SoylentNews is people
SoylentNews
Russia's Elbrus 8CB Microarchitecture: 8-core VLIW on TSMC 28nm
All of the world's major superpowers have a vested interest in building their own custom silicon processors. The vital ingredient to this allows the superpower to wean itself off of US-based processors, guarantee there are no supplemental backdoors, and if needed add their own. As we have seen with China, custom chip designs, x86-based joint ventures, or Arm derivatives seem to be the order of the day. So in comes Russia, with its custom Elbrus VLIW design that seems to have its roots in SPARC.
Russia has been creating processors called Elbrus for a number of years now. For those of us outside Russia, it has mostly been a big question mark as to what is actually under the hood – these chips are built for custom servers and office PCs, often at the direction of the Russian government and its requirements. We have had glimpses of the design, thanks to documents from Russian supercomputing events, however these are a few years old now. If you are not in Russia, you are unlikely to ever get your hands on one at any rate. However, it recently came to our attention of a new programming guide listed online for the latest Elbrus-8CB processor designs.
The latest Elbrus-8CB chip, as detailed in the new online programming guide published this week, built on TSMC's 28nm, is a 333 mm2 design featuring 8 cores at 1.5 GHz. Peak throughput according to the documents states 576 GFLOPs of double precision, with the chip offering four channels of DDR4-2400, good for 68.3 GB/s. The L1 and L2 caches are private, with a 64 kB L1-D cache, a 128 kB L1-I cache, and a 512 kB L2 cache. The L3 cache is shared between the cores, at 2 MB/core for a total of 16 MB. The processor also supports 4-way server multiprocessor combinations, although it does not say on what protocol or what bandwidth.
It is a compiler focused design, much like Intel's Itanium, in that most of the optimizations happen at the compiler level. Based on compiler first designs in the past, that typically does not make for a successful product. Documents from 2015 state that a continuing goal of the Elbrus design is x86 and x86-64 binary translation with only a 20% overhead, allowing full support for x86 code as well as x86 operating systems, including Windows 7 (this may have been updated since 2015).
Previously: Russian Homegrown Elbrus-4C CPU Released
(Score: 1, Interesting) by Anonymous Coward on Tuesday June 02 2020, @07:21AM (10 children)
(Score: 1, Interesting) by Anonymous Coward on Tuesday June 02 2020, @07:42AM
It's a good start. If they are going to just use TSMC's fabs, then they can get a much better node if they are willing to pay for it.
(Score: 4, Funny) by turgid on Tuesday June 02 2020, @09:11AM (1 child)
Nearly fast enough to run Word, Excel, Internet Exploder and MacAfee concurrently, then?
I refuse to engage in a battle of wits with an unarmed opponent [wikipedia.org].
(Score: 1, Insightful) by Anonymous Coward on Tuesday June 02 2020, @09:36AM
2011 called and want their meme back. It's Edgium and Windows Defender nowadays.
(Score: 0) by Anonymous Coward on Tuesday June 02 2020, @10:52AM (1 child)
Ah, so what you're saying there is 'tried and trusted' technology.....well, for given values of 'trusted' obviously, though I do appreciate the Russian point of view that they'd rather have processors with their own backdoor code in the beasties powering their systems rather than having to use foreign ones with Chinese/Israeli/USian backdoor code lurking in them.
(Score: 1) by petecox on Tuesday June 02 2020, @11:17AM
China have their own x86-emulation in (MIPS based) Loongson, should their Zhaoxin partnership with Via ever be Huaweied by sanctions.
(Score: 0) by Anonymous Coward on Tuesday June 02 2020, @01:10PM (3 children)
I thought translation + VLIW was judged an architectural dead end some time ago. Maybe they found a way to make it work.
Did they provide any standard benchmark results?
Would be interesting to see how the machine does in 32 and 64 bit translation mode and well as native.
Also power consumption.
They are only at 1500Mhz, so there is some process scale left in the architecture, if the above are reasonable.
(Score: 2) by DannyB on Tuesday June 02 2020, @01:59PM
Maybe they're trying to troll the planet, but don't realize that the joke is out of date.
Like Trinity in The Matrix Reloaded, using SSH and NMAP to exploit the SSH CRC-32 bug [securityfocus.com] from Feb 08 2001.
Punched cards are like a short piece of 80 channel paper tape.
The lower I set my standards the more accomplishments I have.
(Score: 2) by takyon on Tuesday June 02 2020, @04:59PM
In other news, x86 emulation/translation on ARM could really take off, as long as Intel doesn't sue.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 1) by petecox on Tuesday June 02 2020, @07:11PM
"translation + VLIW was judged an architectural dead end"
Didn't Nvidia license some Transmeta code-morphing IP in their ARM 'Denver' cores?
(Score: 0) by Anonymous Coward on Wednesday June 03 2020, @04:53AM
Define "behind". If it is lacking "management" modules and dumbass design flaws like the one used in Spectre, etc. Then yea it is "behind". Adding features that the end buyer doesn't want or need isn't necessarily progress.