[ / / / / / / / / / / / / / ] [ dir / ausneets / funegros / just / monarchy / vg / vore / voxed / x ][Options][ watchlist ]

/tech/ - Technology

You can now write text to your AI-generated image at https://aiproto.com It is currently free to use for Proto members.
Name
Email
Subject
Comment *
File
Select/drop/paste files here
Password (Randomized for file and post deletion; you may also set your own.)
* = required field[▶ Show post options & limits]
Confused? See the FAQ.
Expand all images

File (hide): 7a899c1892a97b8⋯.jpg (243.61 KB, 1085x1086, 1085:1086, Transmeta_Efficeon_TM8600_….jpg) (h) (u)

[–]

 No.907793>>908038 >>908112 [Watch Thread][Show All Posts]

What could they have done to save themselves?

Couldn't they in-theory fork their own Code Morphing Software to simulate architectures other than x86-64 like ARM or PowerPC? Was that not the whole point of the Transmeta concept?

 No.908038>>908048 >>908077

>>907793 (OP)

should've called themselves Straightmeta cause as we all know trans have a high rate of destroying themselves


 No.908048


 No.908077>>908295

>>908038

ell em ayy oh


 No.908112>>908145

>>907793 (OP)

>Was that not the whole point of the Transmeta concept?

No, the actual CPU core was designed from start to end around x86 emulation. It would be shit for running any other ISA.


 No.908145>>908170

>>908112

>No, the actual CPU core was designed from start to end around x86 emulation. It would be shit for running any other ISA.

Then what was the purpose of using a VLIW core and a hypervisor-level translation layer? I thought the entire point of the Transmeta concept was the ability to swap "architectures" on-the-fly for compatibility with somewhat better performance than emulation


 No.908154>>908178

The whole thing went black and was purchased by a major defense contractor.


 No.908170>>908373

>>908145

The point was to get rid of all the complex scheduling silicon that tries to extract parallelism from inherently sequential code and move all that complexity to software. It was expected that this way more silicon could be dedicated to actually doing shit instead of planning to do shit, and therefore higher performance could be obtained from the same transistor budget.

Of course, like all VLIW power fantasies, this performance never materialised. Optimal scheduling depends on too many dynamic variables to be achieved ahead of time. However, Transmeta found out that their shit-performing failed abortion of a performance chip sipped far less power than competing processors. So they quickly regrouped and started marketing it as a mobile part. The rest is history.


 No.908178

>>908154

sauce?


 No.908295

>>908077

ecks dee


 No.908373>>914157

>>908170

So in other words it's immune to Spector/meltdown because the silicon is inherently in-order


 No.914157

>>908373

In-order cores can still execute code speculatively (and have to, if they aim for decent performance). Also, the x86 operations are still likely reordered as necessary. It's just that the reordering is done by the dynarec firmware rather than by hardware.




[Return][Go to top][Catalog][Screencap][Nerve Center][Cancer][Update] ( Scroll to new posts) ( Auto) 5
11 replies | 0 images | Page ???
[Post a Reply]
[ / / / / / / / / / / / / / ] [ dir / ausneets / funegros / just / monarchy / vg / vore / voxed / x ][ watchlist ]