VIA's new CPU: Isaiah, and its good!

[taltamir]

http://enthusiast.hardocp.com/...wxLCxoZW50aHVzaWFzdA==

This is awesome. via built a new CPU. Its not comparable to the high end intels, but its powerful enough to run crysis, its x86 with 64bit compatibility. And they say it will be SSE4 capable this year.

The thing is pin compatible with their existing architecture, but twice as fast per clock.

Maybe next year it will be via vs intel...

 

[Zenoth]

I can see AMD buying VIA.

Hmmm... no.

It might be the other way around...

 

[taltamir]

ha, you jest but it might still happen

I am hoping for AMD to make a comback with fusion in 2009, for intel to put forth a good video card in 2009, and for via and nvidia to merge and start making complete platforms of quality CPUs and Video cards...

That will be the absolute best situation possible, three companies each with a complete platform combining cpu, gpu, and chipset.

 

[aigomorla]

Originally posted by: Zenoth
I can see AMD buying VIA.

Hmmm... no.

It might be the other way around...

LMAO!

thats if samsung doesnt gobble it up first.




Imagine a large 25000:1 contrast Flat screen TV with an AMD processor built in on an ATI chipset with ATI card.

nice little all in 1 package.

 

[zach0624]

this is pretty sweat, maybe we could have a third major player in the cpu market?

1.8 and playing crysis is great and it says they are going to come out with a dual core soon too so maybe it is possible that these will get under 50w tdps.

 

[larciel]

Competition is good! Keep it up VIA and AMD!!

*runs away with his new e8400 in arms.

 

[Sylvanas]

Ah good stuff to hear VIA is still plugging away, can't wait to see this product in action. I remember my last piece of VIA tech was an MSI KT4V motherboard- great mobo from what I remember, go go competition!

 

[Avalon]

I had posted a thread on Isaiah recently, where we had a small discussion about it. Good to see more news on it.

 

[Amaroque]

That's a cool little chip! :beer: For VIA!

 

[StormRider]

It's great to see more players in the CPU market. What sort of motherboard would be able to take a Via processor? What type of socket does it uses?

 

[Falloutboy]

if they do an all in one board cpu combo that can do 1080p and supports HDMI out its going to be very popular with the HTPC croud

 

[nyker96]

Originally posted by: Zenoth
I can see AMD buying VIA.

Hmmm... no.

It might be the other way around...

who knows maybe VIA will be buying AMD if this thing is a hit!! :=]

 

[taltamir]

btw, intel made a super portable CPU... and this isaiah supposedly kicks its butt.

And yes... this does look like it could make a very attractice HTPC

 

[ArchAngel777]

I'd be glad to see Via make a viable chip. But at this point, I probably will not believe it until some of the major reviewers have the silicon in hand.

 

[d3lt4]

That looks sweet. I wouldn't mind buying one for an HTPC or laptop if it is cheap. And if it overclocks well for HTPC. That would make my soul more at rest than buying an intel chip. Is there an estimated price on these? How good are VIA Mobo's? From my previous experience they are horrible. Would Nvidia bother to make a board for them do you think? Are they any better now?

Hopefully we can have 3 chipset/ CPU makers that all offer competitive products. That would be great for everyone. Now we just need Intel to stop for a bit, AMD to release a great new architecture, and VIA to sell a lot of these.

 

[taltamir]

Is hardocp not a good enough reviewer for you?

 

[Zap]

Originally posted by: StormRider
What sort of motherboard would be able to take a Via processor? What type of socket does it uses?

They currently don't have plans to sell the CPU separately. These are intended to be soldered directly to motherboards, and the planned motherboards are still the usual mini ITX. Makes perfect sense as these new chips are pin compatible with the current VIA C7 chips, so board makers can just immediately start using it. Supposedly (depending on application) the new chip outperforms the C7 anywhere from 2x to 3x at the same GHz. I suppose that wasn't too tough to accomplish, but they did this with a 20W TDP, and expect "normal" usage to only burn around 5W or so. Note that these chips are single core.

Since these are pin compatible, what I'm looking forward to is a future Everex Cloudbook with one of these and a bigger-than-7" screen, with the same price and form factor. Now THAT would be an UMPC worth owning.

 

[n7]

Pretty impressive sounding.

 

[CTho9305]

Some thoughts:

• Based on this photo, it looks like a substantial amount of the design is custom rather than synthesized. I had always assumed Centaur's CPUs were almost completely synthesized.
• It looks like the various parts are placed with routing channels in between them, which has plusses and minuses when compared to directly abutting the components (the plus is reduced design effort; the minus is the area penalty; compared to the 65nm K8 die, there is much more white space.)
• A single-core die + 1MB is 60mm^2. A 65nm dual-core K8 + 512KB per core + a northbridge is 126mm^2, so a K8 + 512KB + half a northbridge is 63mm^2. I guess that puts the total area of the core in the same ball park.
• Isaiah is 25W @ 2GHz @ 1.15V. The BE series of AMD chips is 45W @ 2.3GHz @ 1.25V (for 2 cores+northbridge). If AMD chopped a core and the northbridge, and undervolted to run at 2GHz (instead of using worse bins and keeping it at higher voltages), the power would probably drop substantially, but the Isaiah still doesn't look bad at all.
• Assuming 70M transistors per MB of L2 cache, Isaiah is 94-70=24M transistors. I can't remember numbers for K8 / Core 2, but I think they're both more transistors.
• Their L1 is 64KB, 16-way associative. That's bigger than Intel's L1's, and more associative than AMD's. Their load to use penalty is 4 cycles rather than 3 though. I wonder whether the 4 cycle latency is caused by that cache... or if other aspects (e.g. the memory disambiguation, store-to-load forwarding for different access sizes, etc) were complex enough that they were the critical path, making enough time available to use a large and highly-associative L1.
• Their "7-wide" execution is less than K7 and K8 (3 load or store address, 3 integer, 3 float), but I think it's wider than Core 2 (I count 6, using Via's nomenclature).
• Their decoder sounds like it might be more powerful than Intel's, which can only handle a subset of instructions in all slots except the first slot. Having the front-end run ahead when one instruction is in the microcode ROM is pretty clever. Some of the microcode stuff makes me wonder if they've been reading competitors' patents (or whether everybody handles it the same way when taking a 10,000ft view).
• It's interesting that their division and square root happen on their adder port rather than the multiply port. I wonder if they have dedicated hardware for those operations.
• The thermal management features are cool.
• In the video, the guy puts Silverthorne and Isaiah in the same power range. I think Silverthorne's power is significantly lower than that (.5-2 watts rather than ??-25W).
• This chip was first announced in 2004, for first availability in 1H06!

Originally posted by: taltamir
btw, intel made a super portable CPU... and this isaiah supposedly kicks its butt.

Intel's "super portable" CPU is called Silverthorne (or Bonnell) and it's somewhere in the 0.5-1W range. This thing is 25W at 2GHz. The IPC of Isaiah is probably a lot higher, but even at low frequencies and reduced voltages it's not competing in the same ballpark. The die is also enormous in comparison, which means Silverthorne would be significantly cheaper.

That said, this still sounds pretty cool.

 

[VirtualLarry]

Does Isiah have the extra crypto features that other Via CPUs have?

 

[taltamir]

The 1Ghz isaiah takes less then 1volt. And I think the intel system takes more power overall... at least until intel finishes redesigning the CHIPSET that goes with it to ALSO be low power. via already has a low power chipset.

 

[nerp]

I'm excited about this chip. I've been contemplating building a HTPC soon and now I'm inclined to wait for this to become available.

 

[Amaroque]

Originally posted by: nerp
I'm excited about this chip. I've been contemplating building a HTPC soon and now I'm inclined to wait for this to become available.

Don't hold your breath. It looks great on paper though.

 

[IntelUser2000]

Assuming 70M transistors per MB of L2 cache, Isaiah is 94-70=24M transistors. I can't remember numbers for K8 / Core 2, but I think they're both more transistors.

Here's the problem of finding transistor count of the actual CPU core.

1. The size of the L1 cache is included, so comparison between different L1 cache sized CPUs like Core 2 and K7/K8 is very hard

2. Simple math doesn't work.

Core 2 Conroe: 19 million with L1 cache
K7: 22 million with L1 cache

(The L1 cache of K7 mentioned is 152KB because its 64KB Instruction+64KB Data+3 Pre-decode Bits/Byte (adds 24 KB))

K8 is probably not much more than 22 million, and Penryn isn't too much over Conroe either. Only way to know the actual transistor count is for the manufacturers themselves to reveal it.

Anyway, back to CN Isaiah. Reports say that the two FP units are not symmetrical. One unit is dedicated to FPAdd(FP Add) and other is dedicated to FPMul(FP Multiply).

Still, damn lot better than what their other CPUs used to be. I'd say it'd be up to Pentium M performance range.

 

[CTho9305]

Originally posted by: IntelUser2000

Assuming 70M transistors per MB of L2 cache, Isaiah is 94-70=24M transistors. I can't remember numbers for K8 / Core 2, but I think they're both more transistors.

Here's the problem of finding transistor count of the actual CPU core.

1. The size of the L1 cache is included, so comparison between different L1 cache sized CPUs like Core 2 and K7/K8 is very hard

2. Simple math doesn't work.

Core 2 Conroe: 19 million with L1 cache
K7: 22 million with L1 cache

(The L1 cache of K7 mentioned is 152KB because its 64KB Instruction+64KB Data+3 Pre-decode Bits/Byte (adds 24 KB))

You left out ECC bits and tags

K8 is probably not much more than 22 million, and Penryn isn't too much over Conroe either. Only way to know the actual transistor count is for the manufacturers themselves to reveal it.

Yup. I was trying to come up with a ballpark number.

Anyway, back to CN Isaiah. Reports say that the two FP units are not symmetrical. One unit is dedicated to FPAdd(FP Add) and other is dedicated to FPMul(FP Multiply).

Aren't all CPUs "asymmetrical"? On K7 there's an FP adder pipe, an FP multiply pipe, and a pipe that handles misc stuff (format conversion or stores or something... I forget). The Core microarchitecture has different units on the various ports too, right?