For $3100 USD You Can Have A Fast, Fully-Free-Software Workstation

[zanny]

Free/OSS solution is a pipe dream. The CPU is an IBM ASIC right? Done, moot point etc. You have absolutely 0 control over what has been implemented in the ASIC unless you are guaranteed to have the code that was put to production for inspection. Even if you did, there is absolutely 0 guarantee of a late masking on the CPU.
Unless you produce your own hardware in your own controlled fabs with people you absolutely trust, ... well... I don't think need to explain more.

That is exactly what you do. Open architectures are a hedge against hardware hegemony, that enables consumers of the architecture to develop and fab their own alternative implementations of the architecture, that can consume and execute the same binaries.

You cannot have that security with x86, and to get it on ARM you need to pay ARM Holdings a gratuitous amount of money, and because of the nature of ARM licensing you cannot be guaranteed they will not change their licensing policies in the future just as a user of the architecture. The only way to be confident that future products will not work against your interest is the ability to replace them at will.

[torsionbar28]

Free/OSS solution is a pipe dream. The CPU is an IBM ASIC right? Done, moot point etc. You have absolutely 0 control over what has been implemented in the ASIC unless you are guaranteed to have the code that was put to production for inspection. Even if you did, there is absolutely 0 guarantee of a late masking on the CPU.
Unless you produce your own hardware in your own controlled fabs with people you absolutely trust, ... well... I don't think need to explain more.

I think you're misinterpreting FOSS concepts. The ability to alter the ASIC has nothing to do with it.

[milkylainen]

That is exactly what you do. Open architectures are a hedge against hardware hegemony, that enables consumers of the architecture to develop and fab their own alternative implementations of the architecture, that can consume and execute the same binaries.

You cannot have that security with x86, and to get it on ARM you need to pay ARM Holdings a gratuitous amount of money, and because of the nature of ARM licensing you cannot be guaranteed they will not change their licensing policies in the future just as a user of the architecture. The only way to be confident that future products will not work against your interest is the ability to replace them at will.

Sure. But in realistic terms, "normal" customers are not going to afford a fab built to ensure that hardware does not get modified.

Or you do it like China and Russia. Develop your own hardware, including CPUs. Built in locally trusted fabs.

[milkylainen]

I think you're misinterpreting FOSS concepts. The ability to alter the ASIC has nothing to do with it.

Or maybe we have different views about OSS. My ability to inspect firmware (GPLv3 etc) and in the end actually what constitutes the
hardware is how I view it.

For me, CPUs, GPUs etc, are just written in Verilog, VHDL, SystemC etc. I see no difference between hardware and code executing on it. Both are descriptions in languages. If you cannot inspect what you are running on, you cannot claim any full open source system.
I can accept that it is synthesized by a trusted partner, but never written.

For me it is the same as saying, this is a binary blob, just use it (which I don't like).
I can accept that a friend compiled it for me, but never that I was just given a blob without having source to it.

In reality I have just accepted that hardware is hardware. I run on closed source hardware just as everybody else.

[madscientist159]

Or maybe we have different views about OSS. My ability to inspect firmware (GPLv3 etc) and in the end actually what constitutes the
hardware is how I view it.

For me, CPUs, GPUs etc, are just written in Verilog, VHDL, SystemC etc. I see no difference between hardware and code executing on it. Both are descriptions in languages. If you cannot inspect what you are running on, you cannot claim any full open source system.
I can accept that it is synthesized by a trusted partner, but never written.

For me it is the same as saying, this is a binary blob, just use it (which I don't like).
I can accept that a friend compiled it for me, but never that I was just given a blob without having source to it.

In reality I have just accepted that hardware is hardware. I run on closed source hardware just as everybody else.

There is one critical distinction that I think keeps getting lost in the noise. Software is mutable, hardware (short of FPGAs and outright replacement of ASICs) is immutable. Intel could (theoretically, of course, no one knows the true capabilities of the ME) update your ME firmware remotely without your permission and maybe even without your knowledge. Good luck modifying hardware the same way; not only do most people know when their stuff leaves their possession but in most jurisdictions physically altering/replacing someone's physical property would be illegal, even for the original manufacturer.

No one's talking about hardware interception/modification, the simple fact is if you're a bad enough actor various agencies *will* find a way to get to you. The idea is more to keep malicious and snoopy types out of the majority of people's computers, and rely on the fact that silicon can't easily be modified to target a select few individuals without firmware assisting in one way or another.

Just my $0.02. :-)

[torsionbar28]

Or maybe we have different views about OSS. My ability to inspect firmware (GPLv3 etc) and in the end actually what constitutes the
hardware is how I view it.

For me, CPUs, GPUs etc, are just written in Verilog, VHDL, SystemC etc. I see no difference between hardware and code executing on it. Both are descriptions in languages. If you cannot inspect what you are running on, you cannot claim any full open source system.
I can accept that it is synthesized by a trusted partner, but never written.

For me it is the same as saying, this is a binary blob, just use it (which I don't like).
I can accept that a friend compiled it for me, but never that I was just given a blob without having source to it.

In reality I have just accepted that hardware is hardware. I run on closed source hardware just as everybody else.

I agree that inspection is critical to a FOSS system. No binary blobs. No closed source black boxes. It's got to be totally open for full inspection. And that's exactly what this Raptor system delivers. You cannot modify this POWER8 ASIC any more than you could an intel x86 ASIC. So I'm not entirely sure what it is you're claiming. You have exactly as much control over this IBM POWER8 ASIC as you do any other ASIC. I.e. unless you've got a chip fab in your basement, you are trusting that part of the process to someone else.

[CrystalGamma]

Sad that there is no really good firmware-free GPU, but since the board has IOMMU according to their preliminary specs, I'm not too worried.
Sounds good that the performance is competitive at least. I guess you didn't have access to power measurements, did you? As I said in the comments of the earlier article, I am very interested in the idle power ...

[Kano]

The CPU seems to be overpriced for the performance you get. But maybe some buy it and use Facebook with it. Even without cookies you can identify the handful owners with the unique browser ID.

[milkylainen]

I agree that inspection is critical to a FOSS system. No binary blobs. No closed source black boxes. It's got to be totally open for full inspection. And that's exactly what this Raptor system delivers. You cannot modify this POWER8 ASIC any more than you could an intel x86 ASIC. So I'm not entirely sure what it is you're claiming. You have exactly as much control over this IBM POWER8 ASIC as you do any other ASIC. I.e. unless you've got a chip fab in your basement, you are trusting that part of the process to someone else.

From the point I see it, there is nothing more to add to your statement.
What you are saying is correct. The ASIC itself is pretty much immutable (besides firmware modifiable behavior) . But you cannot account for the behavior for the ASIC. There is absolutely no guarantee that secret opcodes cannot be unlocked with opcode knocking or something else. These opcodes would give you the power to manipulate memory access or just traverse whatever you like in the highest privilege settings of the ASIC. There is no "inspection" of ASIC behavior. You just have to trust it. Ergo, black box.
This is highly incompatible with the view I have with OSS hardware.

[zuxun]

Was openssl single threaded? POWER is made for multithreading.

1 thread vs 1 thread it cannot compete with Xeon.

Many threads vs Many threads it can compete with Xeon. This is how POWER is designed.

I would know this since i am a multithreading guru.

Each POWER7 core runs 4 threads.

Each POWER8 core runs 8 threads. I don't know how many TALOS can run because there is no information in their website.

It is probable that TALOS 8 core can run 64 threads.