "500MHz 030" is pretty exactly what the 68060 card does right now, on real hardware.
|
Well, let's say the average '030 is running at 30Mhz. A 500Mhz '030 is about 16x faster. Now at the same clockspeed, how much faster is an '040 compared to an '030? Now at the same clockspeed, how much faster is an '060 compared to an '040? Now multiply by 110Mhz/30Mhz. What # do you get?
Edit: 'faster' isn't the right word here. IPC is more appropriate. Given that the '030 is probably the most popular Amiga (accelerated) cpu, a x16+ boost in processing power is not an unrealistic claim.Last edited by Lou on 17-Jun-2011 at 11:31 AM.
|
|
Status: Offline |
|
|
umisef
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-Jun-2011 12:07:24
| | [ #83 ] |
|
|
|
Super Member |
Joined: 19-Jun-2005 Posts: 1714
From: Melbourne, Australia | | |
|
| @Lou
Quote:
Much easier --- look at the screenshot posted here, and see the Natami's 120MHz 060 rated as "19.58 times as fast as a 25MHz A3000-030/25".
|
|
Status: Offline |
|
|
Hammer
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-Jun-2011 13:52:40
| | [ #84 ] |
|
|
|
Elite Member |
Joined: 9-Mar-2003 Posts: 5246
From: Australia | | |
|
| @Samurai_Crow
Quote:
Samurai_Crow wrote:
I don't have a Windows machine to try WinUAE on. And the fact that as a large program, Total Chaos AGA overflows the JIT cache making it run slowly on every version of UAE that I've tried. It's rare for an Amiga program to be developed to this point of memory consumption and still use memory efficiently as Total Chaos AGA does. It's reasons like this that Team Chaos Leader has joined the NatAmi team because emulators don't always cut the muster. Look for a Total Chaos version enhanced for the NatAmi's capabilities later on.
|
Which Intel Mac Mini are you using? It's memory support ranges from 667 MHz DDR2 to 1066 Mhz DDR3. Intel Core 2 Duo's FSB is half duplex.
The late 2009 Intel Mac Mini runs the same Intel Core 2 Duo P8700 2.53Ghz/1066Mhz FSB as my old Sony Vaio VGN-FW45 laptop. I also have ASUS G1Sn laptop with Intel Core 2 Duo T7500 2.2Ghz/800Mhz FSB/667Mhz DDR2 memory.
On my ACER Iconia W500 tablet and WinUAE 2.3.2/basic AmigaOS 3.1, Total Chaos AGA r7 seems to run OK.
My host CPU usage is around 68 percent with one CPU core peg'ed around 85 percent. Total Chaos AGA r7 seems to be a basic game compared to Quake 68K. As a turnbase game, it's not same level as Win95/98 era Lords of Magic. There's an Amiga version Battle for Wesnoth http://www.mobygames.com/game/battle-for-wesnoth
AMD C-50 APU doesn't have a FSB and it has complete access to 2GB of DDR3-1066Mhz memory.
Last edited by Hammer on 17-Jun-2011 at 02:19 PM. Last edited by Hammer on 17-Jun-2011 at 02:14 PM. Last edited by Hammer on 17-Jun-2011 at 02:06 PM. Last edited by Hammer on 17-Jun-2011 at 02:03 PM. Last edited by Hammer on 17-Jun-2011 at 02:02 PM. Last edited by Hammer on 17-Jun-2011 at 01:57 PM.
_________________ Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68) Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68) |
|
Status: Offline |
|
|
pavlor
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-Jun-2011 14:14:45
| | [ #85 ] |
|
|
|
Elite Member |
Joined: 10-Jul-2005 Posts: 9578
From: Unknown | | |
|
| @Lou
Quote:
a x16+ boost in processing power is not an unrealistic claim. |
Sure, nice for 68k architecture.
For comparison, 440EP 666 MHz is roughly more than 130 times faster than 68030 33 MHz (in integer operations). |
|
Status: Offline |
|
|
ErikBauer
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-Jun-2011 14:51:18
| | [ #86 ] |
|
|
|
Super Member |
Joined: 25-Feb-2004 Posts: 1141
From: Italy | | |
|
| @pavlor
In fact NatAmi is not engineered to beat PPC or X86, just to be the fastest 68K Amiga ever.
_________________ God created Paula so that Allister Brimble and Dave Whittaker could do music
Check my Amiga gameplays (ITA)! |
|
Status: Offline |
|
|
pavlor
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-Jun-2011 15:05:31
| | [ #87 ] |
|
|
|
Elite Member |
Joined: 10-Jul-2005 Posts: 9578
From: Unknown | | |
|
| @ErikBauer
Quote:
In fact NatAmi is not engineered to beat PPC or X86, just to be the fastest 68K Amiga ever. |
Exactly. That is why I like Natami project. |
|
Status: Offline |
|
|
Samurai_Crow
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-Jun-2011 15:13:38
| | [ #88 ] |
|
|
|
Elite Member |
Joined: 18-Jan-2003 Posts: 2320
From: Minnesota, USA | | |
|
| @Hammer
My Mac Mini has the 667 MHz DDR2 memory. The Core 2 Duo is clocked at 1.83 GHz. It's not the latest model. It's the one I had to completely disassemble to boost it to 2 Gigs of RAM. |
|
Status: Offline |
|
|
AmigaBlitter
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 8:26:17
| | [ #89 ] |
|
|
|
Elite Member |
Joined: 26-Sep-2005 Posts: 3512
From: Unknown | | |
|
| |
Status: Offline |
|
|
itix
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 8:53:13
| | [ #90 ] |
|
|
|
Elite Member |
Joined: 22-Dec-2004 Posts: 3398
From: Freedom world | | |
|
| @AmigaBlitter
Faster than PowerPC 440... _________________ Amiga Developer Amiga 500, Efika, Mac Mini and PowerBook |
|
Status: Offline |
|
|
KimmoK
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 9:04:29
| | [ #91 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2003 Posts: 5211
From: Ylikiiminki, Finland | | |
|
| @AmigaBlitter
I assume those results are with identical clock frequency.
"The study was conducted in simulation and on a PLDA FPGA PCIe card using a ALTERA Stratix4 230C2"
I wonder how high performance such softcore can reach today when maxed out. 100 000 + MIPS ?
Update... I understand that Apollo can be as fast ast e6500 in integer per core (6MIPS per Mhz). Did I get it wrong? if only we could have Apollo softcore running at 1.8Ghz or more... Does Apollo support SMP?
Kickstarter project to have a Apollo chip manufactured @28nm tech... With some onchip RAM it might be nice accelerator chip for old miggies ... later AAA/Hombre on the same die to do a 2Ghz ClassicAmiga replacement.... Last edited by KimmoK on 04-Apr-2012 at 09:41 AM. Last edited by KimmoK on 04-Apr-2012 at 09:37 AM. Last edited by KimmoK on 04-Apr-2012 at 09:06 AM.
_________________ - KimmoK // For freedom, for honor, for AMIGA // // Thing that I should find more time for: CC64 - 64bit Community Computer? |
|
Status: Offline |
|
|
wawa
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 9:25:57
| | [ #92 ] |
|
|
|
Elite Member |
Joined: 21-Jan-2008 Posts: 6259
From: Unknown | | |
|
| @itix
Quote:
Faster than PowerPC 440... |
i ll believe it after seeing a proof.. |
|
Status: Offline |
|
|
olegil
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 10:27:42
| | [ #93 ] |
|
|
|
Elite Member |
Joined: 22-Aug-2003 Posts: 5895
From: Work | | |
|
| @KimmoK
"The APOLLO Core offers a superscalar integer performance of over 300 MIPS at 200 MHz." According to their own website.
Which is a little strange considering their other claims of 3 instructions per clock etc. But possibly the 3 instructions are 1 math, 1 branch and 1 memory access, which means real life applications would have a hard time filling up the pipeline.
And it's a soft core. The FPGA arches have moved from 300MHz 10 years ago to 600MHz today (mostly the focus is on increasing perfomance per logic gate/flip flop per watt in the FPGA world, not on increasing clock frequency per se). So 1.8GHz just ain't gonna happen in an FPGA. _________________ This weeks pet peeve: Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean. |
|
Status: Offline |
|
|
KimmoK
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 10:33:04
| | [ #94 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2003 Posts: 5211
From: Ylikiiminki, Finland | | |
|
| @olegil
Yes. That's why I was hinting about getting asic/SoC done around Apollo core. (futile thing, but strange things are done sometimes, I rather see Power core taken forward... like kickstarter bounty for verisilicon to run PPC460+GPU chip mass production run. Or e6500 core + GPU ...) Last edited by KimmoK on 04-Apr-2012 at 10:33 AM.
_________________ - KimmoK // For freedom, for honor, for AMIGA // // Thing that I should find more time for: CC64 - 64bit Community Computer? |
|
Status: Offline |
|
|
kamelito
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 10:33:38
| | [ #95 ] |
|
|
|
Cult Member |
Joined: 26-Jul-2004 Posts: 813
From: Unknown | | |
|
| @amigadave
AFAIK OS4 for classic require a PPC card so it's not 68k so why should it run on Natami?
Kamel |
|
Status: Offline |
|
|
Channel_Z
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 10:37:57
| | [ #96 ] |
|
|
|
Regular Member |
Joined: 4-Mar-2009 Posts: 305
From: Unknown | | |
|
| @thread
http://www.natami.net/knowledge.php?b=6¬e=44420&order=&x=14
Third post from bottom. So the Apollo page is just a placeholder, and nothing official. The numbers might just be invented. Anyway it is more interesting to see the results in the Cyclone4 in the Natami, since noone can afford to build a system with that Stratix FPGA anyway. |
|
Status: Offline |
|
|
Channel_Z
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 10:42:15
| | [ #97 ] |
|
|
|
Regular Member |
Joined: 4-Mar-2009 Posts: 305
From: Unknown | | |
|
| @itix Quote:
Faster than PowerPC 440... |
The page is comparing the results with an existing PPC440 softcore, not with an ASIC PPC440. |
|
Status: Offline |
|
|
AmigaBlitter
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 4-Apr-2012 10:53:47
| | [ #98 ] |
|
|
|
Elite Member |
Joined: 26-Sep-2005 Posts: 3512
From: Unknown | | |
|
| @Channel_Z
Sure we are talking about softcore, both for PPC440, Nios, and Apollo softcore. Anyway, if all is done correctly, the respective asic version should perform the say way.
This Natami softcore is indeed interesting. Of course you can't run the software core at very high speed in the current FPGA they use for the Natami. But just think about this: if Freesale take a look at the Apollo performance, may consider to buy the softcore and extend the 68000 line up. In fact they use already synthesizable core for some PPC and Arm line.
_________________ retired |
|
Status: Offline |
|
|
number6
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-May-2012 17:11:48
| | [ #99 ] |
|
|
|
Elite Member |
Joined: 25-Mar-2005 Posts: 11540
From: In the village | | |
|
| @thread
Project update from the Natami team:
Post by Claudio Wieland
#6
_________________ This posting, in its entirety, represents solely the perspective of the author. *Secrecy has served us so well* |
|
Status: Offline |
|
|
wawa
| |
Re: Oh, boyz, it's seems that the new baby can beat them all Posted on 17-May-2012 17:38:30
| | [ #100 ] |
|
|
|
Elite Member |
Joined: 21-Jan-2008 Posts: 6259
From: Unknown | | |
|
| @number6
good think it has been clarified at last. nothing unexpected though. the scene is dying off and cant sustain more ambitious hardware projects on its own.
edit: strange only that the project author has now to seclude from the group that has gathered around him and even give up his original project name, even though he is the one who did the actual work. Last edited by wawa on 17-May-2012 at 05:43 PM.
|
|
Status: Offline |
|
|