Slashdot Mirror
New VIA x86 CPU Takes Aim At Intel Silverthorne
Kaz writes "While not operating on the same scale as the two major CPU designers, VIA has been gaining traction in the world of UMPCs and thin clients with its Eden and C7 lines of processors. While past architectures have been considerably out-of-date in terms of modern features, the new Isaiah architecture looks to be very competitive with what AMD and Intel have lined up for future ultra-mobile products. It features an out-of-order, superscalar execution core, 64-bit support, virtualization, and even SSE3 — all on a 94M-transistor, 65nm process die. The initial offering will be single-core only, though VIA says that multi-core ability is already designed in. Is Isaiah going to replace your Core 2 system for gaming? No, but it might give Intel's Silverthorne a run for the money."
Google says: ... VIA C7 @ 1.2 GHz LV (TDP 7 W). VIA C7 @ 1.0 GHz LV (TDP 5 W)
Support VIA C7 @ 1.5 GHz D (TDP 25 W). VIA C7 @ 1.5 GHz (TDP 12 W). VIA C7 @ 1.3 GHz (TDP
So the C7 can be a 5W part too. Which is not too bad for a 1GHz CPU.
I guess the ISAIAH will have such a version too. Sounds interesting, doesn't it?
They misspoke then, the current Via will get 20 watts peak, if you include the systemboard. The processor (c7-d) is 10 watts of that. They have another processor (c7), at somewhat lower speed (1 ghz instead of 1.5), that runs on 2 watts or somesuch.
Please note the 'peak' intel measures its wattage on averages, not peaks like AMD/VIA.
Take a look at the newer VIA VB7001G board. It may be the C7-D processor, but from logicsupply dot com it is $123. Not a hugely cheap board, but quite nicely priced for a mini-ITX board. The only drawback is that some of the cases cost almost more than the M/B itself!
Also the gOS boards are quite nice, though at micro-ATX are harder to fit in to a low power solution... I have two of these, one running my router with a dual Netflex-3 card (yeah I know, older 10/100, but I don't need any faster) and it runs quite well.
I'd be interested to see how this new chip/chipset combo works in say a HTPC and if it does HD content well. None of the current VIA Unichrome chipsets do HD very well.
Mark
I've had nothing but good luck with them. Combined with a mini-itx fanless case, these things make great appliances. Here's a great place to get them:
http://www.logicsupply.com/
At work, we used the mini-itx with fanless case for branch office VPN solutions using linux + openswan (which in turn connected back to checkpoint clusters as well as other branch office openswan gateways). At home, I have a VIA chipset m/b with an Athlon 3000+ processer which has been running great for me for a few years.
VIA's Pico-ITX full systems (not just the chip) have already by clocked at 14w idle, 16w max in pre-release reviews from 6 months ago. The previous generation C7s can easily be throttled back to stay at 5w on the proc as needed. I'm not sure if such functionality is available on the new Pico systems though.
Intel is "shooting for" a 5w processor (no clarification if this is max load, or idle) in 2010.
VIA's Pico-ITX is already available at 1ghz, and the previous generation C7's are available up to 2ghz.
Intel's Silverthorne processor is also aiming for the Pentium M era performance (900mhz - 2.3ghz).
Yes, the initial Silverthorne release is slated for Q1-Q2 2008, but the performance goals you mentioned aren't slated until 2010. So what I'm saying here, is that you can already buy everything that Intel is "shooting for" 2 years before they plan on reaching those goals. With all likelihood, the 2008 release of the Silverthorne will be a 1ghz proc sucking down 20w at peak. Which will put it right in competitive range of the C7 and new Pico-ITX.
-Rick
"Most people in the U.S. wouldn't know they live in a tyrannical state if it walked up and grabbed their junk." - MyFirs
I re-read the post you replied to several times, but couldn't find the place where it advocated replacing the cache memory with DRAM. Instead it was about new memory technology yet to be found. Here's the relevant quote, with emphasis by me:
:-)
"If you can find a CMOS-compatible, high-density (e.g. - SRAM's six transistors per cell is toooo big) memory technology, then we're going to be at the point where we can simply replace the cache memory with on-board memory. If said on-chip memory technology is nonvolatile, then we're talking panacea cakes, batman."
I'm interested to see your non-volatile DRAM.
OK, I grant you that he didn't explicitly talk about performance requirements. However, if he considered DRAM as sufficient, why would he have asked for new memory technology?
The Tao of math: The numbers you can count are not the real numbers.
For a while I was on a mission to build a really power efficient PC. Unfortunately when I got my AC power meter, I learned a number of disappointing things:
but as the article said, this time it's more powerful. The C7 is not particularly strong because of its in-order execution core, and the new CPU appears to fix this.
For the record, my 2 GHz C7 machine can play a 720p h.264 video smoothly, but only without sound :) This is using MPlayer, no hardware acceleration except Xvideo.
Escher was the first MC and Giger invented the HR department.
"What he means is: forget on-chip cache -- on-chip main memory. IOW, instead of having main memory on the motherboard, it would be embedded into your processor, running, presumable, at the same speed as the CPU."
Memory on the die has been done in micro controlers for years. It isn't going to happen on PCs for a long time.
"A CPU like this new VIA CPU might be slow, but if you had sufficient memory integrated right on the CPU die, it would blow the pants off your latest 4+GHz Core 2 Duo."
What is sufficient memory? 4 GB or Maybe 512 MB? There is a reason that they use Static ram for cache. It needs to be fast. So lets say that you get 512 MB on the die are you not going to allow the user to add more memory? Or how about this. You put 512 MB on the die and then let them add memory on the buss if they need more. And then you could have it swap memory from the slower buss memory in to the fast on die memory to speed everything up... Yea and we could call it a cache!
Until you can put the full address space on the die it will not work for anything but microcontrollers.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
"I don't claim to know a whole lot about processor tech, but I know that it wouldn't be hard to add this sort of optimization."
It is hard and it has been done. That is exactly what the cache does.
The tasks that need the memory the most stay in the cache longest.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
My first clue you were full of crap was this: "Silverthorne will be a 1ghz proc sucking down 20w at peak". I'm not sure if you pay attention, but Intel has Core 2 Solo chips running at 1.06/1.2Ghz that peak at 5.5 watts. Silverthorne is a 45nm chip running on a simplified core-2-esque march, and you're making this ridiculous claim that it will "suck down" 20w at peak.
Seriously, 2006 called, it wants its news back.
That's fine that you feel that way. Now show me a news report, white paper, press release, or ANYTHING that backs those numbers up.
The only press release that I could find that had actual numbers on it said that 5w was the goal of the product line by 2010. So if you have something better to go by than armchair techno-forecasting, please, go ahead and post it.
-Rick
"Most people in the U.S. wouldn't know they live in a tyrannical state if it walked up and grabbed their junk." - MyFirs
And after I RTFA it does have the on-board cryptography, and according to the article it will have double the performance of current Via chips. So this could be a very good chip for a low powered business laptop. I know that my P4m rarely gets above 600Mhz when I'm working on it, so I doubt that I'd even notice if you switched it with a Via. And having quick hashing and encryption/decryption would be a really nice feature if I had clients data on it.That's my
ACs don't waste your time replying, your posts are never seen by me.
On single-threaded CPU's, perhaps. But look at the Sun UltraSPARC T1 and T2. They are multithreaded - each core rotates between up to four threads on each clock cycle. When a cache miss occurs, it simply pulls the affected thread from rotation and continues with the remaining threads while fetching the data in the background. This means cache misses have a much smaller impact on performance than they do on single-threaded CPU's. Thus they need much less cache to maintain performance and throughput.
Putting main memory on the die just isn't practical, except for application-specific embedded microprocessors. It would be expensive and wouldn't actually give you much of a speed increase. Cache is a way of using a small amount of fast memory to speed up some slow memory. The nice thing about it is that the speed increase is out of proportion to its size because of common access patterns. And it's transparent.